Abstract
Background
Low back pain (LBP) is a significant public health issue due to its high prevalence and associated disability burden. Clinical practice guidelines recommend non‐pharmacological/non‐surgical interventions for managing pain and function in people with LBP.
Objectives
To provide accessible, high‐quality evidence on the effects of non‐pharmacological and non‐surgical interventions for people with LBP and to highlight areas of remaining uncertainty and gaps in the evidence regarding the effects of these interventions for people with LBP.
Methods
We searched the Cochrane Database of Systematic Reviews from inception to 15 April 2023, to identify Cochrane reviews of randomised controlled trials testing the effect of non‐pharmacological/non‐surgical interventions, unrestricted by language. Major outcomes were pain intensity, function and safety.
Two authors independently assessed eligibility, extracted data and assessed the quality of the reviews using AMSTAR 2 (A MeaSurement Tool to Assess Systematic Reviews) and the certainty of the evidence using GRADE. The primary comparison was placebo/sham.
Main results
We included 31 Cochrane reviews of 644 trials that randomised 97,183 adults with LBP. We have high confidence in the findings of 19 reviews, moderate confidence in the findings of two reviews, and low confidence in the findings of 10 reviews. We present results for non‐pharmacological/non‐surgical interventions compared to placebo/sham or no treatment/usual care at short‐term (≤ three months) follow‐up.
Placebo/sham comparisons
Acute/subacute LBP
Compared to placebo, there is probably no difference in function (at one‐week follow‐up) for spinal manipulation (standardised mean difference (SMD) ‐0.08, 95% confidence interval (CI) ‐0.37 to 0.21; 2 trials, 205 participants; moderate‐certainty evidence).
Data for safety were reported only for heated back wrap. Compared to placebo, heated back wrap may result in skin pinkness (6/128 participants versus 1/130; 2 trials; low‐certainty evidence).
Chronic LBP
Compared to sham acupuncture, acupuncture probably provides a small improvement in function (SMD ‐0.38, 95% CI ‐0.69 to ‐0.07; 3 trials, 957 participants; moderate‐certainty evidence).
Compared to sham traction, there is probably no difference in pain intensity for traction (0 to 100 scale, mean difference (MD) ‐4, 95% CI ‐17.7 to 9.7; 1 trial, 60 participants; moderate‐certainty evidence).
Data for safety were reported only for acupuncture. There may be no difference between acupuncture and sham acupuncture for safety outcomes (risk ratio (RR) 0.68, 95% CI 0.42 to 1.10; I2 = 0%; 4 trials, 465 participants; low‐certainty evidence).
No treatment/usual care comparisons
Acute/subacute LBP
Compared to advice to rest, advice to stay active probably provides a small reduction in pain intensity (SMD ‐0.22, 95% CI ‐0.02 to ‐0.41; 2 trials, 401 participants; moderate‐certainty evidence).
Compared to advice to rest, advice to stay active probably provides a small improvement in function (SMD ‐0.29, 95% CI ‐0.09 to ‐0.49; 2 trials, 400 participants; moderate‐certainty evidence).
Data for safety were reported only for massage. There may be no difference between massage and usual care for safety (risk difference 0, 95% CI ‐0.07 to 0.07; 1 trial, 51 participants; low‐certainty evidence).
Chronic LBP
Compared to no treatment, acupuncture probably provides a medium reduction in pain intensity (0 to 100 scale, mean difference (MD) ‐10.1, 95% CI ‐16.8 to ‐3.4; 3 trials, 144 participants; moderate‐certainty evidence), and a small improvement in function (SMD ‐0.39, 95% CI ‐0.72 to ‐0.06; 3 trials, 144 participants; moderate‐certainty evidence).
Compared to usual care, acupuncture probably provides a small improvement in function (MD 9.4, 95% CI 6.15 to 12.65; 1 trial, 734 participants; moderate‐certainty evidence).
Compared to no treatment/usual care, exercise therapies probably provide a small to medium reduction in pain intensity (0 to 100 scale, MD ‐15.2, 95% CI ‐18.3 to ‐12.2; 35 trials, 2746 participants; moderate‐certainty evidence), and probably provide a small improvement in function (0 to 100 scale, MD ‐6.8, 95% CI ‐8.3 to ‐5.3; 38 trials, 2942 participants; moderate‐certainty evidence).
Compared to usual care, multidisciplinary therapies probably provide a medium reduction in pain intensity (SMD ‐0.55, 95% CI ‐0.83 to ‐0.28; 9 trials, 879 participants; moderate‐certainty evidence), and probably provide a small improvement in function (SMD ‐0.41, 95% CI ‐0.62 to ‐0.19; 9 trials, 939 participants; moderate‐certainty evidence).
Compared to no treatment, psychological therapies using operant approaches probably provide a small reduction in pain intensity (SMD ‐0.43, 95% CI ‐0.75 to ‐0.11; 3 trials, 153 participants; moderate‐certainty evidence).
Compared to usual care, psychological therapies (including progressive muscle relaxation and behavioural approaches) probably provide a small reduction in pain intensity (0 to 100 scale, MD ‐5.18, 95% CI ‐9.79 to ‐0.57; 2 trials, 330 participants; moderate‐certainty evidence), but there is probably no difference in function (SMD ‐0.2, 95% CI ‐0.41 to 0.02; 2 trials, 330 participants; moderate‐certainty evidence).
It is uncertain whether there is a difference between non‐pharmacological/non‐surgical interventions and no treatment/usual care for safety (very low‐certainty evidence).
Authors' conclusions
Spinal manipulation probably makes no difference to function compared to placebo for people with acute/subacute LBP. Acupuncture probably improves function slightly for people with chronic LBP, compared to sham acupuncture. There is probably no difference between traction and sham traction for pain intensity in people with chronic LBP. Compared to advice to rest, advice to stay active probably reduces pain intensity slightly and improves function slightly for people with acute LBP.
Acupuncture probably reduces pain intensity, and improves function slightly for people with chronic LBP, compared to no treatment. Acupuncture probably improves function slightly for people with chronic LBP, compared to usual care. Exercise therapies probably reduce pain intensity, and improve function slightly for people with chronic LBP, compared to no treatment/usual care. Multidisciplinary therapies probably reduce pain intensity, and improve function slightly for people with chronic LBP, compared to usual care. Compared to usual care, psychological therapies probably reduce pain intensity slightly, but probably make no difference to function for people with chronic LBP.
Plain language summary
What are the likely benefits and harms of non‐medicine and non‐surgical treatments for non‐specific low back pain?
Key messages
For acute low back pain (pain lasting less than 6 weeks)
• Advice to stay active probably reduces pain and improves function compared to advice to rest in bed.
For subacute low back pain (pain lasting from 6 to 12 weeks)
• Multidisciplinary therapies probably reduce pain compared to usual care. • Spinal manipulation probably does not improve function compared to placebo (a 'sham' or 'dummy' treatment designed to resemble the actual treatment but lacking active ingredients or the intended therapeutic effect).
For chronic low back pain (pain lasting longer than 12 weeks)
• Acupuncture probably reduces pain and improves function compared to placebo and no treatment/usual care. • Exercise therapies probably reduce pain and improve function compared to placebo and no treatment/usual care. • Traction probably does not reduce pain compared to sham traction. • Multidisciplinary therapy probably reduces pain and improves function compared to usual care. • Psychological therapies probably reduce pain but make no difference to function compared to usual care.
What is low back pain, and how is it treated?
Low back pain is a common health condition that can be associated with disability and poor quality of life. For most cases of low back pain, the cause of pain is unknown and is described as ‘non‐specific’ low back pain. Many types of non‐medicine and non‐surgical treatments are available for people with low back pain of different durations: acute (pain lasting less than 6 weeks), subacute (pain lasting from 6 to 12 weeks) and chronic (pain lasting longer than 12 weeks). There is a need to provide accessible, high‐quality information on the benefits and safety of non‐drug and non‐surgical treatments for healthcare professionals and patients to better manage low back pain.
What did we want to find out?
We wanted to summarise the evidence from Cochrane reviews on the effectiveness and safety of non‐medicine and non‐surgical treatments for adults with non‐specific low back pain.
What did we do? We found 31 reviews that included 644 studies with 97,183 participants. The studies investigated the effects of 27 different types of treatment for low back pain.
What did we find?
For people with acute/subacute low back pain, we found that advice to stay active probably reduces pain in the short term (i.e. up to 3 months) compared to advice to rest in bed. We found that multidisciplinary therapies probably reduce pain in the long term (i.e. at 12 months or longer). Spinal manipulation probably does not improve function in the short term.
For people with chronic low back pain, we found that acupuncture, exercise, and psychological therapies probably reduce pain in the short and medium term (i.e. from 3 to 12 months). Acupuncture and exercise probably also improve function in the short and medium term. Multidisciplinary therapies probably reduce pain and improve function in the short and medium term. Traction probably does not reduce pain in the short term.
We have less confidence in the effects of other non‐medicine and non‐surgical interventions for low back pain.
Non‐medicine and non‐surgical interventions may not be associated with serious adverse (i.e. unwanted, harmful) events.
What are the limitations of the evidence?
We have reduced confidence in the evidence because we judged that 38% of the reviews did not employ the most rigorous methods available. Almost three‐quarters of the reviews were published before 2020, meaning that the evidence they contain may be relatively dated. There is a need to update some Cochrane reviews following recommended guidance.
Because of the quality of the evidence, we are still uncertain about the benefits or risks of many non‐medicine and non‐surgical treatments commonly used in clinical practice for low back pain. We encourage healthcare professionals, patients, and organisations that fund research on low back pain to use this overview to make informed decisions for low back pain treatment.
How current is this evidence?
This overview is current to April 2023. However, one‐third (10 of 31) of the reviews are more than 15 years old, meaning the evidence they contain is even more dated.
Background
Description of the condition
Low back pain (LBP) is the most common musculoskeletal condition and is the leading cause of years lived with disability globally (Abbafati 2020). An increasing and ageing population is contributing to an increase in the number of people with LBP, leading to a significant financial impact on healthcare and support systems (Buchbinder 2018; Cieza 2020). In the USA, LBP and neck pain were responsible for the highest healthcare spending across 154 health conditions, with an estimated annual cost of 134,500 million US dollars (USD) (95% confidence interval 122,400 to 146,900 million USD) (Dieleman 2020).
Non‐specific LBP refers to cases of LBP where a specific cause of LBP (e.g. fracture, epidural abscess, malignancy, cauda equina syndrome, radiculopathy or spinal canal stenosis) cannot be identified (Hancock 2007; Maher 2017). Most acute episodes of non‐specific LBP improve substantially within the first six weeks, even with minimal intervention (Costa 2012). Up to 40% of people experience symptoms beyond three months (Henschke 2008), referred to as chronic non‐specific LBP (Van Tulder 2006). Around 30% of people who recover from LBP report a recurrence within the first year (Da Silva 2017).
Clinical practice guidelines recommend using a biopsychosocial framework to inform the management of non‐specific LBP, acknowledging the interaction of biological, psychological and social factors in shaping the LBP experience (Foster 2018; Hartvigsen 2018). Recently, recommendations have reduced emphasis on pharmacological and surgical interventions and shifted to encouraging non‐pharmacological and non‐surgical interventions such as self‐management, complementary medicine, and physical and psychological therapies as first‐line treatment options for LBP (Foster 2018).
Description of the interventions
Non‐pharmacological and non‐surgical approaches refer to interventions that do not involve using medications, injections or surgery to treat LBP. For brevity, the term non‐pharmacological will be used to cover both types of treatments in this review. Non‐pharmacological interventions for LBP include the use of (1) acupuncture and dry needling, (2) education and advice, (3) lumbar supports, (4) electrophysical modalities, (5) exercise programmes, (6) heat and cold therapies, (7) manual therapy approaches and (8) psychological therapies (Chou 2017; Van Middelkoop 2011).
Non‐pharmacological interventions can be delivered as a single intervention (e.g. exercise) or a multimodal intervention (e.g. exercise and manual therapy). The choice of non‐pharmacological intervention is influenced by a range of factors, including the health professionals’ skills, patients’ and health professionals’ knowledge, preferences and access to resources (Balagué 2012; Poder 2019; Widerström 2019). The non‐pharmacological management of LBP generally aims to improve pain, function, mental health and social participation impacted by the condition (Maher 2017), although the specific aims may vary by type of intervention.
This overview of Cochrane reviews focuses on non‐pharmacological interventions to improve pain and physical function in people with LBP.
How the intervention might work
The mechanisms by which different non‐pharmacological interventions might improve pain and function are often not fully understood and may differ across interventions (Lee 2016; Wun 2021). Commonly proposed mechanism(s) for each class of non‐pharmacological interventions are described below.
Acupuncture and dry needling
Acupuncture and dry needling involve the insertion of one or more needles at specific parts of the body to create analgesic effects (Mu 2020). Acupuncture commonly includes manual stimulation of the needles or stimulation applied with electrical current, heat lamps and moxibustion (a herb burned at the end of the needle) (Lao 1996). Dry needling involves the insertion of a needle in the myofascial tissue to deactivate supposed myofascial trigger points. There are various claims and theories regarding the potential therapeutic mechanism of acupuncture and dry needling. These include physiological mechanisms (e.g. pain gate mechanisms, the activation of descending inhibitory neural pathways, specific local actions on putative muscle “trigger points”) (Melzack 1981), and explanations based on traditional Chinese medicine concepts (Andersson 1995).
Advice and education
Patient education involves sharing knowledge or information with a patient and providing guidance. Patient education can include information on diagnosis, prognosis and the management of LBP (Furlong 2020; Traeger 2015), as well as guidance on self‐management. Patient education may also involve the provision of more complex information, such as teaching patients about the anatomy of the back (Heymans 2004), the neurobiology of pain (Moseley 2015), or the interaction between thoughts, feelings and behaviours (Hansen 2010). Patient education aims to change patients’ attitudes, beliefs, behaviours and motivations that may influence pain and function (Borrell‐Carrio 2004; Furlong 2020).
Assistive materials
Assistive materials involve using devices to change the posture of the body, adapt the worker to the work environment (e.g. lumbar braces, insoles and manual handling aids), or both (Verbeek 2011). Assistive materials are assumed to reduce pain and disability by limiting spinal movement and relieving load and strain on the back (Van Duijvenbode 2008; Verbeek 2011).
Electrophysical modalities
Electrophysical modalities involve using electrical, magnetic and light energy as a treatment. Examples of electrophysical modalities include transcutaneous electrical nerve stimulation (TENS), interferential therapy, non‐invasive brain stimulation, electromagnetic field therapy or laser therapy (Milne 2001; O'Connell 2010; Van Middelkoop 2011). TENS has been proposed to alter the excitability of peripheral nociceptors, stimulate inhibitory interneurons in the spinal cord, and activate supraspinal descending inhibitory neural pathways (Vance 2014). Brain stimulation techniques seek to modulate activity in the areas of the brain involved in pain processing (O'Connell 2010). Low‐level laser therapy is proposed to modulate peripheral neurotransmitters, exerting anti‐inflammatory effects and reducing muscle spasms in the back, leading to reduced pain (Yousefi‐Nooraie 2008).
Exercise therapies
Exercise is defined as planned, structured and repetitive physical activity intended to improve or maintain one or more components of physical fitness (Caspersen 1985). Exercise therapies that have been tested for their effects on LBP include stretching, strengthening and motor control training for the spine; general resistance training; aerobic exercise; Pilates and yoga; tai chi; water‐based exercises; and graded exercise or a combination of two or more types of exercises (multimodal approaches) (Owen 2020). The mechanisms through which exercise improves pain and function may depend on the type of exercise (Smith 2019; Wun 2021). These hypothesised mechanisms include improving flexibility, strength, endurance, motor control and performance in certain activities (Hodges 2003; Rainville 2004), modulating central nervous system activity and immune system function (Sluka 2018), and targeting psychosocial factors (e.g. fear, self‐efficacy and worries) commonly associated with the LBP experience (Gilanyi 2023; Sluka 2018).
Heat and cold therapies
Heat and cold therapies aim to modulate the temperature of body tissues for pain management. Heat therapy includes superficial and deep tissue heating and is assumed to decrease stiffness, reduce muscle spasms, increase blood flow and reduce inflammation (Lewis 2012). Examples of superficial heating include hot water bottles, heated stones and heat wraps, and examples of deep heating include microwave diathermy and therapeutic ultrasound (Ebadi 2020; French 2006). Cold therapy uses ice, cold towels or cold gel packs to reduce the temperature of the tissues, aiming to decrease pain through the reduction of inflammation, oedema and nerve conduction velocity (Algafly 2007; French 2006).
Manual therapies
Manual therapy involves passive movements applied by health professionals intended to benefit musculoskeletal and neurophysiological systems (Bishop 2015). Several manual therapy techniques have been proposed for managing LBP, including manipulation and mobilisation, mechanical or manual traction, massage and neural mobilisation. Manipulation and mobilisation are proposed to work via biomechanical (e.g. reduce internal mechanical stress and improve spinal range of motion) and neurophysiological (e.g. pain gate mechanisms and motor control system) mechanisms (Bialosky 2009; Rubinstein 2019). Mechanical or manual traction has been claimed to increase intervertebral space, inhibit nociception impulses, increase mobility, reduce mechanical stress and revert dysfunctional neuromuscular reflex patterns (Clarke 2006). Massage therapies (e.g. deep tissue, trigger point) are intended to improve circulation, decrease muscle spasms, provide muscle relaxation and increase the range of motion (Bialosky 2009). Neural mobilisation combines movements of the spine and extremities to improve the range of motion, decrease sensitivity and restore homeostasis in the nervous system (Coppieters 2008).
Psychological therapies
Psychological and psychologically informed therapies are a broad family of therapy techniques designed to target psychological factors associated with LBP (Hoffman 2007). Several psychological therapies are used to treat LBP (Pincus 2013); the most common originate from behavioural or cognitive‐behavioural traditions of clinical psychology (Williams 2020). These interventions are believed to operate by targeting mechanisms proposed to shape and maintain behaviour and function (Williams 2020), such as cognition, beliefs, self‐efficacy, pain reactivity, cognitive flexibility and pain behaviours (Burns 2016).
Why it is important to do this overview
Non‐pharmacological interventions are recommended as first‐line care for LBP (Foster 2018; Oliveira 2018). There are many non‐pharmacological interventions available to healthcare professionals (Ahern 2019). People with LBP, clinicians, researchers and healthcare policy‐makers require accessible, high‐quality information on the strength of the evidence for the efficacy, effectiveness and safety of non‐pharmacological interventions for making informed healthcare decisions (Chou 2018; Gopalakrishnan 2013; Lim 2019). Cochrane reviews have investigated the effects of non‐pharmacological interventions and are available to decision‐makers through the Cochrane Library. However, there are multiple overlapping reviews, varying in currency, scope and methodology. This may inhibit decision‐makers' access to and use of this evidence.
A single accessible overview for research consumers is required to improve access to high‐quality information and to describe the non‐pharmacological evidence base, including evidence gaps, for treating LBP (Hunt 2018).
Objectives
To provide accessible, high‐quality evidence on the effects of non‐pharmacological and non‐surgical interventions for people with LBP.
To highlight areas of remaining uncertainty and gaps in the evidence regarding the effects of these interventions for people with LBP.
Methods
This overview was conducted in accordance with guidance in Chapter V of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2024). A protocol for this overview was published in the Cochrane Database of Systematic Reviews (Cashin 2021a).
Criteria for considering reviews for inclusion
Types of studies
We included systematic reviews of randomised controlled trials published in the Cochrane Library. We included only Cochrane systematic reviews due to evidence demonstrating that Cochrane reviews have higher methodological quality and lower overlap in the included studies, which increases confidence in the findings (Higgins 2024). We excluded reviews that included both randomised and non‐randomised designs unless the data for the randomised designs were available separately. We excluded Cochrane systematic reviews withdrawn from the Cochrane Library.
Types of participants
We included systematic reviews of trials in adults, 18 years or older, with LBP without a specific cause of LBP (e.g. fracture, infection, neoplasm, metastasis, epidural abscess, malignancy, cauda equina syndrome, radiculopathy or spinal canal stenosis) of any duration (Maher 2017). LBP was defined as a primary area of pain between the twelfth rib and gluteal fold, with or without associated leg pain (Koes 2007). We excluded systematic reviews of trials in participants with mixed pain conditions and reviews that included studies of participants younger than 18 years unless they reported separate results for the studies of participants 18 years and older.
Types of intervention and comparison
We included systematic reviews of non‐pharmacological and non‐surgical interventions intended to improve pain and function for people with LBP. We included three categories of comparisons:
non‐pharmacological and non‐surgical intervention versus sham/placebo interventions;
non‐pharmacological and non‐surgical intervention versus no intervention (including waiting‐list control and continuation of usual care);
non‐pharmacological and non‐surgical interventions versus other interventions (a different type of non‐pharmacological, pharmacological or surgical intervention).
We defined multimodal therapies as a combination of two types of non‐pharmacological interventions delivered by the same clinician. Note that this is a heterogeneous category and might include physical and psychological interventions or at least two physical interventions involving different classes of treatment. We defined a multidisciplinary therapy as an intervention that involves a physical and at least one other element from the biopsychosocial model (e.g. psychological, social, occupational) delivered by multiple clinicians from two or more professional backgrounds.
We planned to specify the type of comparison in the results section (e.g. sham, placebo, no treatment, usual care or other interventions). When it was not possible to determine if the comparison referred to no treatment or usual care, we used the combined term 'no treatment/usual care'.
Types of outcomes measure
The outcomes for this overview reflect the core outcome set for non‐specific LBP (Chiarotto 2018), and the recommendations of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) (Dworkin 2005). We did not restrict the time of measurement. We grouped outcomes into a short‐term period (≤ three months post‐randomisation), an intermediate‐term period (> three months and ≤ 12 months post‐randomisation), and a long‐term period (> 12 months post‐randomisation). We included the outcome measure closest to the midpoint of the period in cases where outcome data were available for multiple time points within a period. We excluded systematic reviews that did not report our prespecified major and minor outcomes.
Major outcomes
Pain, defined as pain intensity, measured with a continuous self‐report scale (e.g. a visual analogue scale (VAS), numerical rating scale (NRS), the Brief Pain Inventory (BPI) or other validated measure), or in a dichotomous format (e.g. as the proportion of participants in each group who attained a predetermined threshold of improvement).
Physical function, defined as back pain‐related function, measured with a continuous self‐report scale (e.g. Roland‐Morris Disability Questionnaire (RMDQ), Oswestry Disability Index (ODI)), functional testing protocols, or other validated quantitative measures.
Safety, defined as adverse effects including, amongst others, incidence and severity of adverse effects, trial withdrawal due to adverse effects, and incidence of serious adverse effects, as described in the systematic review.
Minor outcomes
Participant ratings of improvement, defined as global perceived effect, measured with a validated tool.
Health‐related quality of life, measured with a validated tool (e.g. the 36‐Item Short Form Health Survey (SF‐36)).
Workplace participation, measured as days to return to work, days of absenteeism or days of reduced work activities.
Search methods for identification of reviews
Electronic searches
We searched the Cochrane Database of Systematic Reviews (The Cochrane Library) using a combination of Medical Subject Headings (MeSH) and keywords (Appendix 1), without restriction, up to 15 April 2023. We managed citations using EndNote and Covidence (Covidence).
Data collection and analysis
Selection of reviews
Two authors (RRNR and AGC) independently screened the title and abstract of each citation against the inclusion criteria. We obtained the full texts of the records meeting the eligibility criteria and screened them again to confirm inclusion in the review. We planned to use a third overview author to resolve disagreements when the two screening authors could not reach a consensus, but this was not required. We provide a PRISMA flow chart describing the screening and review selection process in Figure 1.
1.

Review flow diagram
Data extraction and management
We (RRNR, AGC, BMW) designed, piloted and reviewed a data extraction form before extracting data for the overview. At least two overview authors (RRNR and AGC, MCF, EOH or SS) independently extracted data using the finalised data extraction form. We resolved discrepancies through discussion when necessary. We planned to involve a third independent author to resolve disagreements; however, no unresolved disagreements occurred between the data extractors. We used the data extraction form to capture the following review characteristics.
Objectives of the review
Date of publication
Date of most recent search
Resources searched
Number of included trials
Characteristics of the included participants (e.g. duration of pain, pain severity, sex, age, race/ethnicity, comorbidities, prior treatment history)
Description of interventions and comparisons
Outcomes and time points assessed
Comparisons performed and meta‐analysis details
Assessment of risk of bias or methodological quality of the included trials
Assessment of the certainty of included evidence
When data were presented in tabular or figure format, we extracted data from the included reviews. We planned to contact review authors via email if we could not extract the required information from the reports, but this step was unnecessary.
Assessment of methodological quality of included reviews
Confidence in the included reviews
We used the AMSTAR 2 (A MeaSurement Tool to Assess Systematic Reviews) instrument to rate our confidence in the methods of the included systematic reviews (Shea 2017) (Appendix 2). The 16 items in the AMSTAR 2 instrument provide a broad assessment of systematic review quality. Taken together, they inform a judgement about the degree of confidence to have in the review findings. At least two authors (RRNR and AGC, MCF, EOH or SS) independently assessed the methodological quality of the included reviews to arrive at a confidence rating for each. We resolved discrepancies through consensus. We consulted a third author if disagreements persisted.
We formed an overall judgement about our confidence in each included systematic review by considering seven items of the AMSTAR 2 instrument that are most likely to affect the validity of a review (Shea 2017). The seven ‘critical’ items are: protocol registered before the commencement of the review (item 2); adequacy of the literature search (item 4); justification for excluding individual studies (item 7); risk of bias from individual studies being included in the review (item 9); appropriateness of meta‐analytical methods (item 11); consideration of the risk of bias when interpreting the results of the review (item 13); and assessment of the presence and likely impact of publication bias (item 15). For publication bias (item 15), we considered only reviews that conducted quantitative analyses. If reviews included only a few studies in meta‐analysis, graphical and statistical methods are not appropriate for assessing potential publication bias. In these cases, we assessed whether authors utilised alternative methods to assess for publication bias (e.g. thorough literature searches, the inclusion of trial registries, and the use of sensitivity analyses to determine how many missing ‘null’ studies would be needed to change the interpretation of the results) and whether authors addressed the potential impact of publication bias in their interpretation and discussion of the results (Shea 2017). Our judgements followed the AMSTAR 2 recommendations described by Shea 2017. We regarded the remaining nine items as non‐critical elements. We defined our judgements of confidence as follows:
High overall confidence in the results of the review if there were no or one non‐critical weakness: the systematic review provides an accurate and comprehensive summary of the results of the available studies that address the question of interest.
Moderate if there was more than one non‐critical weakness: the systematic review has more than one weakness but no critical flaws. It may provide an accurate summary of the results of the available studies that were included in the review.
Low if there was one critical flaw with or without non‐critical weaknesses: the review has a critical flaw and may not provide an accurate and comprehensive summary of the available studies that address the question of interest.
Critically low if there was more than one critical flaw with or without non‐critical weaknesses: the review has more than one critical flaw and should not be relied on to provide an accurate and comprehensive summary of the available studies.
Risk of bias in primary studies included in reviews
We reported the risk of bias assessments for the primary studies described in each included review. We reported the risk of bias tool used, the dimensions assessed (e.g. allocation concealment, participant blinding) and the number of 'high', 'moderate', and 'unclear' risks of bias for each domain of the included reviews. We did not repeat or update the reviews’ assessments.
Certainty of evidence in the included reviews
We reported the Grading of Recommendations Assessment, Development and Evaluation (GRADE) judgement of certainty for each core comparison for our primary outcomes (Balshem 2011), as reported in the included systematic reviews.
We independently judged the certainty of evidence using the GRADE approach only when GRADE judgements were not reported in the original systematic review (Meader 2014). The GRADE approach uses five considerations (risk of bias, consistency of effect, imprecision, indirectness and publication bias) to assess the certainty of evidence for each outcome. Two overview authors (RRNR and AGC, MCF, EOH or SS) independently extracted the GRADE assessments for each systematic review or conducted the judgement when necessary. We resolved discrepancies through consensus, and involved a third author to arbitrate, when necessary. We used the following GRADE criteria in our judgements and reported the reasons for downgrading to facilitate the interpretation:
Serious study limitations: we downgraded once if less than 50% of the study population was at low risk of bias across all risk of bias criteria.
Inconsistency: we downgraded once if point estimates varied widely across studies, confidence intervals showed minimal or no overlap, statistical tests for heterogeneity were statistically significant or the I2 statistic was greater than 50%.
Indirectness: we downgraded once if more than 50% of participants were outside the target group.
Imprecision: we downgraded once if there were fewer than 400 participants for continuous outcomes and fewer than 300 events for dichotomous data, or based on the width of the confidence intervals (for continuous variables such as pain intensity and function), crossing either the null or the threshold for a clinically meaningful effect of 10 points on a 0 to 100 scale. We downgraded by two levels if the confidence interval spanned both the proposed clinically meaningful effect and the null effect.
Publication bias: we downgraded once if there was direct evidence of publication bias or if estimates of effect based on small‐scale, industry‐sponsored studies raised suspicion of publication bias.
We considered single studies both inconsistent and imprecise (unless the sample size was greater than 400 participants for continuous data and greater than 300 events for dichotomous data). We downgraded one additional level due to imprecision if the pooled sample size was smaller than 50 participants per arm because evidence suggests that small published studies may be associated with positive results due to publication biases and the greater possibility of producing spurious effects due to chance (Nuesch 2010).
To judge the certainty of evidence from narrative reporting, we reviewed the reasons for not pooling the effects and used the same GRADE criteria described above. Similar decisions have been used in a previous Cochrane overview that reported the certainty of the evidence of narrative effects (O'Connell 2013).
GRADE judgements indicate the following degree of certainty in the conclusions of the systematic review.
High: very certain that the true effect lies close to that of the estimate of the effect.
Moderate: moderately certain in the effect estimate – the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low: certainty in the effect estimate is limited – the true effect may be substantially different from the estimate of the effect.
Very low: very little certainty in the effect estimate, and the true effect is likely to be substantially different from the effect estimate.
Overlap between reviews
Assessing the overlap of randomised controlled trials in the included systematic reviews provides information about the independence of the findings for each intervention and low back pain duration in the overview (Hennessy 2020; Pieper 2014). The overlap calculation and reporting between reviews followed the recommended guidance (Hennessy 2020; Pieper 2014). We examined the degree of overlap of primary studies in the included reviews using a citation matrix of the primary studies (rows) included in each review (columns) to calculate the corrected covered area (CCA) (Pérez‐Bracchiglione 2022; Pieper 2014). We considered a CCA lower than 5% to indicate slight overlap, 5% to 10% as moderate overlap, 10% to 15% as high overlap, and more than 15% as very high overlap (Pieper 2014). We extracted data from the most recent systematic review to manage the overlap in primary studies in the reviews that reported results for the same intervention class and pain duration but with different years of publication. We also presented the CCA for each pair of systematic reviews using the Graphical Representation of Overlap for Overviews to transparently report separate contributions for the total overlap in the overview (Pérez‐Bracchiglione 2022).
Data synthesis
We presented data for the primary and secondary outcomes for each intervention stratified by pain duration (acute or subacute, chronic, mixed pain duration), comparison groups (sham/placebo first, followed by no treatment/usual care and other interventions) and follow‐up period (short‐term, followed by intermediate‐ and long‐term). In the summary of findings tables, we presented results by their level of certainty (i.e. high‐certainty evidence first, followed by moderate‐, low‐ and very low‐certainty evidence). In the ‘Overview of reviews’ tables, we presented results by pain duration, comparison groups and alphabetic order. We presented narrative descriptions of results only where statistical outcome data were not available. We reported separate data for LBP durations: acute or subacute (0 to 12 weeks), chronic (> 12 weeks), and mixed (multiple symptom durations).
We did not conduct any novel statistical synthesis of the effects of interventions or make any indirect comparisons. We converted effect sizes for pain intensity reported on 0 to 10 pain intensity scales to a common scale (0 to 100 scale) to facilitate interpretation.
We classified the size of the mean between‐group difference for pain and function outcomes based on the definitions from the American College of Physicians and the American Pain Society (Chou 2017).
Large effect: ≥ 20 points on a 0 to 100 scale or ≥ 0.8 standardised mean difference (SMD)
Medium effect: > 10 to 20 points on a 0 to 100 scale or > 0.5 to 0.8 SMD
Small effect: 5 to 10 points on a 0 to 100 scale or 0.2 to 0.5 SMD
Little to no difference: < 5 points on a 0 to 100 scale or < 0.2 SMD
No clear evidence of difference: the boundaries of the 95% confidence interval span both sides of the line of no effect
Harmful: boundaries of the 95% confidence interval fall completely within harm
To express the certainty of the evidence, we adopted the following language conventions (Santesso 2020).
'Treatment X reduces/increases…' reflects high‐certainty evidence.
'Treatment X probably reduces/increases…' reflects moderate‐certainty evidence.
'Treatment X may reduce/increase…' reflects low‐certainty evidence.
'It is uncertain whether treatment X reduces/increases…' reflects very low‐certainty evidence.
We present the results of the interventions compared to sham/placebo and no treatment/usual care in the main text. We report the results of the interventions compared to other treatments in Appendix 3.
We summarised the short‐term efficacy results (versus sham/placebo and no treatment/usual care) for pain intensity in the summary of findings tables, as described in Chapter V of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2024). In the 'Overview of reviews' tables, we summarised the short‐term efficacy results for the remaining outcomes (function, safety, participant ratings of improvement, quality of life and workplace participation). We also included a ‘Summary of results' table, as proposed by McClurg 2016, highlighting the effect sizes and certainty of the evidence by considering both the effect size, GRADE score for the primary outcomes at each time point and the reviews’ methodological quality.
Results
The search of the Cochrane Library Issue 4 2023 identified 1577 records. After removing duplicates, we excluded 1510 records by screening titles and abstracts, and excluded a further 21 records after full‐text assessment (Figure 1). The reasons for excluding full‐text articles were: ineligible outcomes (seven), ineligible population (seven), ineligible study design (one), review withdrawal (four), and Cochrane reviews updated (two) (Appendix 4). One of the updated Cochrane reviews has been replaced with two separate reviews, which we included in this overview. We identified 10 review protocols as potentially eligible for inclusion in future overview updates, and described them in Table 1. We included 31 reviews in the overview (Almeida 2023; Dahm 2010; Ebadi 2020; Engers 2008; Franke 2015; French 2006; Furlan 2005; Furlan 2015; Hayden 2005; Hayden 2021a; Henschke 2010; Kamper 2014; Khadilkar 2008; Macedo 2016; Marin 2017; Mu 2020; Oltean 2014; Parreira 2017; Parry 2019; Poquet 2016; Rubinstein 2011; Rubinstein 2012; Saragiotto 2016; Schaafsma 2013; Urrutia 2004; Van Duijvenbode 2008; Walker 2010; Wegner 2013; Wieland 2022; Yamato 2015; Yousefi‐Nooraie 2008).
1. Details of ongoing reviews.
| Study ID | Review aim | Dates/notes |
| Aas 2009 | To determine the effectiveness of workplace interventions in working‐age adults with low back pain. | Protocol published 07 October 2009 |
| Furlan 2011 | To determine the effects of acupuncture for (sub)acute non‐specific low back pain compared to no treatment, sham/placebo acupuncture or other sham procedures, other therapies, the addition of acupuncture to other therapies, and between various techniques of acupuncture. | Protocol published 10 August 2011 |
| IJzelenberg 2011 | To assess the effectiveness of exercise therapy on pain, functional status, recovery, general health and return to work at short, intermediate and long‐term follow‐up as compared to no treatment and other conservative treatments in adults with non‐specific acute low back pain. | Protocol published 05 October 2011 |
| Machado 2012 | To evaluate the effectiveness of the McKenzie method in people with chronic non‐specific low back pain. | Protocol published 14 March 2012 |
| Liu 2013 | To assess the effects of sling exercise therapy on pain intensity, back‐specific pain‐related disability, overall improvement, patient satisfaction with treatment and return to work for patients with chronic low back pain. | Protocol published 01 September 2013 |
| Odebiyi 2013 | To determine the effectiveness of TENS (transcutaneous electrical nerve stimulation) in the management of chronic low back pain. "We will compare TENS (either alone or in combination with another treatment) to placebo, no treatment, or other interventions for chronic LBP and evaluate outcomes in terms of pain intensity, back‐specific functional status, general health, work disability, and patient satisfaction. A secondary objective is to determine the most effective dosage and intensity of TENS for chronic LBP." | Protocol published 30 April 2013 |
| Jesus‐Moraleida 2016 | To evaluate the effectiveness of exercise therapy to improve pain and/or functional performance in older people with non‐specific low back pain compared to no treatment and other conservative treatments. | Protocol published 14 April 2016 |
| Almeida 2020 | To evaluate the benefits and harms of telerehabilitation for people with acute, subacute or chronic non‐specific low back pain, compared to usual care, no treatment, waiting list, any form of education and advice (remotely or face‐to‐face), or a similar face‐to‐face intervention on pain intensity and disability. | Protocol published 10 August 2020 |
| Sridhar 2021 | To assess the effects of different models of evidence‐based care for people with non‐specific low back pain on self‐reported outcomes. | Protocol published 09 December 2021 |
Description of included reviews
We present the main characteristics of the included reviews in Table 2. The 31 reviews included 644 unique randomised controlled trials (RCTs), ranging from three RCTs in Urrutia 2004 to 249 RCTs in Hayden 2021a. The median year of review publication was 2014 and ranged from 2004 to 2023. All reviews searched at least one trial registry database (e.g. Cochrane Central Register of Controlled Trials (CENTRAL); ClinicalTrials.gov) for potential RCTs, but none of the reviews included studies published only in trial registries.
2. Characteristics of the included reviews.
| Review | Date of last search | Total number of participants (RCTs) | Population | Interventions | Comparisons | Outcomes planned |
| Urrutia 2004 | July 2009 | 273 (3 RCTs) | Adults aged 16 to 65 years old; non‐specific LBP that was not relieved with conservative treatment |
Neuroreflexotherapy | Sham/placebo; standard care | Primary:
Secondary:
|
| Hayden 2005 | April 2005 | 2357 (17 RCTs) | Adults with acute, subacute, or chronic non‐specific LBP | Exercise | Placebo or no treatment; other conservative therapy; another exercise | Primary:
Secondary: not reported |
| Furlan 2005 | February 2003 | 1177 (12 RCTs) | Adults (>18 years) with (sub)acute or chronic non‐specific LBP and myofascial pain syndrome in the low back region |
Acupuncture and dry needling | Sham/placebo acupuncture; no treatment, other sham procedure or other therapeutic interventions and types of acupuncture. | Primary:
Secondary:
|
| French 2006 | October 2005 | 856 (5 RCTs) | Adults aged 18 years or over with acute, sub‐acute or chronic non‐specific LBP |
Superficial heat or cold | Placebo; no therapy; other therapies | Primary:
Secondary:
|
| Khadilkar 2008 | July 2007 | 585 (4 RCTs) | Adults aged 18 years and over with chronic LBP pain (> 12 weeks) | Transcutaneous electrical nerve stimulation (TENS) | Placebo | Primary:
Secondary:
|
| Engers 2008 | July 2006 | 7139 (24 RCTs) | Adults aged 16 years of age or older, with acute, subacute, or chronic non‐specific LBP |
Individual patient education | No intervention; individual patient education added to usual care versus usual care only; non‐educational interventions; other types of patient education | Primary:
Secondary: not reported |
| Van Duijvenbode 2008 | December 2006 | 1361 (8 RCTs) | Adults aged 18 to 65 years with acute, subacute, or chronic non‐specific LBP | Lumbar support | No intervention; other types of treatment; lumbar support added to another treatment versus that other treatment; different types of lumbar support | Primary:
Secondary: not reported |
| Yousefi‐Nooraie 2008 | November 2007 | 384 (7 RCTs) | Adults aged 18 years and over with acute, subacute or chronic LBP |
Low‐level laser therapy (LLLT) | Sham treatment; LLLT added to exercise versus exercise; other treatments | Primary:
Secondary:
|
| Dahm 2010 | May 2009 | 1923 (7 RCTs) | Adults between 16 and 80 years of age with acute LBP or exacerbations of chronic pain lasting less than six weeks |
Advice to rest in bed or advice to stay active | Advice to stay active or continue normal daily activities; other treatments; short bed rest versus longer bed rest | Primary:
Secondary: not reported |
| Henschke 2010 | February 2009 | 3438 (RCTs 30) | Adults between 18 and 65 years of age with back pain that had persisted for 12 weeks or more |
Behavioural therapy (operant treatments; cognitive treatment; respondent treatment; combined treatment approaches) | Placebo, no treatment/waiting list; different types of behavioural treatment; behavioural treatment versus other kinds of treatment; behavioural treatment added to another treatment versus the other treatment alone | Primary:
Secondary: not reported |
| Walker 2010 | May 2009 | 2887 (12 RCTs) | Adults aged 18 years or over with non‐specific LBP | Chiropractic (combination of therapies such as manual therapy, thermotherapy, electrotherapy, exercise programs, nutritional advice, orthotics, lifestyle modification and patient education) | No treatment; other therapies | Primary:
Secondary: not reported |
| Rubinstein 2011 | July 2012 | 6070 (26 RCTs) | Adults aged > 18 years of age with low back pain longer than 12 weeks (studies with more than 50% of this study population) |
Spinal manipulation and mobilisation (SMT) | Sham SMT; inert interventions; other interventions; SMT added to another intervention versus that intervention alone | Primary:
Secondary:
|
| Rubinstein 2012 | July 2012 | 2674 (20 RCTs) | Adults > 18 years of age with < six weeks LBP with or without radiating pain |
Spinal manipulation or mobilisation | Sham SMT; inert interventions; other interventions; SMT added to another intervention versus that intervention alone; SMT versus another SMT technique | Primary:
Secondary:
|
| Wegner 2013 | August 2012 | 2762 (32 RCTs) | Adults aged 18 years or older with acute, subacute or chronic LBP with or without sciatica | Traction | Sham/placebo;no treatment; physiotherapy without traction; another type of traction; other treatment | Primary:
Secondary:
|
| Schaafsma 2013 | March 2012 | 4404 (25 RCTs) | Adults > 16 years with work disability related to back pain with acute, subacute or chronic back pain |
Physical conditioning program | Usual care; other treatment | Primary:
Secondary: not reported |
| Oltean 2014 | September 2014 | 2050 (14 RCTs) | Adults aged over 18 years with acute, subacute or chronic non‐specific LBP |
Herbal therapy | Placebo; other herbal therapy; conventional treatment | Primary:
Secondary: not reported |
| Kamper 2014 | March 2014 | 6858 (41 RCTs) | Adults older than 18 years of age with non‐specific chronic LBP |
Multidisciplinary biopsychosocial rehabilitation | Waiting list; usual care; physical treatment; surgery | Primary:
Secondary:
|
| Furlan 2015 | August 2014 | 3096 (25 RCTs) | Adults older than 18 years with acute, subacute or chronic non‐specific LBP |
Therapeutic massage | Inactive (e.g. sham therapy, no treatment, waiting list); active (e.g. manipulation, mobilisation, TENS) | Primary:
Secondary: safety |
| Franke 2015 | June 2014 | 500 (12 RCTs) | Adults older than 18 years with non‐specific LBP |
Muscle energy technique (MET) | Sham; no treatment; other therapies; MET plus any intervention versus that same intervention alone | Primary:
Secondary:
|
| Yamato 2015 | June 2015 | 510 (10 RCTs) | Adults aged 16 or older with acute, subacute or chronic non‐specific LBP |
Pilates | Minimal intervention or no intervention; other types of exercises | Primary:
Secondary:
|
| Macedo 2016 | April 2015 | 197 (3 RCTs) | Acute or subacute non‐specific LBP with or without leg pain (studies that include > 75% of this pain duration) |
Motor control exercise | Spinal manipulative therapy; other exercise; medical management | Primary:
Secondary:
|
| Poquet 2016 | August 2015 | 643 (4 RCTs) | Adults aged between 18 and 70 years with acute or subacute non‐specific LBP |
Back schools | Placebo (sham or attention control); no treatment; another treatment | Primary:
Secondary:
|
| Saragiotto 2016 | April 2015 | 2431 (29 RCTs) | Chronic non‐specific LBP with or without leg pain or recurrent LBP |
Motor control exercise (MCE) | Placebo; no treatment; another active treatment; MCE added to other interventions | Primary:
Secondary:
|
| Parreira 2017 | November 2016 | 4105 (30 RCTs) | Adults aged 18 to 70 years with chronic non‐specific LBP |
Back schools | No treatment; medical care; passive physiotherapy; exercises | Primary:
Secondary:
|
| Marin 2017 | July 2016 | 981 (9 RCTs) | Adults between 18 and 65 years of age with current episode greater than six weeks and less than 12 weeks of non‐specific LBP |
Multidisciplinary biopsychosocial rehabilitation | Usual care; other treatment | Primary:
Secondary:
|
| Parry 2019 | January 2019 | 955 (10 RCTs) | Adult workers aged 18 or older, working in sedentary occupations (workers sedentary for more than 50% of the working day) |
Interventions to reduce or break up workplace sitting by encouraging standing or walking at the workplace; interventions targeted at the physical work environment (sit‐stand workstation, treadmill workstation, activity tracker, multi‐component interventions) | Waiting list control/no intervention; written information | Primary:
Secondary:
|
| Ebadi 2020 | January 2020 | 1025 (10 RCTs) | Adults with chronic non‐specific LBP |
Therapeutic ultrasound | Placebo; other interventions | Primary:
Secondary:
|
| Mu 2020 | 30 August | 8270 (33 RCTs) | Adults aged 18 years and older with chronic non‐specific LBP | Acupuncture and dry needling | Sham; no treatment; usual care; other therapies; acupuncture added to another intervention versus another intervention alone; other acupuncture techniques | Primary:
Secondary:
|
| Hayden 2021a | December 2020 | 24,486 (249 RCTs) | Adults with chronic non‐specific LBP |
Exercise | No treatment/ usual care/ placebo; other conservative therapy; another exercise | Primary:
Secondary:
|
| Wieland 2022 | August 2021 | 2223 (21 RCTs) | Adults aged 18 years or older with current chronic non‐specific LBP |
Yoga | Sham; no treatment or waiting list; another active treatment; yoga added to another intervention vs another intervention alone | Primary:
Secondary: work disability |
| Almeida 2023 | August 2022 | 563 (5 RCTs) | Adults aged 18 years or greater with (sub)acute non‐specific LBP | McKenzie method exercise | Minimal intervention (waiting‐list control, placebo, or inert controls); other methods; other types of exercise therapy; all other interventions | Primary:
Secondary:
|
LBP: low back pain; RCT: randomised controlled trial
Declarations of interest and funding
All reviews provided a declaration of interest statement. None of the reviews reported that the authors’ interests would affect the review process. None of the reviews reported any direct funding that would represent a conflict of interest in conducting the systematic review. Nineteen of the 31 reviews reported the funding sources of the included RCTs (Ebadi 2020; French 2006; Furlan 2005; Furlan 2015; Hayden 2021a; Macedo 2016; Marin 2017; Mu 2020; Oltean 2014; Parreira 2017; Parry 2019; Poquet 2016; Rubinstein 2011; Rubinstein 2012; Saragiotto 2016; Van Duijvenbode 2008; Walker 2010; Wieland 2022; Yamato 2015).
Outcomes and time points
All but one review included pain and disability as the primary outcomes: Schaafsma 2013 included workplace participation as the primary outcome. Eight reviews did not include safety outcomes (Ebadi 2020; Engers 2008; Hayden 2005; Oltean 2014; Rubinstein 2011; Rubinstein 2012; Schaafsma 2013; Van Duijvenbode 2008). Only two reviews provided clear definitions regarding the measurement of adverse events in the included trials (Hayden 2021a; Wieland 2022). Although all reviews provided details of the time points for pain and disability outcome assessments, most did not report the time point for the assessment of adverse events.
Analyses
Seventy‐seven percent of included reviews (24/31) planned to conduct subgroup analyses, but only 35% (11/31) performed at least one of the planned analyses (Dahm 2010; Furlan 2005; Hayden 2005; Hayden 2021a; Henschke 2010; Kamper 2014; Parreira 2017; Rubinstein 2011; Rubinstein 2012; Schaafsma 2013; Yousefi‐Nooraie 2008). The reasons for not conducting subgroup analyses were incomplete and insufficient data (e.g. no separate data for acute and chronic LBP and types of interventions, or few studies). Sixty‐five percent of included reviews (20/31) planned sensitivity analyses, but only 29% (9/31) performed at least one prespecified sensitivity analysis (Dahm 2010; Hayden 2021a; Henschke 2010; Kamper 2014; Parreira 2017; Schaafsma 2013; Van Duijvenbode 2008; Wegner 2013; Wieland 2022). The most common reason for conducting sensitivity analyses was to investigate the extent to which the risk of bias of included studies influenced the results of the meta‐analyses (Dahm 2010; Hayden 2021a; Henschke 2010; Kamper 2014; Parreira 2017; Schaafsma 2013; Van Duijvenbode 2008; Wegner 2013; Wieland 2022). The reasons for not conducting sensitivity analyses were incomplete and insufficient data (e.g. no studies at low risk of bias in the analyses and few studies). Forty‐two percent of the included reviews (13/31) discussed the presence of publication bias. Thirty‐two percent (10/31) investigated publication bias using funnel plots (Franke 2015; Furlan 2015; Hayden 2021a; Kamper 2014; Marin 2017; Parreira 2017; Rubinstein 2011; Rubinstein 2012; Saragiotto 2016; Wieland 2022), and 10% (3/31) suggested that the presence of publication bias in the systematic reviews was due to inconsistent data reporting (i.e. primary outcomes and follow‐ups) between protocol/methods and the results in the primary studies, or many published small trials identified in the review (Almeida 2023; Schaafsma 2013; Wegner 2013).
Population
The 31 reviews included a total of 97,183 randomised participants, ranging from 197 in Macedo 2016 to 24,486 in Hayden 2021a. Eight reviews only included participants with acute and subacute LBP (Almeida 2023; Franke 2015; French 2006; Hayden 2005; Macedo 2016; Marin 2017; Poquet 2016; Rubinstein 2012); ten reviews only included participants with chronic LBP (Ebadi 2020; Hayden 2021a; Henschke 2010; Kamper 2014; Khadilkar 2008; Mu 2020; Parreira 2017; Rubinstein 2011; Saragiotto 2016; Wieland 2022); and 14 reviews did not restrict the duration of LBP for included participants (Dahm 2010; Engers 2008; Furlan 2005; Furlan 2015; Hayden 2005; Oltean 2014; Parry 2019; Schaafsma 2013; Urrutia 2004; Van Duijvenbode 2008; Walker 2010; Wegner 2013; Yamato 2015; Yousefi‐Nooraie 2008). Most reviews did not summarise the characteristics of the trial participants at baseline (e.g. pain intensity, sex, age, race/ethnicity, comorbidities). Age was the most common demographic characteristic summarised in the reviews (10 out of 31) (Franke 2015; Hayden 2021a; Kamper 2014; Khadilkar 2008; Macedo 2016; Marin 2017; Mu 2020; Poquet 2016; Saragiotto 2016; Yamato 2015). Race/ethnicity and comorbidities were not reported in the 'Characteristics of included studies' table, probably due to incomplete reporting in the trials.
Interventions
Twenty classes of non‐pharmacological intervention were reported across the 31 included reviews: acupuncture and dry needling (n = 2) (Furlan 2005; Mu 2020); advice and education (n = 4) (Dahm 2010; Engers 2008; Parreira 2017; Poquet 2016); assistive materials (n = 1) (Van Duijvenbode 2008); electrophysical modalities (n = 3) (Ebadi 2020; Khadilkar 2008; Yousefi‐Nooraie 2008); exercise therapies (n = 8) (Almeida 2023; Hayden 2005; Hayden 2021a; Macedo 2016; Saragiotto 2016; Schaafsma 2013; Wieland 2022; Yamato 2015); heat and cold therapies (n = 1) (French 2006); herbal therapies (n = 1) (Oltean 2014); manual therapies (n = 5) (Franke 2015; Furlan 2015; Rubinstein 2011; Rubinstein 2012; Wegner 2013); multidisciplinary therapies (n = 2) (Kamper 2014; Marin 2017); multimodal therapies (n = 2) (Parry 2019; Walker 2010); neuroreflexotherapies (n = 1) (Urrutia 2004); and psychological therapies (n = 1) (Henschke 2010). Exercise therapies were the most investigated class of non‐pharmacological intervention (359 RCTs, 37,171 individuals), followed by manual therapies (102 RCTS, 13,183 individuals) and advice and education (65 RCTs, 13,811 individuals). Treatment duration and number and duration of the sessions varied across intervention classes. The characteristics of the interventions are described in Appendix 5.
Comparisons
The most common control comparison was another active (i.e. real) treatment (28/31, 90%), followed by no treatment/waiting‐list/usual care (20/31, 65%) and placebo/sham (17/31, 55%). Three reviews described the components of usual care, consisting of medical care, counselling, and/or physiotherapy not delivered as part of the trial (Henschke 2010; Kamper 2014; Urrutia 2004). The most common placebo control comparisons were detuned electrical‐physical modalities. Active controls included a different feature of the same intervention and other modes of interventions, including advice, education, acupuncture/dry needling, manual therapies, exercises, medications, psychological interventions, combined physical therapies, and electrophysical modalities.
The comparisons for each intervention and LBP duration identified in the overview results are provided in Appendix 6.
Overlap between reviews
One review (Mu 2020), which investigated the effects of acupuncture, provided updated results for chronic LBP from an earlier review of acupuncture for acute/subacute and chronic LBP (Furlan 2005). For these reviews, we extracted data for chronic low back pain from Mu 2020, and for the acute/subacute population from Furlan 2005 (the only review that reported results for this group) to avoid overlapping primary trials in the overview. We graphically illustrated the overlap between reviews in Figure 2.
2.
We conducted the assessment of the overlap of reviews following recommended guidance (Pérez‐Bracchiglione 2022). Each field (node) of the graphical representation shows the corrected covered area (CCA) for two specific included reviews, demonstrating the hypothetical results of the CCA if the overview had only those two reviews instead of the whole matrix (Pérez‐Bracchiglione 2022). From 465 pairs of reviews, 450 pairs had slight overlap (< 5%), 12 moderate overlap (5% to < 12%), two high (10% to < 15%), and one very high overlap (≥ 15%). Reviews including psychological therapies (Henschke 2010) and multidisciplinary therapies (Kamper 2014) have a very high overlap (CCA = 16.2%), and the reviews including chiropractic interventions (Walker 2010) and spinal manipulation for chronic LBP (Rubinstein 2011), and physical conditioning as part of a return to work (Schaafsma 2013) and multidisciplinary therapies (Kamper 2014) have a high overlap (CCA = 14.7% and 14.9%, respectively). Very high overlap of studies between these systematic reviews indicates that: (1) the findings related to studies that included psychological therapies might consist of those of the studies that included multidisciplinary therapies and vice versa; (2) the findings related to studies that included chiropractic interventions might consist of those that had spinal manipulation interventions and vice versa; (3) the findings related to studies that included physical conditioning might include those with multidisciplinary therapies and vice versa.
Quality of evidence in the primary studies
All 31 reviews used a tool to assess the risk of bias in the included RCTs. One review, Urrutia 2004, used the ‘1997 11 criteria for methodological quality’ previously recommended by the Cochrane Back Review Group (Van Tulder 1997); seven reviews used the Van Tulder 2003 11 criteria for internal validity previously recommended by the Cochrane Back Review Group (Engers 2008; French 2006; Furlan 2005; Hayden 2005; Khadilkar 2008; Van Duijvenbode 2008; Yousefi‐Nooraie 2008); 11 reviews used the Higgins 2011 12 criteria for risk of bias previously recommended by the Cochrane Back Review Group (Franke 2015; Hayden 2021a; Henschke 2010; Macedo 2016; Parry 2019; Poquet 2016; Rubinstein 2012; Saragiotto 2016; Wegner 2013; Wieland 2022; Yamato 2015); 13 reviews used the Furlan 2009 12 criteria for risk of bias previously recommended by the Cochrane Back Review Group (Dahm 2010; Furlan 2015; Henschke 2010; Kamper 2014; Macedo 2016; Oltean 2014; Rubinstein 2011; Saragiotto 2016; Schaafsma 2013; Walker 2010; Wegner 2013; Wieland 2022; Yamato 2015); and six reviews used the Furlan 2015a 12 criteria for risk of bias previously recommended by the Cochrane Back Review Group (Almeida 2023; Ebadi 2020; Hayden 2021a; Marin 2017; Mu 2020; Parreira 2017), as the primary tool or as a complementary tool.
All reviews included at least one RCT assessed as being at unclear or high risk of bias across the investigated domains. The most common reasons for the high risk of bias were lack of blinding of treatment providers (557/710, 78%), lack of blinding of outcome assessors (450/735, 61%), or lack of blinding of participants (428/718, 60%). The most common reasons for unclear risk of bias were co‐interventions (313/719, 43%), compliance acceptability (274/711, 39%), and allocation concealment (289/735, 39%). Most RCTs were rated as having a low risk of bias for incomplete outcome data due to dropouts (515/736, 70%), the similarity of baseline characteristics (301/457, 66%), randomisation (480/735, 65%), and selective outcome reporting due to intention‐to‐treat analysis (382/703, 55%). We describe the details of the risk of bias tools and domains for each review in Table 3.
3. Risk of bias in the included reviews.
| Review | Number of studies assessed | Used GRADE? | Methodological quality assessment tool | Risk of bias assessment (from review authors) |
| Urrutia 2004 | 3 | No | 1997 11 criteria for methodological quality recommended by the Cochrane Back Review Group (Van Tulder 1997) | Random sequence generation: low risk (3/3) Allocation concealment: low risk (2/3) Blinding (participants): low risk (2/3) Blinding (providers): low risk (0/3) Blinding (outcome assessors): (2/3) Incomplete outcome data (dropouts): low risk (2/3) Incomplete outcome data (ITT): not reported Similarity of baseline characteristics: low risk (2/3) Co‐interventions avoided or similar: not reported Compliance acceptable: low risk (2/3) Timing of outcome assessment similar: not reported Free from selective reporting: not reported |
| Hayden 2005 | Not reported | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: not reported Allocation concealment: not reported Blinding (participants): not reported Blinding (providers): not reported Blinding (outcome assessors): not reported Incomplete outcome data (dropouts): not reported Incomplete outcome data (ITT): not reported Similarity of baseline characteristics: not reported Co‐interventions avoided or similar: not reported Compliance acceptable: not reported Timing of outcome assessment similar: not reported Free from selective reporting: not reported |
| Furlan 2005 | 12 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (5/12) Allocation concealment: low risk (4/12) Blinding (participants): low risk (5/12) Blinding (providers): low risk (0/12) Blinding (outcome assessors): low risk (5/12) Incomplete outcome data (dropouts): low risk (7/12) Incomplete outcome data (ITT): low risk (4/12) Similarity of baseline characteristics: low risk (2/12) Co‐interventions avoided or similar: low risk (7/12) Compliance acceptable: low risk (8/12) Timing of outcome assessment similar: low risk (9/12) Free from selective reporting: not reported |
| French 2006 | 5 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (2/5) Allocation concealment: low risk (1/5) Blinding (participants): low risk (0/5) Blinding (providers): low risk (0/5) Blinding (outcome assessors): low risk (4/5) Incomplete outcome data (dropouts): low risk (5/5) Incomplete outcome data (ITT): low risk (3/5) Similarity of baseline characteristics: low risk (5/5) Co‐interventions avoided or similar: low risk (4/5) Compliance acceptable: low risk (2/5) Timing of outcome assessment similar: low risk (5/5) Free from selective reporting: not reported |
| Khadilkar 2008 | 4 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (4/4) Allocation concealment: low risk (2/4) Blinding (participants): low risk (4/4) Blinding (providers): low risk (0/4) Blinding (outcome assessors): low risk (2/4) Incomplete outcome data (dropouts): low risk (2/4) Incomplete outcome data (ITT): low risk (0/4) Similarity of baseline characteristics: low risk (3/4) Co‐interventions avoided or similar: low risk (4/4) Compliance acceptable: low risk (4/4) Timing of outcome assessment similar: low risk (4/4) Free from selective reporting: not reported |
| Engers 2008 | 24 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (16/24) Allocation concealment: low risk (12/24) Blinding (participants): low risk (3/24) Blinding (providers): low risk (3/24) Blinding (outcome assessors): low risk (13/24) Incomplete outcome data (dropouts): low risk (20/24) Incomplete outcome data (ITT): low risk (20/24) Similarity of baseline characteristics: low risk (12/24) Co‐interventions avoided or similar: low risk (16/24) Compliance acceptable: not reported Timing of outcome assessment similar: not reported Free from selective reporting: not reported |
| Van Duijvenbode 2008 | 8 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (3/8) Allocation concealment: low risk (4/8) Blinding (participants): low risk (1/8) Blinding (providers): not reported Blinding (outcome assessors): low risk (2/8) Incomplete outcome data (dropouts): low risk (5/8) Incomplete outcome data (ITT): low risk (5/8) Similarity of baseline characteristics: low risk (2/8) Co‐interventions avoided or similar: low risk (2/8) Compliance acceptable: low risk (1/8) Timing of outcome assessment similar: low risk (8/8) Free from selective reporting: not reported |
| Yousefi‐Nooraie 2008 | 7 | No | 2003 11 criteria for internal validity recommended by the Cochrane Back Review Group (Van Tulder 2003) | Random sequence generation: low risk (4/7) Allocation concealment: low risk (3/7) Blinding (participants): low risk (6/7) Blinding (providers): low risk 6/7) Blinding (outcome assessors): low risk (6/7) Incomplete outcome data (dropouts): low risk (6/7) Incomplete outcome data (ITT): low risk (3/7) Similarity of baseline characteristics: low risk (4/7) Co‐interventions avoided or similar: low risk (6/7) Compliance acceptable: low risk (7/7) Timing of outcome assessment similar: low risk (7/7) Free from selective reporting: not reported |
| Dahm 2010 | 8 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (5/8) Allocation concealment: low risk (6/8) Blinding (participants): low risk (0/8) Blinding (providers): low risk (0/8) Blinding (outcome assessors): low risk (0/8) Incomplete outcome data (dropouts): low risk (5/8) Incomplete outcome data (ITT): low risk (6/8) Similarity of baseline characteristics: low risk (5/8) Co‐interventions avoided or similar: low risk (6/8) Compliance acceptable: low risk (4/8) Timing of outcome assessment similar: low risk (7/8) Free from selective reporting: low risk (0/8) |
| Henschke 2010 | 30 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (13/30) Allocation concealment: low risk (8/30) Blinding (participants): low risk (8/30) Blinding (providers): low risk (1/30) Blinding (outcome assessors): low risk (8/30) Incomplete outcome data (dropouts): low risk (15/30) Incomplete outcome data (ITT): low risk (11/30) Similarity of baseline characteristics: low risk (24/30) Co‐interventions avoided or similar: low risk (9/30) Compliance acceptable: low risk (11/30) Timing of outcome assessment similar: low risk (30/30) Free from selective reporting: low risk (28/30) |
| Walker 2010 | 12 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (4/12) Allocation concealment: low risk (6/12) Blinding (participants): low risk (1/12) Blinding (providers): low risk (0/30) Blinding (outcome assessors): low risk (7/30) Incomplete outcome data (dropouts): low risk (6/12) Incomplete outcome data (ITT): low risk (7/12) Similarity of baseline characteristics: low risk (9/12) Co‐interventions avoided or similar: low risk (4/12) Compliance acceptable: low risk (5/12) Timing of outcome assessment similar: low risk (9/12) Free from selective reporting: (0/30) |
| Rubinstein 2011 | 26 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (19/26) Allocation concealment: low risk (11/26) Blinding (participants): low risk (1/26) Blinding (providers): low risk (0/26) Blinding (outcome assessors): low risk (1/26) Incomplete outcome data (dropouts): low risk (13/26) Incomplete outcome data (ITT): low risk (10/26) Similarity of baseline characteristics: low risk (24/26) Co‐interventions avoided or similar: low risk (4/26) Compliance acceptable: low risk (7/26) Timing of outcome assessment similar: low risk (25/26) Free from selective reporting: low risk (25/26) |
| Rubinstein 2012 | 20 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) | Random sequence generation: low risk (11/20) Allocation concealment: low risk (7/20) Blinding (participants): low risk (0/20) Blinding (providers): low risk (0/20) Blinding (outcome assessors): low risk (0/20) Incomplete outcome data (dropouts): low risk (12/20) Incomplete outcome data (ITT): low risk (10/20) Similarity of baseline characteristics: low risk (8/20) Co‐interventions avoided or similar: low risk (3/20) Compliance acceptable: low risk (6/20) Timing of outcome assessment similar: low risk (19/20) Free from selective reporting: low risk (3/20) |
| Wegner 2013 | 31 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (12/31) Allocation concealment: low risk (6/31) Blinding (participants): low risk (12/31) Blinding (providers): low risk (1/31) Blinding (outcome assessors): low risk (16/31) Incomplete outcome data (dropouts): low risk (19/31) Incomplete outcome data (ITT): low risk (13/31) Similarity of baseline characteristics: low risk (19/31) Co‐interventions avoided or similar: low risk (22/31) Compliance acceptable: low risk (9/31) Timing of outcome assessment similar: low risk (25/31) Free from selective reporting: low risk (23/31) |
| Schaafsma 2013 | 25 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (21/25) Allocation concealment: low risk (17/25) Blinding (participants): low risk (19/25) Blinding (providers): low risk (2/25) Blinding (outcome assessors): low risk (1/25) Incomplete outcome data (dropouts): low risk (22/25) Incomplete outcome data (ITT): low risk (18/25) Similarity of baseline characteristics: low risk (23/25) Co‐interventions avoided or similar: low risk (6/25) Compliance acceptable: low risk (12/25) Timing of outcome assessment similar: low risk (24/25) Free from selective reporting: low risk (25/25) |
| Oltean 2014 | 14 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (9/14) Allocation concealment: low risk (3/14) Blinding (participants): low risk (9/14) Blinding (providers): low risk (8/14) Blinding (outcome assessors): low risk (8/14) Incomplete outcome data (dropouts): low risk (14/14) Incomplete outcome data (ITT): low risk (11/14) Similarity of baseline characteristics: low risk (12/14) Co‐interventions avoided or similar: low risk (9/14) Compliance acceptable: low risk (4/14) Timing of outcome assessment similar: low risk (11/14) Free from selective reporting: low risk (13/14) |
| Kamper 2014 | 41 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (29/41) Allocation concealment: low risk (28/41) Blinding (participants): low risk (41/41) Blinding (providers): low risk (0/41) Blinding (outcome assessors): low risk (0/41) Incomplete outcome data (dropouts): low risk (27/41) Incomplete outcome data (ITT): low risk (1/41) Similarity of baseline characteristics: low risk (31/41) Co‐interventions avoided or similar: low risk (7/41) Compliance acceptable: low risk (7/41) Timing of outcome assessment similar: low risk (40/41) Free from selective reporting: low risk (1/41) |
| Furlan 2015 | 25 | Yes | 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (16/25) Allocation concealment: low risk (10/25) Blinding (participants): low risk (3/25) Blinding (providers): low risk (0/25) Blinding (outcome assessors): low risk (3/25) Incomplete outcome data (dropouts): low risk (22/25) Incomplete outcome data (ITT): low risk (17/25) Similarity of baseline characteristics: low risk (13/25) Co‐interventions avoided or similar: low risk (6/25) Compliance acceptable: low risk (17/25) Timing of outcome assessment similar: low risk (23/25) Free from selective reporting: low risk (12/25) |
| Franke 2015 | 12 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) | Random sequence generation: low risk (7/12) Allocation concealment: low risk (3/12) Blinding (participants): low risk (0/12) Blinding (providers): low risk (0/12) Blinding (outcome assessors): low risk (0/12) Incomplete outcome data (dropouts): low risk (4/12) Incomplete outcome data (ITT): low risk (2/12) Similarity of baseline characteristics: low risk (1/12) Co‐interventions avoided or similar: low risk (2/12) Compliance acceptable: low risk (3/12) Timing of outcome assessment similar: low risk (11/12) Free from selective reporting: low risk (11/12) |
| Yamato 2015 | 10 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (6/10) Allocation concealment: low risk (6/10) Blinding (participants): low risk (2/10) Blinding (providers): low risk (0/12) Blinding (outcome assessors): low risk (8/10) Incomplete outcome data (dropouts): low risk (8/10) Incomplete outcome data (ITT): low risk (7/10) Similarity of baseline characteristics: low risk (9/10) Co‐interventions avoided or similar: low risk (1/10) Compliance acceptable: low risk (5/10) Timing of outcome assessment similar: low risk (10/10) Free from selective reporting: not reported |
| Macedo 2016 | 3 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (3/3) Allocation concealment: low risk (2/3) Blinding (participants): low risk (1/3) Blinding (providers): low risk (0/3) Blinding (outcome assessors): low risk (0/3) Incomplete outcome data (dropouts): low risk (1/3) Incomplete outcome data (ITT): low risk (1/3) Similarity of baseline characteristics: low risk (2/3) Co‐interventions avoided or similar: low risk (0/3) Compliance acceptable: low risk (1/3) Timing of outcome assessment similar: low risk (3/3) Free from selective reporting: low risk (3/3) |
| Poquet 2016 | 4 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) | Random sequence generation: low risk (3/4) Allocation concealment: low risk (0/4) Blinding (participants): low risk (0/4) Blinding (providers): low risk (0/4) Blinding (outcome assessors): low risk (0/4) Incomplete outcome data (dropouts): low risk (3/4) Incomplete outcome data (ITT): low risk (2/4) Similarity of baseline characteristics: low risk (2/4) Co‐interventions avoided or similar: low risk (0/4) Compliance acceptable: low risk (1/4) Timing of outcome assessment similar: low risk (2/4) Free from selective reporting: low risk (0/4) |
| Saragiotto 2016 | 29 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (24/29) Allocation concealment: low risk (13/29) Blinding (participants): low risk (2/29) Blinding (providers): low risk (0/29) Blinding (outcome assessors): low risk (2/29) Incomplete outcome data (dropouts): low risk (15/29) Incomplete outcome data (ITT): low risk (12/29) Similarity of baseline characteristics: low risk (27/29) Co‐interventions avoided or similar: low risk (1/29) Compliance acceptable: low risk (27/29) Timing of outcome assessment similar: low risk (28/29) Free from selective reporting: not reported |
| Parreira 2017 | 30 | Yes | 2015 12 criteria for risk of bias recommended by the Cochrane Back Review Group Cochrane Back and Neck Group (Furlan 2015) | Random sequence generation: low risk (12/30) Allocation concealment: low risk (7/30) Blinding (participants): low risk (0/30) Blinding (providers): low risk (0/30) Blinding (outcome assessors): low risk (9/30) Incomplete outcome data (dropouts): low risk (24/30) Incomplete outcome data (ITT): low risk (6/30) Similarity of baseline characteristics: low risk (5/30) Co‐interventions avoided or similar: low risk (13/30) Compliance acceptable: low risk (16/30) Timing of outcome assessment similar: low risk (23/30) Free from selective reporting: low risk (5/30) |
| Marin 2017 | 9 | Yes | 2015 12 criteria for risk of bias recommended by the Cochrane Back Review Group Cochrane Back and Neck Group (Furlan 2015) | Random sequence generation: low risk (7/9) Allocation concealment: low risk (7/9) Blinding (participants): low risk (0/9) Blinding (providers): low risk (1/9) Blinding (outcome assessors): low risk (0/9) Incomplete outcome data (dropouts): low risk (6/9) Incomplete outcome data (ITT): low risk (8/9) Similarity of baseline characteristics: not reported Co‐interventions avoided or similar: low risk (9/9) Compliance acceptable: low risk (4/9) Timing of outcome assessment similar: low risk (9/9) Free from selective reporting: low risk (8/9) |
| Parry 2019 | 10 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) | Random sequence generation: low risk (2/10) Allocation concealment: low risk (5/10) Blinding (participants): low risk (0/10) Blinding (providers): low risk (0/10) Blinding (outcome assessors): low risk (0/10) Incomplete outcome data (dropouts): low risk (2/10) Incomplete outcome data (ITT): low risk (8/10) Similarity of baseline characteristics: low risk (8/10) Co‐interventions avoided or similar: not reported Compliance acceptable: not reported Timing of outcome assessment similar: not reported Free from selective reporting: not reported |
| Ebadi 2020 | 10 | Yes | 2015 12 criteria for risk of bias recommended by the Cochrane Back Review Group Cochrane Back and Neck Group (Furlan 2015) | Random sequence generation: low risk (3/10) Allocation concealment: low risk (2/10) Blinding (participants): low risk (5/10) Blinding (providers): low risk (0/10) Blinding (outcome assessors): low risk (6/10) Incomplete outcome data (dropouts): low risk (6/10) Incomplete outcome data (ITT): low risk (3/10) Similarity of baseline characteristics: low risk (7/10) Co‐interventions avoided or similar: low risk (4/10) Compliance acceptable: low risk (0/10) Timing of outcome assessment similar: low risk (9/10) Free from selective reporting: low risk (2/10) |
| Mu 2020 | 33 | Yes | 2015 12 criteria for risk of bias recommended by the Cochrane Back and Neck Group (Furlan 2015a) | Random sequence generation: low risk (29/33) Allocation concealment: low risk (19/33) Blinding (participants): low risk (10/33) Blinding (providers): low risk (0/33) Blinding (outcome assessors): low risk (10/33) Incomplete outcome data (dropouts): low risk (18/33) Incomplete outcome data (ITT): low risk (18/33) Similarity of baseline characteristics: low risk (25/33) Co‐interventions avoided or similar: low risk (23/33) Compliance acceptable: low risk (17/33) Timing of outcome assessment similar: low risk (30/33) Free from selective reporting: low risk (30/30) |
| Hayden 2021a | 249 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2015 12 criteria for risk of bias recommended by the Cochrane Back and Neck Group (Furlan 2015a) | Random sequence generation: low risk (178/249) Allocation concealment: low risk (147/249) Blinding (participants): low risk (82/249) Blinding (providers): low risk (65/249) Blinding (outcome assessors): low risk (68/249) Incomplete outcome data (dropouts): low risk (200/249) Incomplete outcome data (ITT): low risk (160/249) Similarity of baseline characteristics: not reported Co‐interventions avoided or similar: low risk (128/249) Compliance acceptable: low risk (117/249) Timing of outcome assessment similar: low risk (241/249) Free from selective reporting: low risk (227/249) |
| Wieland 2022 | 21 | Yes | 2015 11 used the 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Higgins 2011) and 2009 12 criteria for risk of bias recommended by the Cochrane Back Review Group (Furlan 2009) | Random sequence generation: low risk (17/21) Allocation concealment: low risk (17/21) Blinding (participants): low risk (0/21) Blinding (providers): low risk (0/21) Blinding (outcome assessors): low risk (0/21) Incomplete outcome data (dropouts): low risk (8/21) Incomplete outcome data (ITT): low risk (13/21) Similarity of baseline characteristics: low risk (12/21) Co‐interventions avoided or similar: low risk (10/21) Compliance acceptable: low risk (0/21) Timing of outcome assessment similar: low risk (21/21) Free from selective reporting: low risk (6/21) |
| Almeida 2023 | 5 | Yes | 2015 12 criteria for risk of bias recommended by the Cochrane Back and Neck Group (Furlan 2015a) | Random sequence generation: low risk (4/5) Allocation concealment: low risk (2/5) Blinding (participants): low risk (0/5) Blinding (providers): low risk (0/5) Blinding (outcome assessors): low risk (0/5) Incomplete outcome data (dropouts): low risk (4/5) Incomplete outcome data (ITT): low risk (3/5) Similarity of baseline characteristics: low risk (5/5) Co‐interventions avoided or similar: low risk (3/5) Compliance acceptable: low risk (4/5) Timing of outcome assessment similar: low risk (5/5) Free from selective reporting: not reported |
ITT: intention‐to‐treat analysis
Twenty‐three of 31 reviews (74%) used the GRADE approach to assess the certainty of evidence. We conducted additional GRADE assessments for 220 (26%) comparisons of non‐pharmacological interventions that reported the results using quantitative analysis. The eight reviews that did not use the GRADE approach provided a rating of certainty of evidence using methodological criteria described by Van Tulder 1997 or Van Tulder 2003 (Engers 2008; French 2006; Furlan 2005; Hayden 2005; Khadilkar 2008; Urrutia 2004; Van Duijvenbode 2008; Yousefi‐Nooraie 2008).
Methodological quality of included reviews
Using the AMSTAR 2 tool, we reached a judgement of high confidence in 19 of 31 reviews (61%) (Hayden 2005, Almeida 2023; Engers 2008; Franke 2015; Furlan 2015; Hayden 2021a; Khadilkar 2008; Marin 2017; Mu 2020; Parreira 2017; Poquet 2016; Rubinstein 2011; Rubinstein 2012; Saragiotto 2016; Schaafsma 2013; Wegner 2013; Wieland 2022; Yamato 2015; Yousefi‐Nooraie 2008). We had moderate confidence in two of 31 reviews (6%): in both Van Duijvenbode 2008 and Kamper 2014, authors did not perform data extraction in duplicate. Additionally, in Kamper 2014, authors did not report the funding sources for the primary studies, and in Van Duijvenbode 2008, authors did not provide a satisfactory explanation for heterogeneity. We reached a judgement of low confidence in the remaining 10 reviews (32%) (Dahm 2010; Ebadi 2020; French 2006; Furlan 2005; Henschke 2010; Macedo 2016; Oltean 2014; Parry 2019; Urrutia 2004; Walker 2010), primarily because authors did not investigate publication bias and did not justify this omission (Table 4, Table 5, Table 6 and Table 7). We summarised our judgements for each domain of the AMSTAR 2 confidence assessment in Figure 3.
4. AMSTAR 2 quality assessment.
| AMSTAR 2 item | Urrutia 2004 | Hayden 2005 | Furlan 2005 | French 2006 | Khadilkar 2008 | Engers 2008 | Van Duijvenbode 2008 | Yousefi‐Nooraie 2008 | Dahm 2010 |
| 1. Did the research questions and inclusion criteria for the review include the components of PICO? | Yes | Yes | Yes | No | Yes | Yes | Yes | Yes | Yes |
| 2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review, and did the report justify any significant deviations from the protocol? * | Yes | Partial yes | Yes | Yes | Yes | Yes | Yes | Yes | Partial yes |
| 3. Did the review authors explain their selection of the study designs for inclusion in the review? | Yes | Yes | Yes | No | Yes | Yes | Yes | Yes | Yes |
| 4. Did the review authors use a comprehensive literature search strategy? * | Partial yes | Partial yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 5. Did the review authors perform study selection in duplicate? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 6. Did the review authors perform data extraction in duplicate? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 7. Did the review authors provide a list of excluded studies and justify the exclusions? * | No | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 8. Did the review authors describe the included studies in adequate detail? | Yes | Partial yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 9. Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? * | Yes | Partial yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 10. Did the review authors report on the sources of funding for the studies included in the review? | No | No | Yes | Yes | No | No | Yes | No | No |
| 11. If meta‐analysis was performed, did the review authors use appropriate methods for the statistical combination of results? * | NA | Yes | Yes | Yes | NA | NA | NA | Yes | Yes |
| 12. If meta‐analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta‐analysis or other evidence synthesis? | NA | Yes | Yes | Yes | NA | NA | NA | No | Yes |
| 13. Did the review authors account for RoB in primary studies when interpreting/discussing the results of the review? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 14. Did the review authors provide a satisfactory explanation for and discussion of any heterogeneity observed in the results of the review? | No | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes |
| 15. If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? * | NA | Yes | No | No | NA | NA | NA | Yes | No |
| 16. Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Rating of overall confidence | Low | High | Low | Low | High | High | Moderate | High | Low |
NA: not applicable; PICO: population, intervention(s), comparator(s), outcome(s)
5. AMSTAR 2 quality assessment [continued].
| AMSTAR 2 item | Henschke 2010 | Walker 2010 | Rubinstein 2011 | Rubinstein 2012 | Wegner 2013 | Schaafsma 2013 | Oltean 2014 | Kamper 2014 | Furlan 2015 |
| 1. Did the research questions and inclusion criteria for the review include the components of PICO? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review, and did the report justify any significant deviations from the protocol? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 3. Did the review authors explain their selection of the study designs for inclusion in the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 4. Did the review authors use a comprehensive literature search strategy? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 5. Did the review authors perform study selection in duplicate? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 6. Did the review authors perform data extraction in duplicate? | Yes | Yes | Yes | No | Yes | Yes | Yes | No | Yes |
| 7. Did the review authors provide a list of excluded studies and justify the exclusions? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 8. Did the review authors describe the included studies in adequate detail? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 9. Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 10. Did the review authors report on the sources of funding for the studies included in the review? | No | Yes | Yes | Yes | No | No | Yes | No | Yes |
| 11. If meta‐analysis was performed, did the review authors use appropriate methods for the statistical combination of results? * | Yes | Yes | Yes | Yes | Yes | Yes | NA | Yes | Yes |
| 12. If meta‐analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta‐analysis or other evidence synthesis? | Yes | Yes | Yes | Yes | Yes | Yes | NA | Yes | Yes |
| 13. Did the review authors account for RoB in primary studies when interpreting/discussing the results of the review? * | Yes | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes |
| 14. Did the review authors provide a satisfactory explanation for and discussion of any heterogeneity observed in the results of the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 15. If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? * | No | No | Yes | Yes | Yes | Yes | NA | Yes | Yes |
| 16. Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Rating of overall confidence | Low | Low | High | High | High | High | Low | Moderate | High |
NA: not applicable; PICO: population, intervention(s), comparator(s), outcome(s)
6. AMSTAR 2 quality assessment [continued].
| AMSTAR 2 item | Franke 2015 | Yamato 2015 | Macedo 2016 | Poquet 2016 | Saragiotto 2016 | Parreira 2017 | Marin 2017 | Parry 2019 | Ebadi 2020 |
| 1. Did the research questions and inclusion criteria for the review include the components of PICO? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review, and did the report justify any significant deviations from the protocol? * | Yes | Yes | No | Yes | Yes | Yes | Yes | Yes | Yes |
| 3. Did the review authors explain their selection of the study designs for inclusion in the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Partial yes | Yes |
| 4. Did the review authors use a comprehensive literature search strategy? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 5. Did the review authors perform study selection in duplicate? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 6. Did the review authors perform data extraction in duplicate? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 7. Did the review authors provide a list of excluded studies and justify the exclusions? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 8. Did the review authors describe the included studies in adequate detail? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 9. Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 10. Did the review authors report on the sources of funding for the studies included in the review? | No | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 11. If meta‐analysis was performed, did the review authors use appropriate methods for the statistical combination of results? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 12. If meta‐analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta‐analysis or other evidence synthesis? | Yes | Yes | Yes | Yes | Yes | Yes | No | No | Yes |
| 13. Did the review authors account for RoB in primary studies when interpreting/discussing the results of the review? * | Yes | Yes | Yes | Yes | Yes | Yes | Partial yes | No | Yes |
| 14. Did the review authors provide a satisfactory explanation for and discussion of any heterogeneity observed in the results of the review? | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes | No |
| 15. If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? * | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No |
| 16. Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review? | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Rating of overall confidence | High | High | Low | High | High | High | High | Low | Low |
NA: not applicable; PICO: population, intervention(s), comparator(s), outcome(s)
7. AMSTAR 2 quality assessment [continued].
| AMSTAR 2 item | Mu 2020 | Hayden 2021a | Wieland 2022 | Almeida 2023 |
| 1. Did the research questions and inclusion criteria for the review include the components of PICO? | Yes | Yes | Yes | Yes |
| 2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review, and did the report justify any significant deviations from the protocol? * | Yes | Yes | Yes | Yes |
| 3. Did the review authors explain their selection of the study designs for inclusion in the review? | Yes | Yes | Yes | Yes |
| 4. Did the review authors use a comprehensive literature search strategy? * | Yes | Yes | Yes | Yes |
| 5. Did the review authors perform study selection in duplicate? | Yes | Yes | Yes | Yes |
| 6. Did the review authors perform data extraction in duplicate? | Yes | Yes | Yes | Yes |
| 7. Did the review authors provide a list of excluded studies and justify the exclusions? * | Yes | Yes | Yes | Yes |
| 8. Did the review authors describe the included studies in adequate detail? | Yes | Yes | Yes | Yes |
| 9. Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? * | Yes | Yes | Yes | Yes |
| 10. Did the review authors report on the sources of funding for the studies included in the review? | Yes | Yes | Yes | Yes |
| 11. If meta‐analysis was performed, did the review authors use appropriate methods for the statistical combination of results? * | Yes | Yes | Yes | Yes |
| 12. If meta‐analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta‐analysis or other evidence synthesis? | Yes | Yes | Yes | Yes |
| 13. Did the review authors account for RoB in primary studies when interpreting/discussing the results of the review? * | Yes | Yes | Yes | Yes |
| 14. Did the review authors provide a satisfactory explanation for and discussion of any heterogeneity observed in the results of the review? | Yes | Yes | Yes | Yes |
| 15. If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? * | Yes | Yes | Yes | Yes |
| 16. Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review? | Yes | Yes | Yes | Yes |
| Rating of overall confidence | High | High | High | High |
NA: not applicable; PICO: population, intervention(s), comparator(s), outcome(s)
3.

Summary of the judgements for each domain of the AMSTAR 2 confidence assessment. Domain 1: appropriate research question and inclusion criteria. Domain 2*: protocol registered before the commencement of the review. Domain 3: selection of the study designs rationale. Domain 4*: adequacy of the literature search. Domain 5: performing study selection in duplicate. Domain 6: performing data extraction in duplicate. Domain 7*: justification for excluding individual studies. Domain 8: reporting a list of excluded studies and justification for the exclusions. Domain 9*: risk of bias from individual studies being included in the review. Domain 10: reporting the sources of funding for the studies included in the review. Domain 11*: appropriateness of meta‐analytical methods. Domain 12: assessment of the potential impact of the risk of bias in individual studies. Domain 13*: consideration of the risk of bias when interpreting the results of the review. Domain 14: a satisfactory explanation for and discussion of any heterogeneity observed in the review results. Domain 15*: assessment of the presence and likely impact of publication bias. Domain 16: appropriate reporting of potential sources of conflict of interest and funding authors received for conducting the review. * indicates ‘critical’ domains. The details for each study are described in Table 4.
Effect of interventions
The effects of non‐pharmacological interventions compared to placebo/sham or no treatment/usual care on the primary outcomes (pain intensity, disability and safety) for acute/subacute and chronic LBP are presented below.
The short‐term effects of non‐pharmacological interventions on pain intensity compared to placebo/sham or no treatment/usual care are summarised in summary of findings tables (Table 8 and Table 9, respectively). These tables summarise the effects of the non‐pharmacological interventions on pain intensity and the certainty of the evidence (GRADE). The short‐term effects of non‐pharmacological intervention classes on other outcomes are provided in ‘Overview of reviews’ tables (Appendix 7; Appendix 8; Appendix 9; Appendix 10; Appendix 11; Appendix 12; Appendix 13; Appendix 14; Appendix 15).
8. Summary of findings table for pain intensity for non‐pharmacological and non‐surgical interventions compared to sham or placebo at short‐term follow‐up (≤ 3 months).
| Non‐pharmacological interventions for low back pain in adults compared to sham or placebo at short‐term follow‐up (≤ 3 months post‐intervention) | |||||||
| Outcome | Low back pain duration | Intervention and sham or placebo | Relative effect, 95% CI | I2 (%) | Number of participants (RCTs) | Certainty of the evidence (GRADE) | Comments |
| Pain 6‐point verbal pain intensity scale, where 0 is no pain |
Acute/subacute | Heated back wrap versus placebo | MD 1.06, 95% CI 0.68 to 1.45 | ‐ | 258 (2 RCTs) | Lowb | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Heated back blanket versus placebo | MD ‐32.2, 95% CI ‐38.6 to ‐25.7 | ‐ | 90 (1 RCT) | Lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Muscle energy therapy versus sham manual treatment | MD 14.2 favouring sham muscle energy therapy, 95% CI ‐10.1 to 38.5 | ‐ | 20 (1 RCT) | Lowa,b | |
| Pain | Acute/subacute | Spinal manipulation versus placebo (at 1‐week follow‐up) | SMD ‐0.08, 95% CI ‐0.37 to 0.21 | 0 | 205 (2 RCTs) | Moderateb | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Spinal manipulation versus placebo | MD ‐12, 95% CI ‐20 to ‐3.9 | ‐ | 178 (1 RCT) | Lowb,c | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Spinal manipulation versus placebo (detuned electrical physical modality) | MD ‐12, 95% CI ‐20 to ‐3.9 | ‐ | 178 (1 RCT) | Lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Spinal manipulation versus sham adjustments | MD ‐5, 95% CI ‐13.9 to 3.9 | ‐ | 74 (1 RCT) | Very lowa‐c | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Acupuncture versus sham acupuncture | MD ‐10.04, 95% CI ‐17.22 to ‐2.85 | 76 | 1095 (5 RCTs) | Lowa,c | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Exercise versus placebo | MD ‐8.9, 95% CI ‐15.1 to ‐2.6 | 0 | 300 (2 RCTs) | Lowb,d | Review’s post hoc analysis |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Low‐level laser therapy versus sham laser therapy | MD ‐16, 95% ‐27.95 to ‐4.05 | ‐ | 56 (1 RCT) | Lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Spinal manipulation versus placebo (detuned electrical physical modality) | MD ‐6, 95% CI ‐15.82 to 3.82 | ‐ | 72 (1 RCT) | Very lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Spinal manipulation versus sham manipulation | MD ‐3.24, 95% CI ‐13.62 to 7.15 | 53 | 148 (3 RCTs) | Very lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Traction versus sham | MD ‐4, 95% CI ‐17.65 to 9.65 | ‐ | 60 (1 RCT) | Moderateb | |
GRADE: downgraded (by one or two levels, depending on the seriousness of the issue) for study limitations.
aDowngraded for study limitations bDowngraded for imprecision cDowngraded for inconsistency dDowngraded for other factors
We have ordered the interventions in the table by pain duration (acute/subacute, followed by chronic), and then alphabetically by intervention label.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; SMD: standardised mean difference
9. Summary of findings table for pain intensity for non‐pharmacological and non‐surgical interventions compared to no treatment or usual care at short‐term follow‐up (≤ 3 months).
| Non‐pharmacological and non‐surgical interventions for low back pain in adults compared to no treatment or usual care at short‐term follow‐up (≤ 3 months post‐intervention) | |||||||
| Outcome | Low back pain duration | Intervention and no treatment or usual care | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Pain 0 to 100 scale, where 0 is no pain |
Acute | Exercise versus no treatment/usual care | MD 0.59 favouring no treatment, 95% CI ‐11.51 to 12.69 | 86 | 491 (3 RCTs) | Very lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Subacute | Exercise (McKenzie) added to exercise versus no treatment | MD ‐8, 95% CI ‐10.93 to ‐5.07 | ‐ | 194 (1 RCT) | Lowa,b | |
| Pain | Acute/subacute | Massage versus usual care | SMD ‐1.24 95% CI ‐1.85 to ‐0.64 | ‐ | 51 (1 RCT) | Very lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | McKenzie exercise versus no treatment/ usual care | MD ‐ 7.3, 95% CI ‐12 to ‐2.6 | 337 (2 RCTs) | Lowa,b | ||
| Pain 0 to 100 scale, where 0 is no pain |
Acute | Motor control exercise versus no treatment/usual care | MD ‐9.3 favouring motor control, 95% CI ‐20.41 to 1.8 | ‐ | 42 (1 RCT) | Very lowa,b | |
| Pain | Acute/subacute | Multidisciplinary versus usual care | SMD ‐0.4, 95% CI ‐0.74 to ‐0.06 | ‐ | 272 (4 RCTs) | Very lowa‐c | |
| Pain 0 to 100 scale, where 0 is no pain |
Acute/subacute | Spinal manipulation added to usual care versus usual care alone | MD 8.4 favouring usual care alone, 95% CI ‐0.4 to 17.2 | ‐ | 104 (1 RCT) | Lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Acupuncture versus usual care | MD ‐10.11, 95% CI ‐16.80 to ‐3.43 | 0 | 144 (3 RCTs) | Moderatec | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Acupuncture versus no treatment/usual care | MD ‐18.78, 95% CI ‐37.97 to 0.41 | 95 | 817 (2 RCTs) | Lowa‐c | |
| Pain | Chronic | Acupuncture added to exercise versus no treatment/usual care | SMD ‐0.49, 95% CI ‐0.78 to ‐0.2 | 0 | 187 (2 RCTs) | Lowa | Very low‐certainty evidence according to the overview GRADE criteria: downgraded for study limitations and imprecision |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Acupuncture versus usual care | MD ‐18.78 favouring acupuncture, 95% CI ‐37.97 to 0.41 | 95 | 817 (2 RCTs) | Very lowa‐c | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Back school versus no treatment | MD ‐6.1, 95% CI ‐10.8 to ‐2.01 | 18 | 647 (6 RCTs) | Very lowa,c,d | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Back school versus usual care | MD ‐10.16, 95% CI ‐19.11 to ‐1.22 | 62 | 249 (3 RCTs) | Very lowa,b,d | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Electrophysical versus placebo | MD ‐7.12, 95% CI ‐17.99 to 3.7 | 77 | 121 (3 RCTs) | Very lowa‐c | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Exercise versus no treatment/usual care | MD ‐15.2, 95% CI ‐18.3 to ‐12.2 | 75 | 2746 (35 RCTs) | Moderatec | The analysis included two studies with a placebo comparison |
| Pain | Mixed | Massage versus no treatment (inactive interventions) | SMD ‐0.72, 95% CI ‐1.05 to ‐0.39 | 0 | 761 (7 RCTs) | Lowa | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Motor control exercise versus no treatment/usual care | MD ‐10.01, 95% CI ‐15.67 to ‐4.35 | 36 | 291 (4 RCTs) | Moderateb | |
| Pain | Chronic | Multidisciplinary versus no treatment/usual care | SMD ‐0.55, 95% CI ‐0.83 to ‐0.28 | 72 | 879 (9 RCTs) | Moderatec | |
| Pain | Chronic | Multidisciplinary versus no treatment | SMD ‐0.73, 95% CI ‐1.22 to ‐0.24 | 64 | 213 (3 RCTs) | Lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Pilates versus no treatment/usual care | MD ‐14.05, 95% CI ‐18.91 to ‐9.19 | 6 | 265 (6 RCTs) | Lowa,b | |
| Pain | Chronic | Psychological therapies (operant approaches) versus no treatment/usual care | SMD ‐0.43, 95% CI ‐0.75 to ‐0.11 | 0 | 153 (3 RCTs) | Moderateb | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Psychological therapies (progressive muscle relaxation and operant approaches) versus no treatment/usual care | MD ‐5.18, 95% CI ‐9.79 to ‐0.57 | 19 | 330 (2 RCTs) | Moderateb | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Psychological respondent approach (progressive relaxation) versus no treatment | MD ‐19.77, 95% CI ‐34.34 to ‐5.2 | 57 | 74 (3 RCTs) | Lowa,b | |
| Pain | Chronic | Psychological respondent approach (EMG feedback) versus no treatment | SMD ‐0.8, 95% CI ‐1.32 to ‐0.28 | 0 | 64 (3 RCTs) | Lowa,b | |
| Pain | Chronic | Psychological cognitive approaches versus no treatment | SMD ‐0.27, 95% CI ‐0.75 to 0.22 | 10 | 68 (2 RCTs) | Lowa,b | |
| Pain | Chronic | Psychological cognitive behavioural therapy versus no treatment | SMD ‐0.6, 95% CI ‐0.97 to ‐0.22 | 42 | 239 (5 RCTs) | Lowa,b | |
| Pain | Chronic | Workplace interventions (multi‐component) versus no treatment | RR 0.93, 95% CI 0.69 to 1.27 | 0 | 107 (3 RCTs) | Lowa,b | |
| Pain | Chronic | Workplace interventions versus no treatment | SMD ‐0.35, 95% CI ‐0.8 to 0.10 | 0 | 79 (2 RCTs) | Very lowa,b | |
| Pain | Chronic | Workplace interventions (treadmill workstation) versus no treatment | RR 1.05, 95% CI 0.5 to 2.19 | ‐ | 47 (1 RCT) | Very lowa,b | |
| Pain | Chronic | Workplace interventions (activity tracker) versus no treatment | SMD 0.05, 95% CI ‐0.87 to 0.77 | ‐ | 47 (1 RCT) | Very lowa,b | |
| Pain 0 to 100 scale, where 0 is no pain |
Chronic | Yoga versus no treatment/usual care | MD ‐11.05, 95% CI ‐14.22 to ‐7.8 | 0 | 258 (5 RCTs) | Lowa,b | |
GRADE: downgraded (by one or two levels, depending on the seriousness of the issue) for study limitations.
aDowngraded for study limitations bDowngraded for imprecision cDowngraded for inconsistency dDowngraded for publication bias
We have ordered the interventions in the table by pain duration (acute/subacute, followed by chronic and mixed), and then alphabetically by intervention label.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RR: risk ratio; SMD: standardised mean difference
The effects of the interventions compared to other interventions are reported in Appendix 3. A ‘Summary of results’ table provides the effect size, certainty of the evidence, and methodological confidence in the systematic review of non‐pharmacological interventions compared to placebo/sham or no treatment/usual care on pain and function at the short‐term follow‐up (Table 10 and Table 11).
10. Summary of results table, effect estimates of interventions on pain intensity compared to sham/placebo or no treatment/usual care for acute/subacute and chronic low back pain at short‐term follow‐up (< 3 months).
| Acute/subacute | Chronic | ||||||
| Intervention class | Interventions | Sham/placebo |
No treatment/ usual care |
Confidence in the review | Sham/placebo |
No treatment/ usual care |
Confidence in the review |
| Acupuncture/dry needling | Acupuncture | ? | ? | Low (Furlan 2005) |
Small (LOW) |
Mediumc (MODERATE) |
High (Mu 2020) |
| No evidence of a differenced (VERY LOW) | |||||||
| Smalle (LOW) | |||||||
| Dry needling | ? | ? | ? | ? | ? | ||
| Advice/ education | Advice | ? | ? | Low (Dahm 2010) |
? | ? | ? |
| Back school | ? | ? | High (Poquet 2016) |
? | Smallc (VERY LOW) |
High (Parreira 2017) |
|
| Mediumd (VERY LOW) | |||||||
| Individual patient education | ? | ? | High (Engers 2008) |
? | ? | High (Engers 2008) |
|
| Assistive materials | Lumbar support | ? | ? | Moderate (Van Duijvenbode 2008) |
? | ? | Moderate (Van Duijvenbode 2008) |
| Electrophysical modalities | Low‐level laser therapy | ? | ? | Low (Yousefi‐Nooraie 2008) |
Medium (LOW) |
? | Low (Yousefi‐Nooraie 2008) |
| TENS | ? | ? | High (Khadilkar 2008) |
? | ? | High (Khadilkar 2008) |
|
| Ultrasound | ? | ? | ? | No evidence of a differenceb (VERY LOW) |
? | Low (Ebadi 2020) |
|
| Exercise therapies | Exercises | ? | No evidence of a difference (VERY LOW) *acute low back pain |
High (Hayden 2005) |
Small (LOW) |
Medium (MODERATE) |
High (Hayden 2021a) |
| McKenzie | ? | Small (LOW) |
High (Almeida 2023) |
? | ? | ? | |
| Motor control | ? | No evidence of a difference (LOW) |
High (Macedo 2016) |
? | Small (MODERATE) |
High (Saragiotto 2016) |
|
| Pilates | ? | ? | High (Yamato 2015) |
? | Medium (LOW) |
High (Yamato 2015) |
|
| Yoga | ? | ? | ? | ? | Medium (LOW) |
High (Wieland 2022) |
|
| Heat and cold therapies | Heat (superficial) | Mediumb (LOW) *after five days of heat application |
? | Low (French 2006) |
? | ? | Low (French 2006) |
| Largea (LOW) *after heat application | |||||||
| Cold therapy | ? | ? | ? | ? | |||
| Herbal therapies | Herbal intervention | ? | ? | Low (Oltean 2014) |
? | ? | Low (Oltean 2014) |
| Manual therapies | Massage | ? | Large (VERY LOW) |
High (Furlan 2015) |
? | ? | High (Furlan 2015) |
| Muscle energy technique | No evidence of a difference (VERY LOW) |
? | Low (Franke 2015) |
? | ? | Low (Franke 2015) |
|
| Spinal manipulation | Mediumb (LOW) |
? | High (Rubinstein 2012) |
No clear evidence of a difference (VERY LOW) |
? | High (Rubinstein 2011) |
|
| No evidence of a differencea (LOW) | |||||||
| Traction | ? | ? | High (Wegner 2013) |
No evidence of a difference (MODERATE) |
? | High (Wegner 2013) |
|
| Chiropractic | ? | ? | Low (Walker 2010) |
? | ? | Low (Walker 2010) |
|
| Multimodal therapies | Physical conditioning with lifestyle modification | ? | ? | High (Schaafsma 2013) |
? | ? | High (Schaafsma 2013) |
| Workplace intervention to increase standing and walking | ? | ? | High (Parry 2019) |
? | No evidence of a difference (VERY LOW) |
High (Parry 2019) |
|
| Multidisciplinary therapies | Multidisciplinary intervention | ? | Small (VERY LOW) |
High (Marin 2017) |
? | Mediumc (LOW) |
High (Kamper 2014) |
| Mediumd (MODERATE) | |||||||
| Neuroreflexotherapies | Neuroreflexotherapies | ? | ? | Low (Urrutia 2004) |
? | ? | Low (Urrutia 2004) |
| Psychological therapies |
Psychological interventions | ? | ? | ? | ? | Small (MODERATE) |
Moderate (Henschke 2010) |
| Respondent approaches (progressive relaxation) | Medium (LOW) |
||||||
| Respondent approaches (EMG feedback) | Medium (LOW) |
||||||
| Operant approaches | Small (MODERATE) |
||||||
| Cognitive approaches | No evidence of a difference (LOW) |
||||||
| Cognitive‐behavioural approaches | Medium (LOW) |
||||||
The magnitude of effects is based upon the American Pain Society criteria (Chou 2017): large effect (≥ 20 points on a 0 to 100 scale or ≥ 0.8 SMD); medium effect (> 10 to 20 points on a 0 to 100 scale or > 0.5 to 0.8 SMD); small effect (5 to 10 points on a 0 to 100 scale or 0.2 to 0.5 SMD); little to no difference (< 5 points on 0 to 100 scale or < 0.2 SMD); no evidence of a difference (boundaries of the 95% confidence interval span both sides of the line of no effect); harmful (boundaries of the 95% confidence interval fall completely within harm).
The superscript letter accompanying the magnitude of the effects indicates the following specific comparisons:
aSham intervention as a comparison bPlacebo as a comparison cNo treatment as a comparison dUsual care as a comparison eAdded to usual care
No superscript in the magnitude of effect indicates that the analysis did not differentiate between sham and placebo, or no treatment and usual care.
Details in parentheses ( ) accompanying the magnitude of the effects indicate the certainty of the evidence (very low, low, moderate, high) of the effect estimate.
'?' indicates that no quantitative estimates are reported in the systematic reviews, or that there is no systematic review for the intervention in the particular pain duration.
We assessed the included reviews' credibility with AMSTAR 2 (Shea 2017).
11. Summary of results table, effect estimates of interventions on function compared to sham/placebo or no treatment/usual care for acute/subacute and chronic low back pain at short‐term follow‐up (< 3 months).
| Acute/subacute | Chronic | ||||||
| Intervention class | Interventions | Sham/placebo |
No treatment/ usual care |
Confidence in the review | Sham/placebo |
No treatment/ usual care |
Confidence in the review |
| Acupuncture/dry needling | Acupuncture | ? | ? | Low (Furlan 2005) |
Smalla (MODERATE) |
Smallc (MODERATE) |
High (Mu 2020) |
| Mediumd (LOW) | |||||||
| Mediume (LOW) | |||||||
| Dry needling | ? | ? | ? | ? | ? | ||
| Advice/ education | Advice | ? | ? | Low (Dahm 2010) |
? | ? | High (Parreira 2017) |
| Back school | ? | ? | High (Poquet 2016) |
? | Little to no differencec (VERY LOW) |
||
| No evidence of a differenced (VERY LOW) | |||||||
| Individual patient education | ? | ? | High (Engers 2008) |
? | ? | High (Engers 2008) |
|
| Assistive materials | Lumbar support | ? | ? | Moderate (Van Duijvenbode 2008) |
? | ? | Moderate (Van Duijvenbode 2008) |
| Electrophysical modalities | Low‐level laser therapy | ? | ? | Low (Yousefi‐Nooraie 2008) |
? | ? | Low (Yousefi‐Nooraie 2008) |
| TENS | ? | ? | High (Khadilkar 2008) |
? | ? | High (Khadilkar 2008) |
|
| Ultrasound | ? | ? | ? | Small (LOW) |
? | Low (Ebadi 2020) |
|
| Exercise therapies | Exercises | ? | No evidence of a difference (VERY LOW) *acute low back pain |
High (Hayden 2005) |
Small (MODERATE) |
Small (MODERATE) |
High (Hayden 2021a) |
| McKenzie | ? | No evidence of a difference (LOW) |
High (Almeida 2023) |
? | ? | ? | |
| Motor control | ? | No evidence of a difference (VERY LOW) *acute low back pain |
High (Macedo 2016) |
? | Small (VERY LOW) |
High (Saragiotto 2016) |
|
| Pilates | ? | ? | High (Yamato 2015) |
? | Small (LOW) |
High (Yamato 2015) |
|
| Yoga | ? | ? | ? | ? | Small (MODERATE) |
High (Wieland 2022) | |
| Heat and cold therapies | Heat (superficial) | Mediumb | ? | Low (French 2006) |
? | ? | Low (French 2006) |
| Herbal therapies | Herbal intervention | ? | ? | Low (Oltean 2014) |
? | ? | Low (Oltean 2014) |
| Manual therapies | Massage | ? | No evidence of a difference (VERY LOW) |
High (Furlan 2015) |
? | ? | High (Furlan 2015) |
| Muscle energy technique | ? | ? | Low (Franke 2015) |
? | ? | Low (Franke 2015) |
|
| Spinal manipulation | No evidence of a differencea (VERY LOW) |
No evidence of a difference (LOW) |
High (Rubinstein 2012) |
? | ? | High (Rubinstein 2011) |
|
| No evidence of a differenceb (LOW) | |||||||
| Traction | ? | ? | High (Wegner 2013) |
? | ? | High (Wegner 2013) |
|
| Multimodal therapies | Chiropractic | ? | ? | Low (Walker 2010) |
? | ? | Low (Walker 2010) |
| Physical conditioning with lifestyle modification | ? | ? | High (Schaafsma 2013) |
? | ? | High (Schaafsma 2013) |
|
| Workplace intervention to increase standing and walking | ? | ? | High (Parry 2019) |
? | Small (LOW) *type of intervention is multi‐component workplace |
High (Parry 2019) |
|
| No evidence of a difference (LOW) *type of intervention is sit‐stand workstation | |||||||
| Multidisciplinary therapies | Multidisciplinary intervention | ? | Small (VERY LOW) |
High (Marin 2017) |
? | Smallc (VERY LOW) |
Moderate (Kamper 2014) |
| Smalld (LOW) | |||||||
| Neuroreflexotherapies | Neuroreflexotherapies | ? | ? | Low (Urrutia 2004) |
? | ? | Low (Urrutia 2004) |
| Psychological therapies | Psychological interventions | ? | ? | ? | ? | No evidence of a difference (MODERATE) |
Moderate (Henschke 2010) |
| Respondent approaches (progressive relaxation) | Large (VERY LOW) |
||||||
| Respondent approaches (EMG feedback) | No evidence of a difference (VERY LOW) |
||||||
| Operant approaches | No evidence of a difference (LOW) |
||||||
| Cognitive approaches | No evidence of a difference (LOW) |
||||||
| Cognitive‐behavioural approaches | No evidence of a difference (LOW) |
||||||
The magnitude of effects is based upon the American Pain Society criteria (Chou 2017): large effect (≥ 20 points on a 0 to 100 scale or ≥ 0.8 SMD); medium effect (> 10 to 20 points on a 0 to 100 scale or > 0.5 to 0.8 SMD); small effect (5 to 10 points on a 0 to 100 scale or 0.2 to 0.5 SMD); little to no difference (< 5 points on 0 to 100 scale or < 0.2 SMD); no evidence of a difference (boundaries of the 95% confidence interval span both sides of the line of no effect); harmful (boundaries of the 95% confidence interval fall completely within harm).
The superscript letter accompanying the magnitude of the effects indicates the following specific comparisons:
aSham intervention as a comparison bPlacebo as a comparison cNo treatment as a comparison dUsual care as a comparison eAdded to usual care
No superscript in the magnitude of effect indicates that the analysis did not differentiate between sham and placebo, or no treatment and usual care.
Details in parentheses ( ) accompanying the magnitude of the effects indicate the certainty of the evidence (very low, low, moderate, high) of the effect estimate.
'?' indicates that no quantitative estimates are reported in the systematic reviews, or that there is no systematic review for the intervention in the particular pain duration.
We assessed the included reviews' credibility with AMSTAR 2 (Shea 2017).
1. Non‐pharmacological and non‐surgical treatments versus placebo/sham
Acupuncture and dry needling versus placebo/sham
Two Cochrane reviews investigated the effects of acupuncture and dry needling on LBP (Furlan 2005; Mu 2020), with both reviews comparing acupuncture to placebo/sham. One review, judged at low confidence, included 12 trials with 1177 participants with acute or subacute LBP (Furlan 2005). Another review, judged at high confidence, included 33 trials with 8270 participants with chronic LBP (Mu 2020).
Acute/subacute LBP
Pain intensity
Furlan 2005 narratively reported that compared to sham acupuncture, it is uncertain whether acupuncture reduces pain intensity in the short term (1 trial judged as high quality, 40 participants; rated as very low‐certainty evidence).
Function
Furlan 2005 narratively reported that compared to sham acupuncture, it is uncertain whether acupuncture improves function in the short term (1 trial judged as high quality, 40 participants; rated as very low‐certainty evidence).
Safety
There was no study reporting data for safety.
Chronic LBP
Pain intensity
Mu 2020 reported that compared to sham, acupuncture may provide a small reduction in pain intensity in the short term (0 to 100 pain intensity scale, mean difference (MD) ‐10, 95% confidence interval (CI) ‐17.2 to ‐2.9; I2 = 76%; 5 trials, 1095 participants; low‐certainty evidence) and probably provides little to no difference in the intermediate term (0 to 100 pain intensity scale, MD ‐3.8, 95% CI ‐6.6 to ‐1.1; I2 = 0%; 4 trials, 1138 participants; moderate‐certainty evidence). The review included RCTs with different sham acupuncture protocols: superficial needling on sites next to acupoints without stimulation, non‐penetrating needling on sites next to acupoints, and non‐functioning TENS treatment on acupoints.
Function
Mu 2020 reported that compared to sham, acupuncture probably provides a small improvement in function in the short term (standardised mean difference (SMD) ‐0.38, 95% CI ‐0.69 to ‐0.07; I2 = 67%; 3 trials, 957 participants; moderate‐certainty evidence) and the intermediate term (SMD ‐0.16, 95% CI ‐0.27 to ‐0.05; I2 = 0%; 4 trials, 1373 participants; moderate‐certainty evidence). Acupuncture may provide a small improvement in function, measured by the 0 to 70 Pain Disability Index, in the short term (MD ‐3.3, 95% CI ‐6.4 to ‐0.2; I2 = 0%; 2 trials, 285 participants; low‐certainty evidence) and little to no difference in the intermediate term (MD ‐2.9, 95% CI ‐6.3 to ‐0.5; I2 = 0%; 2 trials, 285 participants; low‐certainty evidence).
Safety
Mu 2020 reported that there may be no difference between acupuncture and sham for safety outcomes in the short term (RR 0.68, 95% CI 0.42 to 1.10; I2 = 0%; 4 trials, 465 participants; low‐certainty evidence).
Minor outcomes
Participant ratings of improvement, defined as global perceived effect
Mu 2020 reported that compared to sham acupuncture, acupuncture probably provides a small increase in global assessment of therapy effectiveness in the short term (1‐ to 6‐point improvement scale, MD ‐0.3, 95% CI ‐0.49 to ‐0.1; 1 trial, 747 participants; moderate‐certainty evidence) and the intermediate term (1‐ to 6‐point improvement scale, MD ‐0.2, 95% CI ‐0.39 to ‐0.01; one trial, 751 participants; moderate certainty evidence).
Health‐related quality of life
Mu 2020 reported that compared to sham acupuncture, acupuncture may provide a small improvement in quality of life in the short term (SMD 0.24, 95% CI 0.03 to 0.45; I2 = 0%; 3 trials, 1068 participants; low‐certainty evidence) and probably provides a small improvement in quality of life in the intermediate term (SMD 0.21, 95% CI 0.09 to 0.33; 3 trials, 1066 participants; moderate‐certainty evidence).
Electrophysical modalities versus placebo/sham
Three reviews investigated the effects of electrophysical modalities on LBP (Ebadi 2020; Khadilkar 2008; Yousefi‐Nooraie 2008), with all three reviews comparing electrophysical modalities with placebo/sham. Khadilkar 2008, judged at high confidence, included four trials with 585 participants and investigated the effects of TENS on LBP. Yousefi‐Nooraie 2008, judged at high confidence, included seven trials with 384 participants and investigated the effects of low‐level laser therapy for low back pain. Ebadi 2020, judged at low confidence, included 10 trials with 1025 participants and investigated the effects of therapeutic ultrasound on chronic LBP.
Chronic LBP
Pain intensity
TENS: Khadilkar 2008 narratively reported that TENS may provide little to no difference in pain intensity in the short term compared to placebo (3 trials judged as high quality, 235 participants; rated as low‐certainty evidence).
Low‐level laser therapy: Yousefi‐Nooraie 2008 narratively reported that low‐level laser may provide a small reduction in pain intensity in the short term (three trials judged as high quality, 168 participants) and intermediate term (1 trial judged as high quality, 71 participants) compared to sham intervention. The review did not combine the effects of these studies due to clinical and outcome measures diversity. One study included only elderly participants and reported the results as the percentage of pain relief described by participants as 'excellent', 'good', 'fair' and 'poor'. In this study, at the six‐month follow‐up, 44.7% of the participants in the laser therapy group and 15.2% of the control group reported excellent relief. Another study reported the results as the percentage of pain relief using excellent, good and fair grades. The percentage of effective treatment, not specified in the study, was 94% (15/16) in the laser therapy group and 48% (12/25) in the sham group. One study reported the results using the difference between means. The review reported that laser therapy may provide a medium reduction in pain intensity in the short term compared to sham laser therapy (0 to 100 pain intensity scale, MD ‐16, 95% CI ‐27.95 to ‐4.05; 1 study, 56 participants; low‐certainty evidence).
Therapeutic ultrasound: Ebadi 2020 reported that compared to placebo, it is uncertain whether therapeutic ultrasound reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐7.1, 95% CI ‐18 to 3.70; I2 = 77%; 3 trials, 121 participants; very low‐certainty evidence).
Function
TENS: Khadilkar 2008 narratively reported there may be no difference between TENS and placebo in pain intensity in the short term (2 trials judged as high quality, 410 participants; rated as low‐certainty evidence).
Therapeutic ultrasound: Ebadi 2020 reported that therapeutic ultrasound may provide a small improvement in function in the short term compared to placebo (SMD ‐0.29, 95% CI ‐0.51 to ‐0.07; I2 = 0%; four trials, 325 participants; low‐certainty evidence).
Safety
TENS: Khadilkar 2008 narratively reported that it is uncertain whether there is a difference between TENS and placebo (detuned TENS) in the short term (1 trial with high risk of bias, 145 participants; rated as very low‐certainty evidence). The review reported that a few participants from both groups developed a minor skin irritation at the site of electrode placement.
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported there may be no difference in safety between low‐level laser therapy and sham (detuned laser) in the short term (2 trials judged as high quality, 78 participants; rated as low‐certainty evidence).
Therapeutic ultrasound: Ebadi 2020 reported that one trial reported data for safety (unclear risk of bias for blindness, 225 participants). The review reported that serious adverse events occurred within both therapeutic ultrasound and placebo groups and were unrelated to the intervention (5.9% in the ultrasound group and 3.7% in the placebo group; rated as low‐certainty evidence).
Minor outcomes
Health‐related quality of life
TENS: Khadilkar 2008 narratively reported results from two trials for health‐related quality of life (2 trials with high risk of bias, 205 participants; rated as low‐certainty evidence). The review reported that one trial reported that TENS may not improve quality of life, while another reported that TENS may improve some domains of the quality‐of‐life questionnaire.
Therapeutic ultrasound: Ebadi 2020 reported that compared to placebo, therapeutic ultrasound probably does not improve quality of life in the short term (0 to 100 SF‐36 scale, MD ‐2.7, 95% CI ‐9.9 to 4.4; I2 = 29%; 2 trials, 267 participants; moderate‐certainty evidence).
Workplace participation
TENS: Khadilkar 2008 narratively reported that compared to placebo, it is uncertain whether TENS reduces sick leave in the short term (1 trial with high risk, 350 participants; rated as very low‐certainty evidence).
Mixed LBP
Pain intensity
No studies reported data for pain intensity.
Function
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported that low‐level laser therapy may provide a small improvement in function in the short term compared to sham (0 to 100 Oswestry Disability Index (ODI), MD ‐8.2, 95% CI ‐13.4 to ‐2.9; 1 trial with high risk of bias for allocation concealment, 56 participants; low‐certainty evidence).
Safety
Low‐level laser therapy: Yousefi‐Nooraie 2008 narratively reported that it is uncertain whether there is a difference between low‐level laser therapy and sham on safety in the short term (2 trials, 1 judged as low quality and 1 judged as high quality, 112 participants; rated as very low‐certainty evidence). The review reported that no participants from one trial reported either discomfort or increased pain related to laser or sham. In another trial, two participants in the low‐level laser therapy group abandoned treatment and three needed to use anti‐inflammatories; and three who abandoned treatment and six participants needing anti‐inflammatories were lost to follow‐up.
Minor outcomes
Participant ratings of improvement, defined as global perceived effect
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported that compared to sham, low‐level laser therapy may not increase participant ratings of improvement in the short term (0 to 100 improvement scale, MD ‐9.5, 95% CI ‐20.9 to 1.9; 1 trial with high risk of bias for allocation concealment, 56 participants; rated as low‐certainty evidence).
Exercise therapies versus placebo/sham
Seven Cochrane reviews investigated the effects of exercise on LBP (Almeida 2023; Hayden 2005; Hayden 2021a; Macedo 2016; Saragiotto 2016; Wieland 2022; Yamato 2015). Of these, only Hayden 2021a compared exercise therapies to placebo/sham. Hayden 2021a, judged at high confidence, included 249 trials with 24,486 participants, and investigated the effects of all types of exercises in participants with chronic LBP.
Chronic LBP
Pain intensity
Hayden 2021a conducted a post hoc analysis including only trials that used placebo as a comparison (detuned electrophysical modality). Compared to placebo, exercise may provide a small reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐8.9, 95% CI ‐15.10 to ‐2.60; I2 = 0%; 2 trials with low risk of bias, 300 participants; rated as low‐certainty evidence).
Function
Hayden 2021a conducted a post hoc analysis including only trials that used placebo as a comparison. The review reported that compared to placebo, exercise probably provides a small improvement in function in the short term (0 to 100 pain intensity scale, MD ‐5.6, 95% CI ‐11.20 to ‐0.01; I2 = 33%; 3 trials with low risk of bias, 400 participants; rated as moderate‐certainty evidence).
Safety
No studies reported data for safety.
Heat and cold therapies versus placebo/sham
One review, judged at low confidence, included five trials with 856 participants, and investigated the effects of superficial heat or cold on LBP (French 2006).
Acute/subacute LBP
Pain intensity
French 2006 reported that:
compared to oral placebo, heated back wrap may provide a medium reduction in pain intensity in the short term (after 5 days) (6‐point verbal pain intensity scale, MD ‐1.06, 95% CI ‐0.68 to ‐1.45; I2 = 0%; 2 trials judged as high quality, 258 participants; rated as low‐certainty evidence);
compared to sham (no heated back blanket), it is uncertain whether heated back blanket reduces pain intensity in the short term (immediately after application) (0 to 100 pain intensity scale, MD ‐32.2, 95% CI ‐38.6 to ‐25.7; 1 trial judged as high quality, 90 participants; rated as very low‐certainty evidence).
Function
French 2006 reported that compared to oral placebo, heated back wrap may provide a medium improvement in function in the short term (0 to 24 Roland‐Morris Disability Questionnaire (RMDQ), MD ‐2.1, 95% CI ‐3.07 to ‐1.18; I2 = 0%; 2 trials, 258 participants; rated as low‐certainty evidence).
Safety
French 2006 narratively reported that compared to oral placebo, heated back wrap may result in greater skin pinkness in the short term (6/128 participants in the heated back wrap group compared to 1/130 in the placebo group, 2 trials judged as high quality; rated as low‐certainty evidence).
Herbal therapies versus placebo/sham
Oltean 2014, judged at low confidence, included 14 trials with 2050 participants, and investigated the effects of herbal therapies on LBP.
Acute and subacute LBP
Pain intensity
Oltean 2014 narratively reported that:
compared to placebo, it is uncertain whether capsicum cream (Rado‐Salil) reduces pain intensity in the short term (up to 14 days) (1 trial with high risk of bias, 40 participants; rated as very low‐certainty evidence);
compared to placebo, it is uncertain whether comfrey root extract ointment (S. officinale) reduces pain intensity in the short term (at day 5) (1 trial with high risk of bias, 120 participants; rated as very low‐certainty evidence);
compared to placebo, it is uncertain whether Brazilian arnica extract gel (S. chilensis) reduces pain intensity in the short term (at day 15) (1 trial with high risk of bias, 20 participants; rated as very low‐certainty evidence).
Function
No studies reported data for function.
Safety
No studies reported data for safety.
Chronic LBP
Pain intensity
Oltean 2014 narratively reported that:
two studies compared the effect of H. procumbens extracts (standardised to 50 to 100 mg harpagoside per day) with placebo (two trials with unclear allocation concealment, 315 participants). One trial reported that there may be no difference between H. procumbens extracts and placebo in the short term, and another trial did not report the difference between groups (low‐certainty evidence);
compared to placebo, a higher dose (240 mg) of willow bark extract (S. alba) probably reduces pain intensity in the short term (2 trials with unclear allocation concealment, 261 participants; moderate‐certainty evidence);
compared to placebo, topical capsicum cream or plaster (C. frutescens) may reduce pain intensity (3 trials using unusual measures for pain intensity, 755 participants; moderate‐certainty evidence).
Function
No studies reported data for function.
Safety
No studies reported data for safety.
Manual therapies versus placebo/sham
Five Cochrane reviews investigated the effects of manual therapies on LBP (Franke 2015; Furlan 2015; Rubinstein 2011; Rubinstein 2012; Wegner 2013), with four of five reviews comparing manual therapies to placebo/sham (Franke 2015; Rubinstein 2011; Rubinstein 2012; Wegner 2013). Two reviews, judged at high confidence, investigated the effects of spinal manipulation therapy for acute LBP (20 trials, 2674 participants) (Rubinstein 2012), and chronic LBP (26 trials, 6070 participants) (Rubinstein 2011). Wegner 2013, judged at high confidence, included 32 trials with 2762 participants, and investigated the effects of traction on LBP. Furlan 2015, judged at high confidence, included 25 trials with 3096 participants, and investigated the effects of massage for LBP. Franke 2015, judged at low confidence, included 12 trials with 500 participants and investigated the effects of muscle energy technique on LBP.
Acute and subacute LBP
Pain intensity
Spinal manipulation: Rubinstein 2012 reported that:
compared to sham spinal manipulation, it is uncertain whether spinal manipulation reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐5 favouring spinal manipulation, 95% CI ‐13.9 to 3.90; 1 trial, 74 participants; very low‐certainty evidence);
compared to placebo (detuned electrophysical modality), spinal manipulation may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐12, 95% CI ‐20 to ‐3.90; 1 trial, 178 participants; low‐certainty evidence).
Muscle energy therapy: Franke 2015 reported that compared to sham muscle energy therapy, it is uncertain whether muscle energy therapy reduces pain intensity in the short term (0 to 100 pain intensity scale, MD 15.2 favouring sham muscle energy therapy, 95% CI ‐10.10 to 38.50; 1 trial, 20 participants; very low‐certainty evidence).
Function
Rubinstein 2012 reported that:
compared to placebo (detuned electrophysical modality), spinal manipulation may not improve function in the short term (SMD ‐0.27 favouring spinal manipulation, 95% CI ‐0.58 to 0.04; 1 trial, 178 participants; low‐certainty evidence), and probably does not improve function at very short‐term (one‐week) follow‐up (SMD ‐0.08 favouring spinal manipulation, 95% CI ‐0.37 to 0.21; I2 = 0%; 2 trials, 205 participants; moderate‐certainty evidence);
compared to sham spinal manipulation, it is uncertain whether spinal manipulation improves function in the short term (SMD ‐0.35 favouring spinal manipulation, 95% CI ‐0.76 to 0.06; 1 trial, 94 participants; very low‐certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Participant ratings of improvement
Spinal manipulation: Rubinstein 2012 reported that compared to placebo (detuned electrophysical modality), spinal manipulation may not increase participant ratings of improvement in the short term (RR 0.97, 95% CI 0.85 to 1.10; 1 trial, 239 participants; low‐certainty evidence).
Chronic LBP
Pain intensity
Rubinstein 2011 reported that:
compared to placebo (detuned electrophysical modality), it is uncertain whether spinal manipulation reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐6.0 favouring spinal manipulation, 95% CI ‐15.80 to 3.80; 1 trial, 72 participants; very low‐certainty evidence);
compared to sham manipulation, it is uncertain whether spinal manipulation reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3.2 favouring spinal manipulation, 95% CI ‐13.60 to 7.20; I2 = 53%; 3 trials, 148 participants; very low‐certainty evidence) and in the intermediate term (0 to 100 pain intensity scale, MD 7.1 favouring sham, 95% CI ‐5.20 to 19.40; 1 trial, 55 participants; very low‐certainty evidence).
Wegner 2013 reported that compared to sham traction (without traction force), traction probably does not reduce pain intensity in the short term (0 to 100 pain intensity scale, MD ‐4 favouring traction, 95% CI ‐17.70 to 9.70; 1 trial, 60 participants; moderate‐certainty evidence).
Function
Rubinstein 2011 reported that compared to sham manipulation, it is uncertain whether spinal manipulation improves function in the short term (SMD ‐0.45 favouring spinal manipulation, 95% CI ‐0.97 to 0.06; 1 trial, 55 participants; very low‐certainty evidence) and the intermediate term (SMD 0, 95% CI ‐0.16 to 0.56; 1 trial, 55 participants; very low‐certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Participant ratings of improvement
Spinal manipulation: Rubinstein 2011 narratively reported that compared to placebo (detuned electrophysical modality), it is uncertain whether spinal manipulation increases participant ratings of improvement in the short term (rated as very low‐certainty evidence). One trial demonstrated a significant difference favouring spine manipulation in the short term (six weeks) but not at three months. One trial did not report enough information for this comparison.
Workplace participation
Spinal manipulation: Rubinstein 2011 reported that compared to placebo (detuned electrophysical modality), it is uncertain whether spinal manipulation improves workplace participation in the short term (risk ratio (RR) 1.29 favouring spinal manipulation, 95% CI 1.00 to 1.65; 1 trial, 72 participants; very low‐certainty evidence).
Neuroreflexotherapy versus placebo/sham
Urrutia 2004, judged at low confidence, included three trials with 273 participants, and investigated the effects of neuroreflexotherapy (i.e. temporary implantation of several epidermal devices into trigger points in the back and into referred tender points in the ear) in people with LBP.
Mixed LBP
Pain intensity
Urrutia 2004 narratively reported that compared to sham, it is uncertain whether neuroreflexotherapy reduces pain intensity in the short term (2 trials at low risk of bias using unusual pain intensity measure for spontaneous pain, referred pain and tenderness pain, 89 participants; rated as very low‐certainty evidence).
Function
Urrutia 2004 narratively reported that compared to sham, it is uncertain whether neuroreflexotherapy improves function in the short term (COOP Functional Chart, 88% substantial improvement for neuroreflexotherapy and 3% substantial improvement for sham neuroreflexotherapy, with 2 trials at low risk of bias, 169 participants; rated as very low‐certainty evidence).
Safety
Urrutia 2004 narratively reported that it is uncertain whether there is a difference between neuroreflexotherapy and sham neuroreflexotherapy on safety in the short term. The review reported data from one trial showing that 65% of the participants experienced adverse effects in the sham neuroreflexotherapy group and 9% experienced adverse effects in the neuroreflexotherapy group (1 trial, 91 participants; rated as very low‐certainty evidence). The main adverse effects were gastric discomfort secondary to anti‐inflammatory consumption. The review reported no difference between the groups in discomfort in the skin tautness, and none of the adverse effects were important enough to require early extraction of the epidermal devices.
Minor outcomes
Health‐related quality of life
Urrutia 2004 narratively reported that compared to sham, it is uncertain whether neuroreflexotherapy improves quality of life in the short term (1 trial with low risk of bias, 78 participants; rated as very low‐certainty evidence).
Workplace participants
Urrutia 2004 narratively reported that compared to sham, it is uncertain whether neuroreflexotherapy reduces the number of days off work in the short term (4.7 days off work in the neuroreflexotherapy group and 12.5 days off work in the sham neuroreflexotherapy) (1 trial with low risk of bias, 91 participants; rated as very low‐certainty evidence).
Acupuncture and dry needling versus no treatment/usual care
Two Cochrane reviews investigated the effects of acupuncture and dry needling on LBP (Furlan 2005; Mu 2020). Only Mu 2020 compared acupuncture to no treatment/usual care. The review, judged at high confidence, included 33 trials with 8270 participants with chronic LBP (Mu 2020).
Chronic LBP
Pain intensity
Mu 2020 reported that:
compared to no treatment, acupuncture probably provides a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.1, 95% CI ‐16.80 to ‐3.40; I2 = 0%; 3 trials, 144 participants; moderate‐certainty evidence);
compared to usual care, it is uncertain whether acupuncture reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐18.8 favouring acupuncture, 95% CI ‐38 to 0.40; I2 = 95%; 2 trials, 817 participants; very low‐certainty evidence), but it probably provides a medium reduction in pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐12.3, 95% CI ‐15.30 to ‐9.30; I2 = 0%; 2 trials, 804 participants; moderate‐certainty evidence);
when added to usual care, acupuncture may provide a small reduction in pain intensity in the short term (SMD ‐0.49, 95% CI ‐0.78 to ‐0.20; I2 = 0%; 2 trials, 187 participants; low‐certainty evidence).
Function
Mu 2020 reported that:
compared to no treatment, acupuncture probably provides a small improvement in function in the short term (SMD ‐0.39, 95% CI ‐0.72 to ‐0.06; I2 = 0%; 3 trials, 144 participants; moderate‐certainty evidence);
compared to usual care, acupuncture may provide a medium improvement in function, as measured by the 0 to 70 Pain Disability Index, in the short term (MD ‐11, 95% CI ‐17.20 to ‐4.80; 1 trial, 74 participants; low‐certainty evidence), but its beneficial effect is uncertain in the intermediate term (MD ‐6.4 favouring acupuncture, 95% CI ‐13.10 to ‐0.30; 1 trial, 63 participants; very low‐certainty evidence). Acupuncture probably provides a small improvement in function, as measured by the 0 to 100 Hannover Function Ability Questionnaire (HFAQ), in the short term (MD 9.4, 95% CI 6.2 to 12.7; 1 trial, 734 participants; moderate‐certainty evidence) and the intermediate term (SMD ‐0.44, 95% CI ‐0.55 to ‐0.33; I2 = 0%; 4 trials, 1310 participants; moderate‐certainty evidence);
when added to usual care, acupuncture may provide a medium improvement in function in the short term (SMD ‐0.51, 95% CI ‐0.80 to ‐0.22; I2 = 0%; 2 trials, 187 participants; low‐certainty evidence).
Safety
Mu 2020 did not report data for safety.
Minor outcomes
Participant ratings of improvement
Mu 2020 reported that compared to usual care, acupuncture probably provides a small increase in global assessment of therapy effectiveness in the short term (6‐point improvement scale, MD ‐0.3, 95% CI ‐0.5 to ‐0.1; 1 trial, 730 participants; moderate‐certainty evidence) and the intermediate term (1‐6 participant ratings of improvement scale, MD ‐0.7, 95% CI ‐1.0 to ‐0.3; 1 trial, 738 participants; moderate‐certainty evidence).
Workplace participation
Mu 2020 reported that compared to no treatment, acupuncture probably provides a medium increase in workplace participation in the short term (SMD ‐0.56, 95% CI ‐0.64 to ‐0.49; I2 = 0%; 2 trials, 2937 participants; moderate‐certainty evidence).
Quality of life
Mu 2020 reported that:
compared to no treatment, acupuncture may provide small improvements in physical health‐related quality of life (0 to 100 SF‐36 physical quality of life scale, MD 5.1, 95% CI 3.1 to 7.1; I2 = 58%; 2 trials, 2837 participants; low‐certainty evidence) and probably provides little to no improvement in mental health‐related quality of life (0 to 100 SF‐36 mental quality of life scale, MD 1.9, 95% CI 1.1 to 2.6; I2 = 0%; 2 trials, 2837 participants; moderate‐certainty evidence) in the short term;
compared to usual care, acupuncture probably provides little to no improvement in physical health‐related quality of life in the short term (0 to100 SF‐12 physical quality of life scale, MD 4.2, 95% CI 2.8 to 5.6; 1 trial, 731 participants; moderate‐certainty evidence), but provides a small improvement in the intermediate term (0 to 100 SF‐12 physical quality of life scale, MD 5.8, 95% CI 4.4 to 7.2; 1 trial, 737 participants; moderate‐certainty evidence);
compared to usual care, acupuncture probably provides little to no difference in mental health‐related quality of life in the short term (0 to 100 SF‐12 mental quality of life scale, MD 1.9, 95% CI 0.3 to 3.6; 1 trial, 731 participants; moderate‐certainty evidence), but probably provides a small improvement in the intermediate term (0 to 100 SF‐12 mental quality of life scale, MD 5.8, 95% CI 4.4 to 7.2; 1 trial, 737 participants);
when added to usual care, acupuncture may provide a small improvement in quality of life in the short‐term (0 to 100 SF‐36 general health scale, MD 7.0, 95% CI 1.0 to 13.1; 1 trial, 143 participants; low‐certainty evidence).
Advice and education versus no treatment/usual care
Four Cochrane reviews investigated advice or education (Dahm 2010; Engers 2008; Parreira 2017; Poquet 2016), with all four comparing the intervention to no treatment/usual care. Engers 2008, judged at high confidence, included 24 trials with 7139 participants, and investigated the effects of individual patient education for low back pain. Two reviews investigated the effects of back school on acute and subacute LBP (judged at high confidence, four trials with 643 participants) (Poquet 2016), and chronic LBP (judged at high confidence, 31 trials with 4105 participants) (Parreira 2017). Dahm 2010, judged at low confidence, included eight trials with 1923 participants, and compared the effects of advice to rest in bed with advice to stay active for acute LBP.
Acute and subacute LBP
Pain intensity
Engers 2008 narratively reported there may be no difference between individual patient education (pamphlet or a 20‐minute educational session) and no treatment on pain intensity in the short term (2 trials with low risk of bias, 803 participants) and in the intermediate term (1 trial judged as high quality, 314 participants; rated as low‐certainty evidence).
Function
Engers 2008 narratively reported there may be no difference between individual patient education (pamphlet or a 20‐minute educational session) and no treatment on function in the short term (2 trials with high and low risk of bias, 407 participants; rated as low‐certainty evidence), but its beneficial effect is uncertain in the intermediate term (1 trial with low risk of bias, 93 participants; rated as very low‐certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Workplace participation
Engers 2008 narratively reported that:
compared to no treatment, individual education using pamphlets may not improve return to work in the short and intermediate term (1 trial judged as high quality, 489 participants; rated as low‐certainty evidence);
compared to usual care, individual education (20 minutes) may not improve return to work in the short and intermediate term (1 trial at high risk for randomisation, 314 participants; rated as low‐certainty evidence);
compared to usual care, two hours of patient education may increase return to work in the short and intermediate term (2 trials, 1 high quality and another low quality, 1432 participants; rated as low‐certainty evidence).
Chronic LBP
Pain intensity
Back school (a combination of advice and exercise). Parreira 2017 reported that:
compared to no treatment, it is uncertain whether back school reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐6.1 favouring back school, 95% CI ‐10.8 to ‐2.0; I2 = 18.6%; 6 trials, 647 participants; very low‐certainty evidence), or in the intermediate term (0 to 100 pain intensity scale, MD ‐4.3 favouring back school, 95% CI ‐14.4 to 5.7; I2 = 70%; 4 trials, 257 participants; very low‐certainty evidence);
compared to usual care, it is uncertain whether back school reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.2 favouring back school, 95% CI ‐19.1 to ‐1.2; I2 = 62%; 3 trials, 249 participants; very low‐certainty evidence), or in the intermediate term (0 to 100 pain intensity scale, MD ‐9.65 favouring back school, 95% CI ‐22.5 to 3.2; I2 = 89%; 3 trials, 545 participants; very low‐certainty evidence).
Function
Back school: Parreira 2017 reported that:
compared to no treatment, it is uncertain whether back school improves function in the short term (0 to 100 function scale, MD ‐3.4 favouring back school, 95% CI ‐6.7 to ‐0.1; I2 = 0%; 3 trials, 426 participants; very low‐certainty evidence), or in the intermediate term (0 to 100 function scale, MD ‐5.9 favouring back school, 95% CI ‐12.1 to 0.2; I2 = 0%; 3 trials, 191 participants; very low‐certainty evidence);
compared to usual medical care, it is uncertain whether back school improves function in the short term (0 to 100 function scale, MD ‐1.2 favouring back school, 95% CI ‐7.0 to 4.6; I2 = 0%; 2 trials, 130 participants; very low‐certainty evidence), or in the intermediate term (0 to 100 function scale, MD ‐6.3 favouring back school, 95% CI ‐10.9 to ‐1.8; I2 = 0%; 3 trials, 331 participants; very low‐certainty evidence).
Safety
No studies reported data for safety.
Mixed LBP
Pain intensity
Engers 2008 narratively reported that compared to usual care, it is uncertain whether individual patient education (booklet) reduces pain intensity in the short term (1 trial judged as low quality, 68 participants; rated as very low‐certainty evidence).
Function
Engers 2008 narratively reported there may be no difference between individual education (booklet) and usual care in function in the short term (2 studies with high and low quality, 605 participants; rated as low‐certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Participant ratings of improvement
Engers 2008 narratively reported that compared to usual care, individual education (booklet) may not improve participant ratings of improvement in the short term (1 study judged as high quality, 294 participants; rated as low‐certainty evidence).
Workplace participation
Engers 2008 narratively reported that compared to usual care, individual education (booklet) may not improve return to work in the short and intermediate term (2 studies judged as low quality, 936 participants; rated as low‐certainty evidence).
Assistive materials versus no treatment/usual care
One review, judged at moderate confidence, included eight trials with 1361 participants, and investigated the effects of lumbar supports on LBP (Van Duijvenbode 2008).
Chronic LBP
Pain intensity
Van Duijvenbode 2008 narratively reported that compared to no treatment, it is uncertain whether a lumbar support (flexible corset) reduces pain intensity in the short term (1 trial judged as low quality, 79 participants; rated as very low‐certainty evidence).
Function
Van Duijvenbode 2008 narratively reported that compared to no treatment, it is uncertain whether lumbar support (flexible corset) improves function in the short term (1 trial judged as low quality, 79 participants; rated as very low‐certainty evidence).
Safety
Van Duijvenbode 2008 did not report data for safety.
Mixed LBP
Pain intensity
Van Duijvenbode 2008 narratively reported that compared to no treatment, it is uncertain whether lumbar support reduces pain intensity in the short term (4 trials mostly judged as low quality, 1170 participants; rated as very low‐certainty evidence).
Function
Van Duijvenbode 2008 narratively reported that compared to no treatment, it is uncertain whether lumbar support improves function in the short term (3 trials mostly judged as low quality, 410 participants; rated as very low‐certainty evidence).
Safety
Van Duijvenbode 2008 did not report data for safety.
Minor outcomes
Participant ratings of improvement
Van Duijvenbode 2008 narratively reported that compared to no treatment, it is uncertain whether lumbar support increases global impression of improvement in the short term (2 trials judged as low quality, 790 participants; rated as very low‐certainty evidence).
Exercise therapies versus no treatment/usual care
Seven Cochrane reviews investigated the effects of exercise on LBP (Almeida 2023; Hayden 2005; Hayden 2021a; Macedo 2016; Saragiotto 2016; Wieland 2022; Yamato 2015), with all seven comparing exercise to no treatment/usual care. Hayden 2005, judged at high confidence, included 17 trials with 2357 participants, and investigated the effects of all types of exercises in participants with acute and subacute LBP. Hayden 2021a, judged at high confidence, included 249 trials with 24,486 participants, and investigated the effects of all types of exercises in participants with chronic LBP. Yamato 2015, judged at high confidence, included 10 trials with 510 participants, and investigated the effects of Pilates exercise on LBP. Macedo 2016 judged at low confidence, included three trials with 197 participants, and investigated the effects of motor control exercise for acute LBP. Almeida 2023, judged at high confidence, included five trials with 538 participants, and investigated the effects of McKenzie exercises for participants with acute and subacute LBP. Saragiotto 2016, judged as high confidence, included 29 trials with 2431 participants, and investigated the effects of motor control exercise for chronic LBP. Wieland 2022, judged as high confidence, included 21 trials with 2223 participants, and investigated the effects of yoga exercise for chronic LBP.
Acute and subacute LBP
Pain intensity
Exercise: Hayden 2005 reported that for acute LBP, compared to no treatment/usual care, it is uncertain whether exercise reduces pain intensity in the short term (0 to 100 pain intensity scale, MD 0.6 favouring no treatment, 95% CI ‐11.5 to 12.7; I2 = 86%; 3 trials, 491 participants; rated as very low‐certainty evidence); in the intermediate term, there may be no difference (0 to 100 pain intensity scale, MD ‐1.4 favouring exercise, 95% CI ‐9.5 to 6.8; I2 = 62%; 3 trials, 491 participants; rated as low‐certainty evidence); and it is uncertain in the long term (0 to 100 pain intensity scale, MD ‐0.8 favouring exercise, 95% CI ‐7.2 to 5.5; 1 trial with high risk of bias for randomisation, rated as very low‐certainty evidence, 318 participants).
Motor control: Macedo 2016 reported that when added to usual medical care, it is uncertain whether motor control exercise reduces pain intensity for acute LBP in the short term (0 to 100 pain intensity scale, MD ‐9.3 favouring motor control, 95% CI ‐20.4 to 1.8; 1 trial, 42 participants; very low‐certainty evidence).
McKenzie: Almeida 2023 reported that compared to no treatment/usual care, McKenzie exercises may provide a small reduction in pain intensity in the short term (at the 2‐week follow‐up) (0 to 100 pain intensity scale, MD ‐7.3, 95% CI ‐12 to ‐2.6; I2 = 0%; 2 trials, 337 participants; low‐certainty evidence), but not in the intermediate term (0 to 100 pain intensity scale, MD ‐5, 95% CI ‐14.3 to 4.3; I2 = 0%; 1 trial, 180 participants; low‐certainty evidence) for acute and subacute LBP.
Function
Exercise: Hayden 2005 reported that for acute LBP, compared to no treatment/usual care, it is uncertain whether exercise improves function in the short term (0 to 100 function scale, MD ‐2.8 favouring exercise, 95% CI ‐15.4 to 9.7; I2 = 91%; 3 trials, 491 participants; rated as very low‐certainty evidence), the intermediate term (0 to 100 function scale, MD ‐1.4 favouring exercise, 95% CI ‐9.5 to 6.8; I2 = 62%; 3 trials, most judged as low quality, 491 participants; rated as low‐certainty evidence), and the long term (0 to 100 function scale, MD ‐0.8 favouring exercise, 95% CI ‐7.2 to 5.5; 1 trial judged as low quality, 318 participants; rated as low‐certainty evidence).
Motor control: Macedo 2016 reported that when added to usual care, it is uncertain whether motor control reduces pain intensity in the short term (0 to 100 function scale, MD ‐2.4, 95% CI ‐4.9 to 0.1; 1 trial, 45 participants; very low‐certainty evidence) for acute LBP.
McKenzie: Almeida 2023 reported that compared to no treatment/usual care, there is no evidence that McKenzie exercises improve function in the short term (at the 2‐week follow‐up) (0 to 100 function scale, MD ‐2.7 favouring McKenzie, 95% CI ‐7.5 to 2; I2 = 0%; 2 trials, 328 participants; low‐certainty evidence), and the intermediate term (0 to 100 function scale, MD ‐0.9 favouring McKenzie, 95% CI ‐7.3 to 5.6; I2 = 0%; 1 trial, 180 participants; low‐certainty evidence) for acute and subacute LBP.
Safety
No studies reported data for safety.
Chronic LBP
Pain intensity
Exercise: Hayden 2021a reported that compared to no treatment/usual care (including two studies with placebo comparators), exercise probably provides a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐15.2, 95% CI ‐18.3 to ‐12.2; I2 = 75%; 35 trials, 2746 participants; moderate‐certainty evidence), the intermediate term (0 to 100 pain intensity scale, MD ‐10.6, 95% CI ‐15 to ‐6.1; I2 = 79%; 17 trials, 1491 participants; moderate‐certainty evidence), and in the intermediate to long term (closest to 12 months of follow‐up) (0 to 100 pain intensity scale, MD ‐11.8, 95% CI ‐22 to ‐1.6; I2 = 88%; 5 trials, 498 participants; moderate‐certainty evidence).
Motor control: Saragiotto 2016 reported that compared to no treatment/usual care, motor control exercise probably provides a small to medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.01, 95% CI ‐15.7 to ‐4.4; I2 = 36%; 4 trials, 291 participants; moderate‐certainty evidence) and the long term (0 to 100 pain intensity scale, MD ‐13, 95% CI ‐18.5 to ‐7.4; I2 = 0%; 3 trials, 279 participants; moderate‐certainty evidence), and may provide a medium reduction in pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐12.6, 95% CI ‐20.5 to ‐4.7; I2 = 35%; 4 trials, 348 participants; low‐certainty evidence).
Pilates: Yamato 2015 reported that compared to no treatment/usual care, Pilates may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐14.1, 95% CI ‐18.9 to ‐9.2; I2 = 6%; 6 trials, 265 participants; low‐certainty evidence), and probably provides a medium reduction in pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐10.5, 95% CI ‐18.5 to ‐2.6; I2 = 0%; 2 trials, 146 participants; moderate‐certainty evidence).
Yoga: Wieland 2022 reported that compared to no treatment/usual care, yoga exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐11.1, 95% CI ‐14.2 to ‐7.8; I2 = 0%; 5 trials, 258 participants; low‐certainty evidence), probably provides a small reduction in pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐5.4, 95% CI ‐2.6 to ‐2.2; I2 = 40%; 9 trials, 946 participants; moderate‐certainty evidence), but there may be no difference in the long term (0 to 100 pain intensity scale, MD ‐5.9 favouring yoga, 95% CI ‐12.3 to 0.5; I2 = 68%; 3 trials, 521 participants; low‐certainty evidence).
Function
Exercise: Hayden 2021a reported that exercise compared to no treatment/usual care (including three studies with placebo comparators) probably provides a small improvement in function in the short term (0 to 100 function scale, MD ‐6.8, 95% CI ‐8.3 to ‐5.3; I2 = 38%; 38 trials, 2942 participants; moderate‐certainty evidence), and the intermediate term (0 to 100 function scale, MD ‐6.6, 95% CI ‐8.3 to ‐4.8; I2 = 21%; 31 trials, 2555 participants; moderate‐certainty evidence), and little to no difference in the long term (0 to 100 function scale, MD ‐4.4, 95% CI ‐7.3 to ‐1.5; I2 = 0%; 6 trials, 550 participants; moderate‐certainty evidence).
Motor control: Saragiotto 2016 reported that compared to no treatment/usual care, motor control exercise may provide a small improvement in function in the short term (0 to 100 function, MD ‐8.6, 95% CI ‐14.8 to ‐2.5; I2 = 74%; 5 trials, 332 participants; very low‐certainty evidence) and the long term (0 to 100 function scale, MD ‐6, 95% CI ‐9.8 to ‐2.1; I2 = 35%; 4 trials, 348 participants; low‐certainty evidence). In the intermediate term, motor control exercise probably provides a small improvement in function (0 to 100 function scale, MD ‐5.5, 95% CI ‐9.2 to ‐1.8; I2 = 27%; 4 trials, 348 participants; moderate‐certainty evidence).
Pilates: Yamato 2015 reported that compared to usual care, Pilates may provide a small improvement in function in the short term (0 to 100 function scale, MD ‐8.0, 95% CI ‐13.2 to ‐2.7; I2 = 56%; 5 trials, 248 participants; low‐certainty evidence), and probably provides a medium improvement in function in the intermediate term (0 to 100 function scale, MD ‐11.2, 95% CI ‐18.4 to ‐3.9; I2 = 0%; 2 trials, 146 participants; moderate‐certainty evidence).
Yoga: Wieland 2022 reported that compared to no treatment/usual care, yoga exercise probably provides a small improvement in function in the short term compared to no treatment/usual care (0 to 100 function scale, SMD ‐0.41, 95% CI ‐0.61 to ‐0.21; I2 = 6%; 8 trials, 474 participants; moderate‐certainty evidence), and may lead to a small reduction in the intermediate term (0 to 100 function scale, SMD ‐0.36, 95% CI ‐0.5 to ‐0.21; I2 = 38%; 11 trials, 1157 participants; low‐certainty evidence) and the long term (0 to 100 function scale, SMD ‐0.27, 95% CI ‐0.45 to ‐0.10; I2 = 0%; 3 trials, 532 participants, low‐certainty evidence).
Safety
Motor control: Saragiotto 2016 narratively reported that it is uncertain whether there is a difference between motor control and minimal intervention on safety in the short term (1 trial with low risk of bias, 154 participants; rated as very low‐certainty evidence).
Pilates: Yamato 2015 narratively reported that it is uncertain whether there is a difference between Pilates and a minimal intervention on safety in the short term (1 trial with low risk of bias, 86 participants; rated as very low‐certainty evidence).
Minor outcomes
Participant ratings of improvement
Motor control: Saragiotto 2016 reported that compared to usual care, motor control exercise may provide a small increase in participant ratings of improvement in the short term (‐5 to 5 participant ratings of improvement scale, MD 1.3, 95% CI 0.3 to 2.3; 1 trial, 154 participants; low‐certainty evidence), the intermediate term (‐5 to 5 participant ratings of improvement scale, MD 1.2, 95% CI 0.31 to 2.09; 1 trial, 154 participants; low‐certainty evidence), and the long term (‐5 to 5 participant ratings of improvement scale, MD 1.5, 95% CI 0.61 to 2.39; 1 trial, 154 participants; low‐certainty evidence).
Pilates: Yamato 2015 reported that compared to no treatment (booklet), Pilates may provide a small increase in participant ratings of improvement in the short term (‐5 to 5 participant ratings of improvement scale, MD 1.5, 95% CI 0.7 to 2.3; 1 trial, 86 participants; low‐certainty evidence), but there may be no difference in the intermediate term (‐5 to 5 participant ratings of improvement scale, MD 0.7 favouring yoga, 95% CI ‐0.11 to 1.51; 1 trial, 86 participants; low‐certainty evidence).
Quality of life
Yoga: Wieland 2022 reported that:
compared to no treatment/usual care, yoga may provide a small improvement in physical health‐related quality of life in the short term (SMD 0.5, 95% CI 0.05 to 0.95; I2 = 0%; 2 trials, 81 participants; low‐certainty evidence), but there may be no difference in the intermediate term (SMD 0.16, 95% CI ‐0.13 to 0.46; I2 = 52%; 3 trials, 434 participants; low‐certainty evidence), or long term (SMD 0.17, 95% CI ‐0.07 to 0.41; 1 trial, 264 participants; low‐certainty evidence).
Compared to no treatment/usual care, it is uncertain whether yoga improves mental health‐related quality of life in the short term (SMD ‐0.15, 95% CI ‐1.24 to 0.93; I2 = 67%; 2 trials, 81 participants; very low‐certainty evidence), probably does not improve mental health‐related quality of life in the intermediate term (SMD 0.21, 95% CI 0 to 0.41; I2 = 9%; 3 trials, 434 participants; moderate‐certainty evidence), and may not improve mental health‐related quality of life in the long term (SMD 0.07, 95% CI ‐0.17 to 0.31; 1 trial, 264 participants; low‐certainty evidence).
Herbal therapies versus no treatment/usual care
Oltean 2014, judged at low confidence, included 14 trials with 2050 participants, and investigated the effects of herbal therapies on LBP.
Acute and subacute LBP
Pain intensity
Oltean 2014 narratively reported that compared to ‘conventional treatment’, it is uncertain whether lavender essential oil applied by acupressure massage reduces pain intensity in the short term (at four weeks; 1 trial with high risk of bias, 20 participants; rated as very low‐certainty evidence).
Function
Oltean 2014 did not report data for function.
Safety
Oltean 2014 did not report data for safety.
Manual therapies versus no treatment/usual care
Five Cochrane reviews investigated the effects of manual therapies on LBP (Franke 2015; Furlan 2015; Rubinstein 2011; Rubinstein 2012; Wegner 2013), with all five reviews comparing the intervention to no treatment/usual care. Two reviews, judged at high confidence, investigated the effects of spinal manipulation therapy for acute LBP (20 trials with 2674 participants) (Rubinstein 2012), and chronic LBP (26 trials with 6070 participants) (Rubinstein 2011). Wegner 2013, judged at high confidence, included 32 trials with 2762 participants, and investigated the effects of traction on LBP. Furlan 2015, judged at high confidence, included 25 trials with 3096 participants, and investigated the effects of massage for LBP. Franke 2015, judged at low confidence, included 12 trials with 500 participants, and investigated the effects of muscle energy technique on LBP.
Acute/subacute LBP
Pain intensity
Massage: Furlan 2015 reported that compared to usual care (not described in detail), it is uncertain whether massage reduces pain intensity in the short term (SMD ‐1.24, 95% CI ‐1.85 to ‐0.64; 1 trial, 51 participants; very low‐certainty evidence).
Function
Massage: Furlan 2015 reported that compared to usual care (not described in detail), it is uncertain whether massage improves function in the short term (SMD ‐0.5, 95% CI ‐1.06 to 0.06; 1 trial, 51 participants; very low‐certainty evidence).
Spinal manipulation: Rubinstein 2012 reported that there may be no difference between spinal manipulation and ‘inert intervention’ (booklet) in function in the short term (SMD ‐0.27 favouring spinal manipulation, 95% CI ‐0.58 to 0.04; 1 study, 178 participants; low‐certainty evidence).
Safety
Massage: Furlan 2015 reported that there may be no difference between massage and usual care in the short term on safety (risk difference (RD) 0, 95% CI ‐0.07 to 0.07; 1 trial, 51 participants; rated as low‐certainty evidence).
Mixed LBP
Pain intensity
Massage: Furlan 2015 reported that compared to inactive interventions, massage may provide a medium reduction in pain intensity in the short term (SMD ‐0.75, 95% CI ‐0.90 to ‐0.60; I2 = 0%; 7 trials, 761 participants; low‐certainty evidence), but its beneficial effect is uncertain in the intermediate term (SMD 0.02, 95% CI ‐0.15 to 0.18; I2 = 0%; 3 trials, 615 participants; very low‐certainty evidence).
Function
Massage: Furlan 2015 reported that compared to inactive interventions, massage may provide a medium improvement in function in the short term (SMD ‐0.72, 95% CI ‐1.05 to ‐0.39; I2 = 74%; 6 trials, 725 participants; low‐certainty evidence), but its beneficial effect is uncertain in the intermediate term (SMD ‐0.16, 95% CI ‐0.31 to 0.01; I2 = 0%; 3 trials, 615 participants; very low‐certainty evidence).
Safety
Massage: Furlan 2015 reported that it is uncertain whether there is a difference between massage and inactive interventions on safety in the short term (RD 0.06, 95% CI 0 to 0.11; I2 = 72%; 4 trials, 624 participants; very low‐certainty evidence).
Multimodal therapies versus no treatment/usual care
Three reviews investigated the effects of multimodal therapies on LBP (Parry 2019; Schaafsma 2013; Walker 2010). Walker 2010, judged at low confidence, included 12 trials with 2887 participants, and investigated the effects of chiropractic (combination of spinal manipulation, massage, thermotherapies, electrophysical therapies, exercise, nutritional advice, lifestyle modification and education) on LBP. Schaafsma 2013, judged at high confidence, included 25 trials with 4404 participants, and investigated the effects of physical conditioning (combination of exercise, functional restoration, duties modification, education and behavioural approaches) as part of a return‐to‐work strategy to reduce sickness absence for workers with LBP. This review reported separate results for light physical conditioning (fewer than five sessions) and intense physical conditioning (more than five sessions). Parry 2019, judged at high confidence, included 10 trials with 955 participants, and investigated the effects of workplace interventions for increasing standing or walking (e.g. sit‐stand workstation, treadmill workstation, activity tracker, break‐reminding, policy modification, modification to office layout) in sedentary workers with LBP.
Parry 2019, Schaafsma 2013 and Walker 2010 compared multimodal therapies to no treatment/usual care only. However, Walker 2010 did not provide data for pain, function or safety outcomes of chiropractic intervention compared to sham/placebo or no treatment/usual care. Walker 2010 provided data on chiropractic intervention compared to other interventions only. Parry 2019 provided data for pain and function outcomes only. Schaafsma 2013 provided data for workplace participation only.
Acute/subacute LBP
None of the studies reported data for the major outcomes in this population.
Minor outcomes
Workplace participation
Physical conditioning: Schaafsma 2013 reported that compared to usual care:
it is uncertain whether light physical conditioning improves workplace participation (sick leave) in the intermediate term (SMD ‐0.18, 95% CI ‐0.45 to 0.10; 1 trial, 201 participants; rated as very low‐certainty evidence) and may not improve workplace participation (sick leave) in the long term (SMD ‐0.3, 95% CI ‐0.69 to 0.09; 1 trial, 112 participants; rated as low certainty evidence);
it is uncertain whether intense physical conditioning improves workplace participation (return to work) in the intermediate term (SMD ‐0.03, 95% CI ‐0.22 to 0.15; I2 = 74%; 3 trials, 447 participants; rated as very low‐certainty evidence), but intense physical conditioning may provide a small improvement in workplace participation in the intermediate term (SMD ‐0.39, 95% CI ‐0.76 to ‐0.02; 2 trials, 257 participants; rated as low‐certainty evidence).
Chronic LBP
Pain intensity
Workplace interventions: Parry 2019 reported that:
compared to no treatment, it is uncertain whether workplace interventions reduce pain intensity in the short term (SMD ‐0.35, 95% CI ‐0.8 to 0.10; I2 = 0%; 2 trials, 79 participants; rated as very low‐certainty evidence);
compared to no treatment, it is uncertain whether a workplace intervention using a treadmill workstation reduces pain intensity in the short term (RR 1.05 favouring no treatment, 95% CI 0.5 to 2.19; 1 trial, 47 participants; rated as very low‐certainty evidence);
compared to no treatment, it is uncertain whether a workplace intervention using an activity tracker reduces pain intensity in the short term (SMD 0.05 favouring activity tracker, 95% CI ‐0.87 to 0.77; I2 = 0.41; 2 trials, 31 participants; rated as very low‐certainty evidence) or in the long term (0 to 100 pain intensity scale, MD ‐6.9 favouring activity tracker, 95% CI ‐37.2 to 23.4; 1 trial, 9 participants; rated as very low‐certainty evidence);
compared to no treatment, a workplace intervention using multiple components (organisational, environmental and individual strategies) may not reduce pain intensity in the short term (RR 0.93 favouring multi‐component, 95% CI 0.69 to 1.27; I2 = 0%; 3 trials, 107 participants; low‐certainty evidence), intermediate term (0 to 100 pain intensity scale, MD ‐4 favouring multi‐component, 95% CI ‐19.5 to 11.5; 1 trial; low‐certainty evidence), or long term (RR 0.89 favouring multi‐component, 95% CI 0.57 to 1.4; I2 = 0%; 2 trials, 67 participants; low‐certainty evidence).
Function
Workplace interventions: Parry 2019 reported that:
compared to no treatment, multi‐component workplace interventions may provide a small improvement in function in the short term (0 to 100 function scale, MD ‐8.8, 95% CI ‐17.4 to ‐0.14; 1 trial, 27 participants; rated as low‐certainty evidence);
compared to no treatment, a workplace intervention using a sit‐stand workstation may not improve function in the short term (0 to 24 RMDQ, MD 0.4 favouring sit‐stand, 95% CI ‐2.7 to 1.90; 1 trial, 46 participants; rated as low certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Workplace participation
Physical conditioning: Schaafsma 2013 reported that compared to usual care:
light physical conditioning may not improve workplace participation (return to work) in the intermediate term (SMD ‐0.33 favouring light physical conditioning, 95% CI ‐0.67 to 0.02; 1 trial with low risk of bias, 138 participants; low‐certainty evidence) and probably does not improve workplace participation (return to work) in the long term (SMD ‐0.34 favouring light physical conditioning, 95% CI ‐0.69 to 0.01; 1 trial with low risk of bias, 138 participants; moderate‐certainty evidence);
intensive physical conditioning probably provides a small reduction in sick leave in the intermediate term (SMD ‐0.23, 95% CI ‐0.42 to ‐0.03; I2 = 0%; 5 trials, 1093 participants; rated as moderate‐certainty evidence) but probably does not decrease sick leave in the long term (SMD ‐0.26 favouring intensive physical conditioning, 95% CI ‐0.61 to 0.10; I2 = 53%; 3 trials, 297 participants; moderate‐certainty evidence).
Mixed LBP
Pain intensity
Chiropractic care: Walker 2010 narratively reported that 16 out of 106 participants who received chiropractic care reported minor, transient exacerbations of symptoms. The review did not report the number of participants experiencing symptom exacerbations in the no‐treatment group (2 trials with high and low risk of bias, 301 participants; rated as very low‐certainty evidence).
Function
No studies reported data for function.
Safety
No studies reported data for safety.
Minor outcomes
Workplace participation
Schaafsma 2013 narratively reported results from two studies that reported data for workplace participation. Compared to no treatment/usual care, it is uncertain whether a lumbar support intervention improves workplace participation in the short term (2 trials judged as low quality, 550 participants; rated as very low‐certainty evidence).
Multidisciplinary therapies versus no treatment/usual care
Two Cochrane reviews investigated the effects of multidisciplinary biopsychosocial interventions on LBP (Kamper 2014; Marin 2017), with both reviews comparing the interventions to no treatment/usual care. Kamper 2014, judged at moderate confidence, included 41 trials with 6858 participants, and investigated the effects of multidisciplinary biopsychosocial intervention on chronic LBP. Marin 2017, judged at high confidence, included nine trials with 981 participants, and investigated the effects of multidisciplinary biopsychosocial intervention on subacute LBP.
Acute/subacute LBP
Pain intensity
Marin 2017 reported that compared to usual care, it is uncertain whether multidisciplinary therapies reduce pain intensity in the short term (SMD ‐0.4, 95% CI ‐0.74 to ‐0.06; I2 = 44%; 4 trials, 272 participants; very low‐certainty evidence) or the intermediate term (SMD ‐0.34, 95% CI ‐1 to 0.31; I2 = 73%; 2 trials, 155 participants; very low‐certainty evidence), but probably reduces pain intensity in the long term (SMD ‐0.46, 95% CI ‐0.7 to ‐0.21; I2 = 17%; 5 trials, 532 participants; moderate‐certainty evidence).
Function
Marin 2017 reported that compared to usual care, it is uncertain whether multidisciplinary therapies improve function in the short term (SMD ‐0.38, 95% CI ‐0.63 to ‐0.14; I2 = 0%; 4 trials, 272 participants; very low‐certainty evidence) or the intermediate term (SMD ‐0.44, 95% CI ‐1.09 to 0.22; I2 = 72%; 2 trials, 151 participants; very low‐certainty evidence), but may provide a small improvement in function in the long term (SMD ‐0.44, 95% CI ‐0.87 to ‐0.01; I2 = 61%; 3 trials, 240 participants; low‐certainty evidence).
Safety
No studies reported data for safety.
Minor outcomes
Workplace participation
Marin 2017 reported that compared to usual care:
it is uncertain whether multidisciplinary therapies increase the likelihood of returning to work in the long term (OR 3.19 favouring multidisciplinary therapies, 95% CI 1.46 to 6.98; I2 = 0%; 3 trials, 170 participants; very low‐certainty evidence);
it is uncertain whether multidisciplinary therapies reduce sick leave in the short term (1 trial with high risk of bias, 33 participants; rated as very low‐certainty evidence), but multidisciplinary therapies may provide a small reduction in sick leave in the long term (SMD ‐0.38, 95% CI ‐0.66 to ‐0.1; I2 = 40%; low‐certainty evidence);
it is uncertain whether multidisciplinary therapies improve the time to return to work in the long term (1 trial with high risk of bias, 196 participants; rated as very low‐certainty evidence).
Chronic LBP
Pain intensity
Kamper 2014 reported that:
compared to no treatment, multidisciplinary therapies may provide a medium reduction in pain intensity in the short term (SMD ‐0.73, 95% CI ‐1.22 to ‐0.24; I2 = 64%; 3 trials, 213 participants; rated as low‐certainty evidence);
compared to usual care, multidisciplinary therapies probably provide a medium reduction in pain intensity in the short term (SMD ‐0.55, 95% CI ‐0.83 to ‐0.28; I2 = 72%; 9 trials, 879 participants; moderate‐certainty evidence), the intermediate term (SMD ‐0.6, 95% CI ‐0.85 to ‐0.34; I2 = 63%; 6 trials, 740 participants; moderate‐certainty evidence), and the long term (SMD ‐0.21, 95% CI ‐0.37 to ‐0.04; I2 = 25%; 7 trials, 821 participants; moderate‐certainty evidence).
Function
Kamper 2014 reported that:
compared to no treatment, multidisciplinary therapies may provide a small improvement in function in the short term (SMD ‐0.49, 95% CI ‐0.76 to ‐0.22; I2 = 0%; 3 trials, 213 participants; rated as low‐certainty evidence);
compared to usual care, multidisciplinary therapies probably provide a small improvement in function in the short term (SMD ‐0.41, 95% CI ‐0.62 to ‐0.19; I2 = 58%; 9 trials, 939 participants; moderate certainty evidence), intermediate term (SMD ‐0.43, 95% CI ‐0.66 to ‐0.19; I2 = 59%; 6 trials, 786 participants; moderate‐certainty evidence), and long term (SMD ‐0.23, 95% CI ‐0.40 to ‐0.06; I2 = 19%; 6 trials, 722 participants; moderate‐certainty evidence).
Safety
Kamper 2014 narratively reported that one trial found no adverse events in the multidisciplinary therapies group, and it was unclear whether adverse events in the usual care group were recorded (1 trial rated as high risk of bias, 134 participants; rated as very low‐certainty evidence).
Minor outcomes
Quality of life
Kamper 2014 reported that compared to usual care:
it is uncertain whether multidisciplinary therapies improve physical health‐related quality of life in the short term (0 to 100 quality of life scale, MD 13.5 favouring multidisciplinary therapies, 95% CI ‐9.1 to 36; I2 = 93%; 2 trials, 144 participants; rated as very low‐certainty evidence), and the intermediate term (0 to 100 quality of life scale, MD 7.4 favouring multidisciplinary therapies, 95% CI ‐5 to 18.8; I2 = 72%; 2 trials, 144 participants; rated as low‐certainty evidence);
it is uncertain whether multidisciplinary therapies improve quality of life in mental health‐related quality of life in the short term (0 to 100 quality of life scale, MD 15.3 favouring multidisciplinary therapies, 95% CI 2.1 to 28.4; I2 = 73%; 2 trials, 144 participants; rated as very low‐certainty evidence) and intermediate term (0 to 100 quality of life scale, MD 7.6, 95% CI 1.7 to 13.5; I2 = 0%; 2 trials, 144 participants; rated as low‐certainty evidence).
Workplace participation
Kamper 2014 reported that compared to usual care, multidisciplinary therapies probably do not improve workplace participation in the short term (OR 1.07 favouring multidisciplinary therapies, 95% CI 0.6 to 1.9; I2 = 0%; 2 trials, 373 participants; moderate‐certainty evidence), intermediate term (OR 1.6 favouring multidisciplinary therapies, 95% CI 0.52 to 4.91; I2 = 80%; 3 trials, 457 participants; moderate‐certainty evidence), and long term (OR 1.04 favouring multidisciplinary therapies, 95% CI 0.73 to 1.47; I2 = 31%; 7 trials, 1307 participants; moderate‐certainty evidence).
Neuroreflexotherapy versus no treatment/usual care
Urrutia 2004, judged at low confidence, included three trials with 273 participants, and investigated the effects of neuroreflexotherapy (i.e. temporary implantation of several epidermal devices into trigger points in the back and into referred tender points in the ear) in people with LBP.
Mixed LBP
Pain intensity
Urrutia 2004 narratively reported that compared to usual care, it is uncertain whether neuroreflexotherapy reduces pain intensity in the short term (1 trial with low risk of bias using an unusual pain intensity measure for spontaneous pain, referred pain and tenderness pain, 104 participants; rated as very low‐certainty evidence).
Function
Urrutia 2004 narratively reported that compared to usual care, it is uncertain whether neuroreflexotherapy improves function in the short term (1 trial with low risk of bias, 104 participants; rated as low‐certainty evidence).
Safety
Urrutia 2004 narratively reported that two participants receiving usual care experienced epigastralgia from the use of anti‐inflammatory drugs. The study did not report safety outcomes related to skin irritation secondary to the epidermal material (1 trial, 104 participants; rated as very low‐certainty evidence).
Minor outcomes
Quality of life
Urrutia 2004 narratively reported that compared to usual care, it is uncertain whether neuroreflexotherapy improves quality of life (1 trial with low risk of bias, 60 participants; rated as very low‐certainty evidence).
Psychological therapies versus no treatment/usual care
One Cochrane review, judged at low confidence, included 31 trials with 3438 participants, and investigated the effects of psychological therapies on LBP (Henschke 2010). The review included several psychological therapies, including operant approaches (addressing external factors associated with pain such as attention, rest and reduction in pain levels by a proposed quick‐solution intervention), respondent approaches (dealing with the perceived muscle tension through progressive relaxation or biofeedback), cognitive approaches (addressing thoughts, feelings, beliefs or a combination of the three), and cognitive behavioural approaches (combining cognitive and behavioural approaches).
Chronic LBP
Pain intensity
Henschke 2010 reported that:
compared to no treatment, respondent approaches such as progressive relaxation may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐19.8, 95% CI ‐34.3 to ‐5.2; I2 = 57%; 3 trials, 74 participants; low‐certainty evidence);
compared to no treatment, respondent approaches such as EMG (electromyographic) feedback may provide a medium reduction in pain intensity in the short term (SMD ‐0.8, 95% CI ‐1.32 to ‐0.28; I2 = 0%; 3 trials, 64 participants; low‐certainty evidence);
compared to no treatment, operant approaches probably provide a small reduction in pain intensity in the short term (SMD ‐0.43, 95% CI ‐0.75 to ‐0.11; I2 = 0%; 3 trials, 153 participants; moderate‐certainty evidence);
compared to no treatment, cognitive approaches may not reduce pain intensity in the short term (SMD ‐0.27 favouring cognitive approaches, 95% CI ‐0.75 to 0.22; I2 = 10%; 2 trials, 68 participants; low‐certainty evidence);
compared to no treatment, cognitive behavioural approaches may provide a medium reduction in pain intensity in the short term (SMD ‐0.6, 95% CI ‐0.97 to ‐0.22; I2 = 42%; 5 trials, 239 participants; low‐certainty evidence);
compared to usual care, psychological therapies (progressive muscle relaxation and combination of behavioural approaches addressing fear, exercise and self‐management) probably provide a small reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐5.2, 95% CI ‐9.8 to ‐0.6; I2 = 19%; 2 trials, 330 participants; moderate‐certainty evidence), but probably do not reduce pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐4.3, 95% CI ‐9.3 to 0.7; I2 = 0%; 2 trials, 319 participants; moderate‐certainty evidence).
Function
Henschke 2010 reported that:
compared to no treatment, operant approaches may not improve function in the short term (SMD ‐1.18 favouring operant approaches, 95% CI ‐3.53 to 1.18; I2 = 0%; 2 trials, 87 participants; low‐certainty evidence);
compared to no treatment, it is uncertain whether respondent approaches such as EMG feedback improve function in the short term (SMD ‐0.17 favouring EMG feedback, 95% CI ‐1.56 to 1.22; I2 = 78%; 2 trials, 44 participants; very low‐certainty evidence);
compared to no treatment, respondent approaches such as progressive relaxation may provide a large improvement in function in the short term (SMD ‐0.88, 95% CI ‐1.36 to ‐0.39; I2 = 0%; 3 trials, 74 participants; low‐certainty evidence);
compared to no treatment, cognitive approaches may not improve function in the short term (SMD ‐0.15 favouring cognitive approaches, 95% CI ‐0.64 to 0.33; I2 = 0%; 2 trials, 68 participants; low‐certainty evidence);
compared to no treatment, cognitive behavioural approaches may not improve function in the short term (SMD ‐0.15 favouring cognitive behavioural approaches, 95% CI ‐0.64 to 0.33; I2 = 49%; 4 trials, 134 participants; low‐certainty evidence);
compared to usual care, psychological therapies combined (progressive muscle relaxation and combination of behavioural approaches addressing fear, exercise and self‐management) may not improve function in the short term (SMD ‐0.2 favouring psychological therapies, 95% CI ‐0.41 to 0.02; I2 = 0%; 2 trials, 330 participants; moderate‐certainty evidence), and the intermediate term (SMD ‐0.12 favouring psychological therapies, 95% CI ‐0.34 to 0.10; I2 = 0%; 2 trials, 319 participants; moderate‐certainty evidence).
Safety
Henschke 2010 did not report data for safety.
3. Non‐pharmacological and non‐surgical treatments versus other interventions
Please refer to Appendix 3 for results.
Discussion
Summary of main results
The objective of this overview was to summarise the evidence from Cochrane reviews of non‐pharmacological and non‐surgical interventions for adults with non‐specific LBP. We identified gaps in the evidence for all intervention classes and comparisons of interest. We noted inconsistencies in the approaches used by the included reviews to evaluate the RCT evidence.
We summarised 31 systematic reviews that included 97,183 participants from 644 RCTs. There were 12 intervention classes: acupuncture and dry needling (Furlan 2005; Mu 2020), advice and education (Dahm 2010; Engers 2008; Parreira 2017; Poquet 2016), lumbar support (Van Duijvenbode 2008), electrophysical modalities (Ebadi 2020; Khadilkar 2008; Yousefi‐Nooraie 2008), exercise therapies (Almeida 2023; Hayden 2005; Hayden 2021a; Macedo 2016; Saragiotto 2016; Schaafsma 2013; Wieland 2022; Yamato 2015), heat and cold therapies (French 2006), herbal therapies (Oltean 2014), manual therapies (Franke 2015; Furlan 2015; Rubinstein 2011; Rubinstein 2012; Wegner 2013), multidisciplinary therapies (Kamper 2014; Marin 2017), multimodal therapies (Parry 2019; Schaafsma 2013; Walker 2010), neuroreflexotherapies (Urrutia 2004), and psychological therapies (Henschke 2010). Overall, outcome data for pain intensity were reported more often than function. Safety data were reported in 14 of 31 (45%) reviews. Most reviews rated the evidence as low or very low certainty due to imprecision (i.e. few participants) and the low quality of the included studies. We have high confidence in the findings of 19 reviews (61%) based on the AMSTAR 2 evaluation (Shea 2017).
Most data were available for chronic LBP (413/679 comparisons, 61%), followed by acute or subacute LBP (191/679 comparisons, 28%) and mixed populations or populations for whom pain duration was not clearly reported (75/679 comparisons, 11%). The best available evidence for sham/placebo and no treatment/usual care comparisons were of moderate certainty.
Based on moderate‐certainty evidence, for acute LBP, advice to stay active probably provides a small reduction in pain intensity in the short term compared to advice to rest in bed, and for subacute LBP, multidisciplinary therapies probably provide a small reduction in pain intensity in the long term compared to usual care. Unfortunately, there is no evidence for the effectiveness of advice to stay active compared to no treatment/usual care. The effects of advice to stay active may result from the intervention's effectiveness or the harmful effects of advice to rest. Spinal manipulation probably does not improve function in the very short term (one week) compared to placebo. For chronic LBP, exercise therapies and multidisciplinary therapies probably provide a medium reduction in pain intensity in the short, intermediate and long term compared to no treatment/usual care. Psychological therapies probably provide a small reduction in pain intensity in the short and intermediate term compared to usual care, and acupuncture probably provides a medium reduction in pain intensity compared to no treatment. The effect of acupuncture is reduced when compared to sham/placebo, which may be due to the invasive characteristic of acupuncture, which can increase treatment expectations and pain reduction (Braithwaite 2020; Kaptchuk 2006).
Overall, the size of the effect for LBP treatments is smaller for function outcomes than for pain intensity outcomes. We found that the overall effect of interventions that require participants' active participation, such as exercise and multidisciplinary therapies, appeared to be associated with greater improvements in function than electrophysical modalities and manual therapies for chronic LBP. Psychological therapies probably do not improve function for chronic LBP, though a recent network meta‐analysis found that when psychological therapies (i.e. cognitive behavioural therapy, pain education) are added to physical therapies, they probably provide a small to medium improvement in function for chronic LBP (Ho 2022). Although this overview found that the effects of multimodal therapies are uncertain, recently, high‐quality randomised controlled trials showed that new psychologically informed multimodal therapies might provide long‐term improvements in function compared to placebo (Ashar 2021; Bagg 2022) and usual care (Kent 2023).
We found moderate‐certainty evidence that some non‐pharmacological interventions probably provide other benefits for people with chronic LBP. For example, acupuncture may provide benefits for quality of life, workplace participation, and the impression of improvement in the treatment of chronic LBP in the short term. Intensive physical conditioning (but not multidisciplinary therapies) probably reduces sick leave in the intermediate term compared to usual care for chronic LBP. There is moderate‐certainty evidence that yoga exercise probably does not improve quality of life, at least in the intermediate term for chronic LBP compared to no treatment/usual care. There is also moderate‐certainty evidence that therapeutic ultrasound does not improve quality of life in the short term compared to sham ultrasound for chronic LBP.
Safety of interventions for LBP
A minority of the reviews (12/31, 39%) reported data on safety outcomes in the sham/placebo or no treatment/usual care comparisons due to few trials reporting safety outcomes in non‐pharmacological interventions. Only one review provided an effect estimate of safety outcomes between the intervention and the comparison group (Furlan 2015). The best available evidence for safety is of low or very low certainty. The narrative and quantitative reporting for safety suggests that non‐pharmacological interventions may not be associated with serious adverse events. For chronic LBP, we found low‐certainty evidence that multidisciplinary therapies are associated with fewer adverse events than surgery in the long term.
The best available evidence for a head‐to‐head comparison
The best available evidence for non‐pharmacological interventions compared to other interventions was high‐certainty evidence for chronic LBP. We found high‐certainty evidence that motor control exercise provides little to no difference in function in the intermediate and long term compared to other exercises for chronic LBP. We found high‐certainty evidence that: spinal manipulation provides little to no difference in pain intensity in the short and intermediate term compared to other interventions for mixed LBP duration; spinal manipulation provides little to no difference in function in the short and intermediate term compared to other interventions; and spinal manipulation does not improve quality of life compared to other interventions. Theoretically, the best review design to summarise head‐to‐head comparisons is through network meta‐analysis. Current Cochrane reviews for back pain have not yet included network meta‐analyses.
Overall completeness and applicability of evidence
This overview summarises published Cochrane reviews of all RCTs examining non‐pharmacological interventions for adults with non‐specific LBP. However, because 74% (23/31) of the reviews were published before 2020, and 49% (12/31) were published before 2015, important RCT evidence may not have been included. There is a need to update several Cochrane reviews for non‐pharmacological interventions. Some reviews were updated but not published in the Cochrane Database of Systematic Reviews. For example, an update of the 2012 Cochrane review for spinal manipulation on chronic back pain was published seven years later in The BMJ (Rubinstein 2019), with both reviews having similar conclusions. An updated systematic review of exercises for chronic back pain that included head‐to‐head comparisons of different exercise modalities was published in the Journal of Physiotherapy (Hayden 2021b). There were two main differences: (1) the non‐Cochrane review found McKenzie exercises led to a moderate reduction in pain intensity, whereas our Cochrane overview did not include a review of McKenzie exercises for chronic LBP; and (2) Hayden 2021b found that Pilates showed medium to large effects in reducing pain intensity and moderate improvements in function, while our overview reported a medium reduction in pain and small improvements in function. It is important to highlight that Cochrane systematic reviews on McKenzie exercises and Pilates for chronic LBP are underway and will be published soon. Therefore, additional RCTs identified in recent publications might alter the results of the Cochrane reviews, particularly those relating to Pilates and McKenzie exercises for chronic LBP (Hayden 2021a), multimodal therapies (Bagg 2022), and psychological therapies for chronic LBP (Ho 2022). There are also several non‐pharmacological interventions that were not covered by Cochrane reviews, such as other types of patient education for LBP (e.g. pain science education) (Wood 2019), virtual reality for chronic LBP (Ahern 2020), psychological therapies for acute LBP (Garland 2020), and TENS for acute LBP (Binny 2019). Cochrane reviews with few included RCTs (e.g. motor control for acute LBP (Macedo 2016)) may potentially benefit from updates.
Most reviews included similar measures for pain intensity (0 to 100 visual analogue scale (VAS), 0 to 10 numerical rating scale (NRS)) and function (0 to 100 Oswestry Disability Index (ODI) or 0 to 24 Roland‐Morris Disability Questionnaire (RMDQ)). Fewer reviews reported results for safety (14/31, 45%) or quality of life (9/31, 29%), which have been identified as important patient outcomes for low back pain (Chiarotto 2018). We were not always sure at which time point(s) the reviews reported results for safety outcomes. To improve the consistency of outcome selection and assist decision‐making, trialists and review authors should consider the recommendations of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) (Dworkin 2005), and core outcome domains for non‐specific LBP (Chiarotto 2018).
Comparison groups need to be considered when interpreting reported treatment effect estimates. Reviews reported data for no‐treatment comparisons (20/31, 65%) more often than sham/placebo comparisons (16/31, 52). No‐treatment comparisons control for contextual factors associated with the intervention (e.g. expectation and attention), which may impact the interpretation of the effect of the intervention (Cashin 2021b). Usual care as a comparison is problematic as usual care could vary over time and across regions, making interpretation of effect estimates challenging (Pascoe 2022). We did not identify any sham/placebo comparisons for lumbar supports, advice, education, multimodal therapies, multidisciplinary therapies, and psychological therapies. There are challenges in developing credible sham conditions for complex non‐drug interventions, especially where the active ingredient/element of the treatment is unclear (Hohenschurz‐Schmidt 2022; Moore 2015). There is a need for researchers to use recommended guidance to design comparable sham/placebo interventions (Hohenschurz‐Schmidt 2022), clearly define and report usual care comparisons (Pascoe 2022), and be explicit about the mechanisms and elements of complex interventions (Cashin 2021b).
The quality of the systematic reviews should also be considered when interpreting the results. For example, some systematic reviews judged as low confidence, according to AMSTAR 2, reported the effects on acute LBP of acupuncture (Furlan 2005), advice (Dahm 2010), low‐level laser therapy (Yousefi‐Nooraie 2008), ultrasound (Ebadi 2020), heat and cold therapies (French 2006), herbal therapy (Oltean 2014), muscle energy therapies (Franke 2015), chiropractic (Walker 2010), and neuroreflexology (Urrutia 2004). We organised the effect size, the certainty of the evidence in the effects according to the systematic reviews (i.e. GRADE), and the confidence of the systematic review (i.e. AMSTAR 2) in Table 10 and Table 11. Interventions with moderate‐certainty evidence (i.e. the highest‐certainty evidence in our overview for sham/placebo and no treatment/usual care comparisons) from systematic reviews with high confidence according to AMSTAR 2 should receive special consideration to guide the management of back pain. Of the interventions with moderate‐certainty evidence (i.e. acupuncture, advice to stay active, exercise therapies, traction, multidisciplinary therapies and psychology therapies), we judged only the review that assessed psychological therapies as not having high confidence (Henschke 2010). Therefore, findings regarding psychological therapies should be interpreted with caution.
Quality of the evidence
We used AMSTAR 2 to evaluate the quality of the included systematic reviews. We reached judgements of overall high confidence in 19 of the 31 reviews (60%), overall moderate confidence in two (6%), and overall low confidence in 10 (30%). Only one review explicitly reported that there was no protocol published before the review (Macedo 2016), and one other review did not provide a list of excluded studies (Urrutia 2004). Three reviews did not fully account for the risk of bias in individual studies when interpreting the results of the interventions (Hayden 2021a; Oltean 2014; Parry 2019). Six of the 24 reviews that performed meta‐analysis (25%) did not adequately investigate the presence of publication bias or satisfactorily discuss its likely impact on the results of the review (Table 7 and Figure 1). Although evidence suggests that Cochrane reviews present better overall methodological quality than non‐Cochrane reviews (Goldkuhle 2018; Moseley 2009; Useem 2015), we reached a judgement of low confidence in a third of the included Cochrane reviews of non‐pharmacological interventions for LBP. The publication year may explain the lower confidence rating these reviews attain. The median publication year of the 10 low‐confidence Cochrane reviews was 2013 (range 2004 to 2020), while the median publication year of the nine high‐confidence Cochrane reviews was 2014 (range 2008 to 2022). Standards for conducting and reporting reviews have evolved considerably in recent years (e.g. the development of the AMSTAR 2 instrument (Shea 2017)), and improvements in review quality may take time to be implemented, even with strict Cochrane editorial processes.
Potential biases in the overview process
We conducted this overview according to a published protocol (Cashin 2021b). We used a sensitive search strategy, which was designed under expert guidance from the Cochrane Back and Neck Review Group. The search of the Cochrane Database of Systematic Reviews was conducted from its inception up to April 2023 to identify published reviews and planned or ongoing reviews (Cochrane review protocols). For this reason, we are confident that this overview offers a current summary of Cochrane reviews investigating the effect of non‐pharmacological and non‐surgical interventions in adults with non‐specific LBP.
Many of the included reviews (20/31, 65%) were authored by members of this overview team (ADF, CGM, MWvT). We attempted to reduce potential bias in the appraisal of the included reviews by delegating data extraction and quality assessment to overview authors (RRNR and AGC, MCF, EOH or SS) who were not authors of the included reviews.
We included only Cochrane reviews in this overview and excluded non‐Cochrane reviews. While the inclusion of non‐Cochrane reviews may produce more comprehensive results, published research suggests that non‐Cochrane reviews have lower quality and more imprecise effect estimates than Cochrane reviews (Goldkuhle 2018; Moseley 2009; Useem 2015). Thus, we argue that this overview presents a conservative and trustworthy summary of the effects of non‐pharmacological interventions in adults with LBP. Another limitation is that over three‐quarters (76%) of these reviews were published before 2020. In addition to the possibility of differing interpretations due to the inclusion of recent RCTs in the meta‐analyses, it is important to note that interpretations may change because older reviews used tools that are no longer recommended for assessing the risk of bias and certainty of evidence.
It is important to be aware that we relied on the reporting (effect estimates, risk of bias of the included trials, and quality/certainty of the evidence) of the included reviews. Therefore, the results of our overview may incorporate possible problems with reporting from the original trials and the included systematic reviews. For example, most reviews explicitly stated that they included participants with non‐specific LBP. However, some analyses may also include some participants with sciatica or other specific LBP conditions, despite our intention to exclude analyses where review authors explicitly stated that participants with sciatica were included (Dahm 2010; Walker 2010). We rated the certainty of the evidence using the GRADE approach in 74 comparisons (indicated as ‘rated as’ in the Results section). We judged the certainty of the evidence for a few comparisons (37/685, 5%) using a different criterion for assessing risk of bias (Van Tulder 1997; Van Tulder 2003), which was reported in three reviews (French 2006; Furlan 2015; Van Duijvenbode 2008). We acknowledge that the GRADE judgements (including our judgements) introduce an element of subjectivity in the assessment of the certainty of the evidence, which potentially can lead to inconsistency in the judgements of the certainty of the evidence for some estimates. For example, low‐certainty evidence for the effect of acupuncture versus no treatment would have been judged as very low certainty according to the overview GRADE criteria. In the 'summary tables' and 'Overview tables', we indicated a few instances where our GRADE judgements differed from the reviews’ judgements. Finally, the overall overlap of primary studies was low in the overview.
Agreements and disagreements with other studies or reviews
Previous authoritative reviews and guidelines, such as The Lancet LBP series and American College of Physicians (ACP) guidelines on back pain treatment, provided recommendations based on systematic reviews and individual RCTs to supplement existing systematic reviews with subsequently published RCTs (Foster 2018; Qaseem 2020). We supplemented these authoritative reviews and guidelines by investigating additional outcomes, such as return to work and, particularly, quality of life, which is considered a core outcome for back pain (Chiarotto 2018). We identified two recent overviews that investigated the effect of exercise in people with acute LBP (Karlsson 2020) and chronic LBP (Grooten 2022). These overviews reached similar conclusions regarding the effects and certainty of the evidence of exercises on LBP. However, our overview was more conservative in judging the certainty of the evidence on function (very low‐certainty instead of low‐certainty evidence) for motor control exercise and for the evidence on pain intensity for Pilates (very low‐certainty instead of low‐certainty evidence), and less conservative for yoga (moderate‐certainty instead of low‐certainty evidence). The difference in the certainty of evidence in these overviews might be due to the year of the included reviews, which might have included new RCTs, or the subjective element of the GRADE approach.
We identified some discrepancies between the guideline recommendations and our overview. We identified that advice to stay active, a recurrent recommendation in clinical practice guidelines for acute and subacute LBP (Oliveira 2018), has not been compared to no treatment or sham/placebo. The beneficial effect of advice to stay active was observed against advice to rest. This type of comparison does not clarify whether advice to stay active is beneficial or advice to rest is harmful to people with LBP. We found moderate‐certainty evidence that multidisciplinary therapies reduce pain intensity for subacute LBP. However, clinical practice guidelines rarely recommend this intervention for subacute LBP (Foster 2018; Oliveira 2018). In contrast, a few clinical guidelines recommend spinal manipulation for subacute LBP as a second‐line or adjuvant treatment option (Foster 2018). We found low‐certainty evidence for no evidence of a difference in pain intensity between spinal manipulation and sham or spinal manipulation plus usual care for subacute LBP. We found only low‐certainty evidence that spinal manipulation provides a medium reduction in pain intensity compared to placebo. Massage is also commonly recommended to treat acute/subacute LBP. Despite this, we found only very low‐certainty evidence that massage provides a large reduction in pain intensity compared to usual care for acute/subacute LBP, and low‐certainty evidence that massage provides a medium effect in pain intensity compared to no treatment for mixed pain duration of LBP. The clinical guidelines' recommendations for spinal manipulation and massage over multidisciplinary therapies for acute/subacute LBP may be related to the complexity of offering multidisciplinary therapies in primary care. In 2016 and 2020, the National Institute for Health and Care Excellence (NICE) in the United Kingdom was against acupuncture for low back pain (NICE 2016). However, in 2021, the NICE guideline provided recommendations for acupuncture for chronic primary pain (which includes chronic low back pain) (NICE 2021). The 2021 NICE guideline acknowledged the limitation of studies blinding acupuncture but decided to recommend acupuncture for chronic pain due to some evidence of the beneficial effects of the intervention compared to sham intervention (NICE 2021).
Authors' conclusions
Implications for practice.
This overview summarises the evidence from Cochrane reviews of randomised controlled trials (RCTs) of non‐pharmacological and non‐surgical interventions for adults with non‐specific low back pain (LBP). It included 31 Cochrane reviews assessing a total of 644 RCTs. Researchers, clinicians and policy‐makers can use this overview to assist them in decision‐making and knowledge translation.
For acute/subacute LBP, we found that advice to stay active probably provides a small reduction in pain intensity compared to advice to rest in bed for acute LBP. We found that multidisciplinary therapies probably provide a small reduction in pain intensity in the long term for subacute LBP. We found low‐ or very low‐certainty evidence for the effects of massage and spinal manipulation on pain intensity.
For chronic LBP, we found that acupuncture probably provides a small to medium reduction in pain intensity and a small improvement in function, quality of life, perception of improvement and workplace participation compared to no treatment. We found that exercises (e.g. motor control, Pilates, yoga) probably provide a small to medium reduction in pain intensity and a small improvement in function. We found that traction probably does not reduce pain intensity. We found that multidisciplinary therapies probably provide a small to medium reduction in pain intensity and improvement in function, but probably do not improve workplace participation. We found that psychological therapies (e.g. operant approaches and progressive relaxation) probably provide a small reduction in pain intensity. We found that multimodal therapies, such as light physical conditioning as part of a return‐to‐work strategy, probably do not increase return to work of participants with chronic LBP, but intensive physical conditioning probably provides a reduction in sick leave. There is low‐ or very low‐certainty evidence that non‐pharmacological and non‐surgical interventions are not associated with serious adverse events. We acknowledge that some of the evidence from these reviews is more than 10 years old, and implications for practice may change with the inclusion of newer RCTs.
Our overview is consistent with recommendations from most current international clinical practice guidelines for non‐pharmacological and non‐surgical treatments of LBP. Our overview identified that most of these recommendations for acute and chronic low back pain are based on moderate‐ and low‐certainty evidence.
Effective non‐pharmacological and non‐surgical interventions have moderate‐certainty evidence, at best. Most of these effective interventions provide a small effect on pain and function in people with LBP. There are more effective treatment options for chronic LBP than for acute LBP. For chronic LBP, some interventions probably improve quality of life and workplace participation, in addition to improving pain and function. Non‐pharmacological and non‐surgical interventions appear to be relatively safe, but this evidence is of low or very low certainty. There is a need to prioritise non‐pharmacological and non‐surgical interventions with moderate‐certainty evidence over those with low‐ or very low‐certainty evidence in the current model of care for LBP.
Implications for research.
There is a need to update most of the published Cochrane reviews and complete the 10 published Cochrane review protocols on non‐pharmacological interventions for LBP. These protocols aim to evaluate the effects of new or current interventions in a particular population with non‐specific LBP or with a specific comparison group. We recommend that future updated reviews follow updated guidance from the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2024). The implementation of updated guidance might improve the consistency of methods and the confidence in future Cochrane systematic reviews. Researchers should consider critical domains on AMSTAR 2 (A MeaSurement Tool to Assess Systematic Reviews). In some cases, future reviews should incorporate formal screening for trustworthiness. For example, an included Cochrane systematic review identified duplication of data in two eligible studies, which may be missed in other systematic reviews (Hayden 2021a). Some tools have been proposed for evaluating the integrity and trustworthiness of clinical studies, but there is a need for further evaluation of the validity and performance of these tools (Grey 2020; Parker 2022).
New RCTs investigating non‐pharmacological or non‐surgical interventions should follow core outcomes for non‐specific LBP (Chiarotto 2018), as well as the IMMPACT (Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials) recommendations for chronic pain outcomes (Dworkin 2005), to reduce inconsistencies between trials and improve the synthesis of results in the systematic reviews. Future RCTs should also follow the Consolidated Standards of Reporting Trials (CONSORT) statement to facilitate complete and transparent reporting and a full description of the trial, including participants’ demographic and clinical characteristics and methodological procedures (Schulz 2010). Details of the intervention and comparison groups should be reported using the TIDieR (Template for Intervention Description and Replication) reporting guidelines (Hoffmann 2014).
There are many trials comparing non‐pharmacological and non‐surgical interventions with other interventions. The number of trials is expected to increase since non‐pharmacological interventions are commonly used in combination with pharmacological or other non‐pharmacological interventions in clinical practice, creating new potential interventions for LBP. Network meta‐analysis is considered the most appropriate analytical method to compare several interventions for a clinical problem with each other (Bagg 2018). Our overview did not find any Cochrane review investigating the effect of non‐pharmacological interventions on LBP using network meta‐analysis. The use of network meta‐analysis would facilitate reaching conclusions about which treatments or combinations of treatments are most effective and fill the methodological gaps of traditional systematic reviews that use pairwise analyses in the field of LBP. With the growing number of network meta‐analyses, future overviews might be more informative regarding the effectiveness of non‐pharmacological interventions for LBP (Efthimiou 2016). However, network meta‐analysis is challenging, and most still have low methodological quality (Gao 2019).
This overview shows that after more than four decades with 640 unique trials that randomised more than 90,000 participants, the overall effect of existing non‐pharmacological interventions for LBP is probably small. Therefore, research should investigate innovative and cost‐effective treatment strategies and process evaluations of existing treatments to optimise the effects of non‐pharmacological interventions for LBP.
Finally, more research is necessary to identify the people who will benefit from a particular non‐pharmacological and non‐surgical intervention. Individualising care for people with LBP may maximise treatment outcomes. Currently, only a few RCTs have enough power to investigate subgroup effects (Edwards 2022). More efficient interventions and recruitment strategies or innovative research methods may increase the number of participants in future trials and potentially provide reliable information about the effect of non‐pharmacological interventions in subgroups of people with LBP.
History
Protocol first published: Issue 8, 2021
Acknowledgements
We thank Prof Roger Chou and Dr Bruno Saragiotto for their comments on this overview.
Appendices
Appendix 1. Search strategy
Search strategy: The Cochrane Library (up to 15 April 2023)
#1 MeSH descriptor back pain explode all trees #2 MeSH descriptor pain explode all trees #3 (back or spine or spinal) adj2 pain #4 lumbar* or lumbo*
#5 backache* or back ache*
#6 (#1 OR #2 OR #3 OR #4 OR #5)
[Limited to Cochrane reviews]
Appendix 2. AMSTAR‐2 assessment criteria
AMSTAR‐2 is a 16‐item critical appraisal tool to assist in identifying high‐confidence systematic reviews. There is no summary score, but an overall rating based on weaknesses across seven critical domains*.
Did the research questions and inclusion criteria for the review include the components of PICO?
Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review and did the report justify any significant deviations from the protocol? *
Did the review authors explain their selection of the study designs for inclusion in the review?
Did the review authors use a comprehensive literature search strategy? *
Did the review authors perform study selection in duplicate?
Did the review authors perform data extraction in duplicate?
Did the review authors provide a list of excluded studies and justify the exclusions? *
Did the review authors describe the included studies in adequate detail?
Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? *
Did the review authors report on the sources of funding for the studies included in the review?
If meta‐analysis was performed, did the review authors use appropriate methods for statistical combination of results? *
If meta‐analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta‐analysis or other evidence synthesis?
Did the review authors account for RoB in individual studies when interpreting/discussing the results of the review? *
Did the review authors provide a satisfactory explanation for, and discussion of, any heterogeneity observed in the results of the review?
If they performed quantitative synthesis, did the review authors carry out adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? *
Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review?
Ratings of the overall confidence of the systematic reviews are as follows:
High: zero or one non‐critical weakness: the systematic review provides an accurate and comprehensive summary of the results of the available studies that address the question of interest.
Moderate: more than one non‐critical weakness*: the systematic review has more than one weakness, but no critical flaws. It may provide an accurate summary of the results of the available studies that were included in the review.
Low: one critical flaw with or without non‐critical weaknesses: the review has a critical flaw and may not provide an accurate and comprehensive summary of the available studies that address the question of interest.
Critically low: more than one critical flaw with or without non‐critical weaknesses: the review has more than one critical flaw and should not be relied on to provide an accurate and comprehensive summary of the available studies.
Appendix 3. Effects of non‐pharmacological and non‐surgical interventions compared to other interventions
Acupuncture and dry needling
Two Cochrane reviews investigated the effects of acupuncture and dry needling on LBP (Furlan 2005, Mu 2020). One review, judged at low confidence, included 12 trials with 1177 participants with acute and subacute LBP (Furlan 2005). Another review, judged at high confidence, included 33 trials with 8270 participants with chronic LBP (Mu 2020).
Major outcomes
Pain intensity
Acute and subacute LBP
Furlan 2005 narratively reported:
There may be no difference between acupuncture and naproxen (500 mg daily for ten days) on pain intensity in the short term (one trial judged as high quality, 57 participants, rated as low certainty evidence).
It is uncertain whether there is a difference between a single session of dry‐needling Yaotongxue (Extra 29, EX‐UE 7) and a single session of bilateral dry‐needling of SI3 on pain intensity in the short term (one trial judged as low quality, 150 participants, rated as very low certainty evidence).
Chronic LBP
Mu 2020 reported:
There may be no difference between acupuncture and TENS on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐9.5 favouring acupuncture, 95% CI ‐19.5 to 0.5, I2=0%, three trials, 116 participants, low certainty evidence).
Compared to baclofen, acupuncture may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐13.6, 95% CI ‐22.6 to ‐4.6, one trial, 40 participants, low certainty evidence).
Compared to dry needling, acupuncture may provide little to no difference in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐1.2, 95% CI ‐1.3 to ‐1.1, one trial, 60 participants, low certainty evidence), but its beneficial effect is uncertain in the intermediate term (0 to 100 pain intensity scale, MD 1.4, 95% CI 0.6 to 2.2, one trial, 60 participants, very low certainty).
It is uncertain whether there is a difference between acupuncture and pulsed radiofrequency therapy on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3.0 favouring pulsed radiofrequency therapy, 95% CI ‐19.7 to 13.7, one trial, 65 participants, very low certainty evidence).
Compared to a fast‐in and fast‐out acupuncture technique, rotation acupuncture of needles may provide little to no difference in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐1.0, CI ‐1.9 to ‐0.1, one trial, 31 participants, low certainty evidence).
Compared to 0.25 mm diameter needles acupuncture, acupuncture with 0.9 mm diameter needles may provide little to no difference in pain intensity in the short term (0 to 100 pain intensity scale, MD 2, 95% CI 0.6 to 3.4, one trial, 31 participants, low certainty evidence).
There may be no difference between standardised acupuncture and individualised acupuncture in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐1.3, 95% CI ‐9.5 to 6.9, one trial, 144 participants, low certainty evidence) and in the intermediate term (MD ‐3.2, 95% CI ‐11.8 to 5.4, one trial, 139 participants, low certainty evidence).
When added to exercise, it is uncertain whether acupuncture reduces pain intensity in the short term (0 to 100 pain intensity scale, MD 0.49 favouring exercise without acupuncture, 95% CI ‐2.4 to 3.4, one trial, 45 participants, very low certainty evidence).
When added to botulinum toxin, it is uncertain whether acupuncture reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3.8 favouring acupuncture with botulinum toxin, 95% CI ‐9.2 to 1.6, one trial, 43 participants, very low certainty evidence).
Mixed LBP
Furlan 2015 narratively reported:
It is uncertain whether there is a difference between dry needling at distal trigger point and dry needling at proximal trigger point in pain intensity in the short term (immediately after) (one trial judged as low quality, 20 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between acupuncture with needles inserted and removed immediately and acupuncture with needles inserted and removed after 10 minutes in pain intensity in the intermediate term (one trial judged as low quality, 20 participants, rated as very low certainty evidence).
Function
Chronic LBP
Mu 2020 reported:
It is uncertain whether there is a difference between acupuncture and TENS on pain intensity in the short term (0 to 24 RMDQ, MD ‐0.49 favouring acupuncture, 95% CI ‐2.61 to 1.63, two trials, 61 participants, very low certainty evidence).
It is uncertain whether there is a difference between acupuncture and pulsed radiofrequency therapy on pain intensity in the short term (0 to 50 ODI, MD 3.4 favouring pulsed radiofrequency therapy, 95% CI ‐11.82 to 18.62, one trial, 65 participants, very low certainty evidence).
Compared to baclofen, acupuncture may provide a small improvement in function in the short term (on a 0 to 24 RMDQ, MD ‐2.3, 95% CI ‐4.55 to ‐0.05, one trial, 40 participants, low certainty evidence).
Compared to dry needling on trigger points, acupuncture may provide little to no difference in function in the short term (0 to 24 RMDQ, MD ‐1.08, 95% CI ‐1.34 to ‐0.82, one trial, 60 participants, low certainty evidence). However, in the intermediate term dry‐needling on trigger points may provide a small improvement in function compared to acupuncture (0 to 50 ODI, MD 5.02, 95% CI 2.05 to 7.99, one trial, 60 participants, low certainty evidence).
Compared to a fast‐in and fast‐out technique acupuncture, rotation acupuncture of needles may provide a small improvement in function in the short term (0 to 50 ODI, MD ‐3.5, 95% CI ‐4.82 to ‐2.18, one trial, 60 participants, low certainty).
There may be no difference between individualised acupuncture and standardised acupuncture in function in the short term (0 to 24 RMDQ, MD 1.7 favouring individualised acupuncture, 95% CI ‐4.34 to 7.74, one trial, 144 participants, low certainty evidence), and probably there is no difference in the intermediate term (0 to 24 RMDQ, SMD 0.03 favouring individualised acupuncture, 95% CI ‐0.16 to 0.22, one trial, 427 participants, moderate certainty evidence).
There may be no difference between individualised acupuncture and standardised acupuncture on global assessment of therapy effectiveness in the short term (0 to 70 Pain Disability Index, MD 0.3, 95% CI ‐3.96 to 4.56, one trial, 144 participants, low certainty evidence) or in the intermediate term (0 to 70 Pain Disability Index, MD ‐0.7, 95% CI ‐8.49 to 7.09, one trial, 139 participants, low certainty evidence).
When added to exercise, it is uncertain whether acupuncture improve function in the short term (0 to 50 OD, MD 3.94 favouring exercise, 95% CI ‐7.84 to 15.72, one trial, 45 participants, very low certainty evidence) or in the intermediate term (0 to 50 ODI, MD ‐1.42 favouring acupuncture added to exercise, 95% CI ‐14.03 to 11.19, one trial, 44 participants, very low certainty evidence).
When added to botulinum toxin, acupuncture may provide a medium improvement in function in the short term (0 to 50 ODI, MD ‐11.91 favouring acupuncture added to botulinum toxin, 95%CI ‐19.02 to ‐4.8, one trial, 43 participants, low certainty evidence).
Mixed LBP
Furlan 2015 narratively reported that it is uncertain whether there is a difference between dry needling at distal trigger point and dry needling at proximal trigger point on function in the short term (immediately after) (one trial judged as low quality, 20 participants, rated as very low certainty evidence).
Safety
Chronic LBP
Mu 2020 reported:
It is uncertain whether there is a difference between acupuncture and TENS in the risk experiencing adverse events in the short term (RR 1.22 favouring TENS, 95% CI 0.48 to 3.13, two trials, 76 participants, very low certainty evidence).
It is uncertain whether there is a difference between acupuncture and dry needling in the risk of experiencing adverse events in the short term (RR 9.71 favouring dry needling, 95% CI 0.56 to 168.32, one trail, 62 participants, very low certainty evidence).
There may be no difference between standardised acupuncture and individualised acupuncture in the risk of experiencing adverse events in the short term (RR 0.97 favouring standardised acupuncture, 95% CI ‐0.32 to 2.93, one trial, 299 participants, low certainty evidence).
When added to exercise, it is uncertain whether acupuncture increases the risk of experiencing adverse events in the short term (RR 11.48 favouring exercise, 95% CI 0.67 to 196.07, one trial, 45 participants, very low certainty evidence).
Minor outcomes
Participant ratings of improvement
Acute/sub‐acute LBP
Furlan 2005 reported:
It is uncertain whether there is a difference between acupuncture with moxibustion and Chinese herbal medicine on overall improvement in the long term (RR 1.17 favouring acupuncture with moxibustion, 95% CI 1.03 to 1.33, one trial judged as low quality, 100 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between dry needling added to injection with lidocaine and injection with lidocaine and steroids or cooling spray over the trigger point area and these interventions alone on global improvement in the short term (one trial for each comparison judged as low quality, range of 33 to 36 participants, rated as very low certainty evidence).
Chronic LBP
Furlan 2005 narratively reported that when added to a regimen of physiotherapy, it is uncertain whether dry needling at trigger points increases participant ratings of improvement in the short term (one trial judged as low quality, 56 participants, rated as very low certainty evidence).
Mu 2020 reported:
Compared to TENS, acupuncture may provide little to no difference on global assessment of therapy effectiveness in the short term (one trial with high risk of bias, 60 participants, low certainty evidence).
It is uncertain whether there is a difference between acupuncture and dry needling in global assessment of therapy effectiveness in the short term (one trial with high risk of bias, 60 participants, rated as very low certainty evidence), or in the intermediate term (one trial with high risk of bias, 60 participants, rated as very low certainty evidence).
When added to distant acupoints, it is uncertain whether local acupoints increase global assessment of therapy effectiveness in the short term (RR 0.92 favouring local and distant acupoints, 95% CI 0.77 to 1.10, I2=0%, two trials, 289 participants, very low certainty evidence) or in the intermediate term (RR 0.89 favouring local and distant acupoints, 95% CI 0.78 to 1.0, I2=0%, two trials, 289 participants, very low certainty evidence).
Compared to acupuncture with regular needles acupuncture, it is uncertain whether acupuncture with heated needles increases global assessment of therapy effectiveness in the short term (two trials with high risk of bias reporting mixed effects, 133 participants, rated as very low certainty evidence).
Quality of life
Chronic LBP
Mu 2020 reported:
It is uncertain whether there is a difference between acupuncture and pulsed radiofrequency therapy on quality of life in the short term (0 to 100 SF‐36 quality of life scale, MD 4.2 favouring acupuncture, 95% CI ‐1.9 to 10.3, one trial, 65 participants, very low certainty evidence).
It is uncertain whether there is a difference between acupuncture with 0.25 diameter needles and acupuncture with 0.9 mm diameter needles on quality of life in the short term (0 to 100 SF‐36 quality of life scale, MD ‐5.3 favouring 0.25 diameter needles, 95% CI ‐11.3 to 0.7, one trial, 31 participants, very low certainty evidence).
There may be no difference between standardised acupuncture and individualised acupuncture on mental health‐related quality of life in the short term (0 to 100 SF‐36 mental‐related quality of life scale, 0.5 favouring individualised acupuncture, 95% CI ‐3.1 to 4.1, one trial, 144 participants, low certainty evidence) or in the intermediate term (0 to 100 SF‐36 mental‐related quality of life scale, MD ‐1.4 favouring standardised acupuncture, 95% CI ‐4.6 to 1.8, one trial, 139 participants, low certainty evidence).
When added to exercise, it is uncertain whether acupuncture improves quality of life in the short term (0 to 100 EuroQol (EQ‐5D) quality of life scale, MD ‐0.05 favouring acupuncture added to exercise, 95% CI ‐0.3 to 0.2, one trial, 45 participants, very low certainty evidence) and the intermediate term (0 to 100 EQ‐5D quality of life scale, MD ‐0.1 acupuncture added to exercise, 95% CI ‐0.3 to 0.1, one trial, 44 participants, very low certainty evidence).
Advice and education
Four Cochrane reviews investigated the effects of advice or education on LBP (Engers 2008, Dahm 2010, Poquet 2016, Parreira 2017). Engers 2008, judged at high confidence, included 24 trials with 7139 participants and investigated the effects of individual patient education for low back pain. Dahm 2010, judged at low confidence, included eight trials with 1923 participants and compared the effects of advice to rest in bed with advice to stay active for acute LBP. The other two reviews investigated the effects of back school on acute and subacute LBP (judged at high confidence, four trials with 643 participants) (Poquet 2016) and chronic LBP (judged at low confidence, 31 trials with 4105 participants) (Parreira 2017).
Major outcomes
Pain intensity
Acute and subacute LBP
Advice: Dahm 2010 reported:
For acute LBP, there may be no difference between advice to stay active and advice to rest in bed in pain intensity in the short term (SMD ‐0.02 favouring advice to stay active, 95% CI 0.16 to ‐0.20, I2=92%, three trials, 480 participants, low certainty evidence). However, the review reported that a sensitivity analysis excluding one trial with high risk of bias and a different population (hospitalised combat trainees) demonstrated that advice to stay active probably provides a small reduction in pain intensity in the short term (SMD ‐0.22, 95% CI ‐0.02 to ‐0.41, two trials, 401 participants, rated as moderate certainty evidence).
For acute LBP, there may be no difference between advice to rest in bed and non‐educational interventions in pain intensity in the short term (two trials with low risk of bias, 472 participants, rated as low certainty evidence).
For acute LBP, it is uncertain whether there is a difference between advice to short rest in bed (2‐3 days) and advice to long rest in bed (7 days) in pain intensity in the short term (SMD 0.2 favouring advice to long rest in bed, 95% CI ‐0.15 to 0.55, one trial with high risk of bias, 47 participants, very low certainty evidence).
For acute LBP, there may be no difference between advice to stay active and other interventions (besides bed rest) in pain intensity (0‐100 pain intensity scale, MD 0, 95% CI ‐3.9 to 3.9, one trial, 103 participants, low certainty evidence).
Chronic LBP
Individual education: Engers 2008 reported:
It is uncertain whether there is a difference between individual patient education and physiotherapy (exercise, manual therapy, back school) in pain intensity in the short or intermediate term (one trial judged as low quality, 302 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between individual patient education and spinal stabilisation in pain intensity in the short or intermediate term (one trial judged as low quality, 302 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between individual patient education and yoga exercise in pain intensity in the intermediate term (one trial judged as low quality, 188 participants, rated as very low certainty evidence).
Back school: Parreira 2017 reported:
It is uncertain whether there is a difference between back school and passive physiotherapy (electrophysical modalities and heat therapy) in pain intensity in the short term (0 to 100 pain intensity scale, MD 2 favouring passive physiotherapy, 95% CI ‐9.5 to 13.4, I2=93%, three trials, 290 participants, very low certainty evidence) and the intermediate term (0 to 100 pain intensity scale, MD ‐16.9 favouring back school, 95% CI ‐66.6 to 32.8, I2=99%, three trials, 290 participants, very low certainty evidence).
It is uncertain whether there is a difference between back school and exercise in pain intensity the short term (0 to 100 pain intensity scale, MD ‐2.1 favouring back school, 95% CI ‐14.6 to 10.5, I2=84%, five trials, 416 participants, very low certainty evidence) or intermediate term (0 to 100 pain intensity scale, MD ‐4.5 favouring back school, 95% CI ‐19.4 to 10.5, I2=94%, four trials, 619 participants, very low certainty evidence).
Mixed LBP
Individual education: Engers 2008 narratively reported that when added to physician delivered individual patient education (booklet), it is uncertain whether individual patient education (booklet) reduces pain intensity in the short term (one trial judged as low quality, 68 participants, rated as very low certainty evidence).
Advice: Dahm 2010 reported that it is uncertain whether there is a difference between advice to rest and advice to stay active in pain intensity in the short term (SMD ‐0.1, 95% CI ‐0.94 to 0.74, one trial, 22 participants, very low certainty evidence).
Function
Acute and subacute LBP
Advice: Dahm 2010 reported:
For acute LBP, compared to advice to rest in bed, advice to stay active probably provides a small improvement in function in the short term (SMD ‐0.29, 95% CI ‐0.09 to ‐0.49, two trials, 400 participants, moderate certainty evidence).
For acute LBP, there may be no difference between advice to rest in bed and non‐educational interventions in function (e.g., exercise, manipulation, medicine) in the short term (two trials with low risk of bias, 472 participants, rated as low certainty evidence) and the intermediate term (two trials with low risk of bias, 472 participants, rated as low certainty evidence).
For acute LBP, compared to other treatments (exercise), advice to stay active may provide a small improvement in function in the short term (at four weeks) (0 to 100 function scale, ‐8.6, 95% CI ‐3.32 to 13.88, one trial, 103 participants, low certainty evidence), but there is no evidence of a difference at three months.
Individual education: Engers 2008 narratively reported that there may be no difference between individual patient education and non‐educational interventions (e.g., chiropractic manipulation, manual therapy, McKenzie therapy, immediate x‐ray, cognitive behavioural group therapy, electrophysical therapies) on pain intensity in the short term (two trials, high and low quality, 485 participants, rated as low certainty evidence).
Back school: Poquet 2016 reported:
It is uncertain whether there is a difference between back school and myofascial therapy in function in the short term (0‐24 RMDQ, MD ‐1.54 favouring back school, 95% CI ‐1.12 to 7.9, one trial, 63 participants, very low certainty evidence) and intermediate term (0‐24 RMDQ, MD ‐1.58, 95% CI ‐4.02 to 5.3, one trial, 61 participants, very low certainty evidence).
It is uncertain whether there is a difference between back school and manipulation in function in the short term (0‐24 RMDQ, MD ‐0.16 favouring back school, 95% CI ‐2.5 to 2.18, one trial, 59 participants, very low certainty evidence) and the intermediate term (0‐24 RMDQ, MD 0.19 favouring manipulation, 95% CI ‐2.3 to 2.68, one trial, 55 participants, very low certainty evidence).
It is uncertain whether there is a difference between back school and manipulation with myofascial therapy in function in the short term (0‐24 RMDQ, MD 0.53 favouring manipulation with myofascial therapy, 95% CI ‐1.6 to 2.66, one trial, 62 participants, very low certainty evidence) or the intermediate term (0‐24 RMDQ, MD ‐0.08 favouring back school, 95% CI ‐2.33 to 2.17, one trial, 62 participants, very low certainty evidence).
When added to back care programme (rest, analgesics, massage, cold/hot packs and electrophysical modalities), back school may provide little to no difference in function in the short (0 to 24 RMDQ, MD 1.78 95% CI 0.52 to 3.04, one trial, 156 participants, low certainty evidence), but there may be no difference in the intermediate term (0 to 24 RMDQ, MD 0.81 95% CI ‐0.44 to 2.06, one trial, 156 participants, low certainty of evidence).
Chronic LBP
Individual education: Engers 2008 narratively reported:
It is uncertain whether there is a difference between individual patient education and physiotherapy (exercise, manual therapy, back school) on function in the short and the intermediate term (one trial judged as low quality, 302 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between individual patient education and spinal stabilisation in the short term, but individual patient education may provide less improvement compared to spinal stabilisation in function in the intermediate term (one trial judged as low quality, 302 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between back school (6 sessions 60 minutes) and individual patient education (one trial judged as low quality, 188 participants) or yoga exercise (one trial with low quality, 101 participants, rated as very low certainty evidence) in function in the short or intermediate term.
It is uncertain whether there is a difference between individual patient education and spinal stabilisation in function in the short term or in the intermediate term (one trial judged as low quality, 101 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between individual patient education and exercise in function in the short term or in the intermediate term (one trial judged as low quality, 101 participants, rated as very low certainty evidence).
Back school: Parreira 2017 reported:
It is uncertain whether there is a difference between back school and passive physiotherapy (electrophysical modalities and heat therapy) in function in the short term (0 to 100 function scale, MD 2.6 favouring passive physiotherapy, 95% CI ‐15.8 to 21.0, I2=81%, two trials, 180 participants, very low certainty evidence) or in the intermediate term (0 to 100 function scale, MD 6.9 favouring passive physiotherapy, 95% CI ‐4.9 to 18.6, I2=74%, two trials, 190 participants, very low certainty evidence).
It is uncertain whether there is a difference between back school and exercise in function in the short term (0 to 100 function scale, MD ‐1.7 favouring back school, 95% CI ‐8.6 to 5.4, I2=85%, six trials, 471 participants, very low certainty evidence) or in the intermediate term (0 to 100 function scale, MD 1.6 favouring exercise, 95% CI ‐3.9 to 7.0, I2=87%, six trials, 766 participants, very low certainty evidence).
Mixed LBP
Individual education: Engers 2008 narratively reported:
Compared to massage, individual patient education (book and videotapes) may result in less improvement in function in the short and intermediate‐term (one trial judged as high quality, 262 participants, rated as low certainty evidence).
There may be no difference between individual patient education (book and videotapes) and acupuncture in function in the short and intermediate‐term (one trial judged as high quality, 262 participants, rated as low certainty evidence).
When added to physiotherapy, individual patient education (booklet) may not improve function in the short or intermediate term (one trial judged as high quality but with high risk of bias for randomisation and blinding of patients and providers, 975 participants, rated as low certainty evidence).
It is uncertain whether there is a difference between booklet and booklet with oral information and telephone feedback (one trial judged as high quality, 294 participants, rated as very low certainty evidence).
When added to exercise, it is uncertain whether a booklet (the Book Home) improves function in the short term (one trial judged as low quality, 311 participants, rated as very low certainty evidence).
Advice: Dahm 2010 reported that there may be no difference between advice to bed rest short time and advice to bed active long time in function in the short term (one trial judged as low quality, 189 participants).
Safety
Acute and subacute LBP
Back school: Poquet 2016 reported:
It is uncertain whether there is a difference between back school and myofascial therapy in the short term (RR 1.59 favouring myofascial therapy, 95% CI 0.48 to 5.3, one trial 99 participants, very low certainty evidence).
It is uncertain whether there is a difference between back school and manipulation (RR 1.02 favouring manipulation, 95% CI 0.35 to 2.94, one trial 97 participants, very low certainty evidence) or the combination of manipulation with myofascial therapy (RR 0.93 favouring back school, 95% CI 0.34 to 2.57, one trial 100 participants, very low certainty evidence) in the short term.
Minor outcomes
Participant ratings of improvement
Acute and subacute LBP
Individual education: Engers 2008 narratively reported that there may be no difference between individual patient education and non‐educational interventions (physiotherapy, x‐ray, interferential therapy) on global improvement (three trials judged as low quality, 454 participants, rated as low certainty evidence) in the short term. The review narratively reported that chiropractic manipulation or McKenzie exercise may increase global improvement in the short term compared to individual patient education (two trials judged as high quality, 203 participants, rated as low certainty evidence).
Chronic LBP
Individual education: Engers 2008 narratively reported that it is uncertain whether there is a difference between individual patient education and exercise on global improvement in the short or intermediate term (one trial judged as low quality, 101 participants, rated as very low certainty evidence).
Mixed LBP
Individual education: Engers 2008 narratively reported:
There may be no difference between individual patient education (booklet and videotapes) and massage or acupuncture on global improvement in the short or intermediate term (one trial judged as high quality, 262 participants, rated as low certainty evidence).
When added to physiotherapy, individual patient education (booklet) may result in less global improvement in the short or intermediate term (one trial judged as high quality, 975 participants, rated as low certainty evidence).
When added to verbal information, it is uncertain whether an education booklet increases global impression of improvement in the short term (one trial judged as low quality, 294 participants, rated as very low certainty evidence).
Workplace participation
Acute and subacute LBP
Advice: Dahm 2010 reported:
There may be no difference between advice to rest and advice to stay active in return to work and sick leave (two trials with low risk of bias, 472 participants, rated as low certainty evidence).
There may be no difference between advice to rest and other interventions in return to work in the short term (two trials with low risk of bias, 472 participants, rated as low certainty evidence).
There may be no difference between advice to stay active combined with other interventions (exercise, manipulation, medicine) and other interventions alone in sick leave (days of sick leave, MD 1.6 favouring stay active, 95% CI ‐0.25 to 3.45, one trial with low risk of bias, 107 participants, rated as low certainty evidence).
Compared to advice to rest, advice to stay active may provide a small reduction in the number of days of sick leave in the short term (number of days of sick leave 4.5, 95% CI 1.37 to 7.63, one trial with high risk of bias for blinding, 123 participants, rated as low certainty evidence).
Individual education: Engers 2008 narratively reported:
There may be no difference between individual patient education (booklet or 1.5 hours of education) and referral for x‐ray or work visit in return to work in the short term (two trials judged as high quality, 265 participants) and the intermediate term (one trial judged as high quality, 164 participants, rated as low certainty evidence).
Compared to cognitive behavioural therapy (6 sessions of 2 hours), individual patient education may provide less improvement in return to work (days of sick leave) in the intermediate term (one trial judged as high quality, 243 participants, rated as low certainty evidence).
Back school: Poquet 2016 reported:
It is uncertain whether there is a difference between back school and physiotherapy in the risk of experiencing sick leave <21 days (RR 1.37 favouring combined physiotherapy, 95% CI 1 to 1.87, one trial with high risk of bias, 116 participants, rated as very low certainty evidence) or >21 days (RR 0.64 favouring back school, 95% CI 0.41 to 1.01, one trial with high risk of bias, 116 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between short‐wave diathermy and back school to decrease the risk of experiencing sick leave >21 days in the short term (RR 0.53, 95% CI 0.35 to 0.80, one trial with high risk of bias, 121 participants, rated as very low certainty evidence).
When added to back care programme (rest, analgesics, massage, cold/hot packs and electrophysical modalities), it is uncertain whether back school improves workplace participation (work absence) in the short term (RR 0.76, 95% CI 0.32 to 1.84, one trial, 168 participants, very low certainty of evidence) or in the intermediate term (RR 0.94, 95% CI 0.76 to 1.16, one trial, 168 participants, very low certainty of evidence).
Chronic LBP
Individual education: Engers 2008 narratively reported that it is uncertain whether there is a difference between individual patient education and back school (6 sessions 60 minutes) on global improvement in the intermediate term (one trial judged as low quality, 188 participants, rated as very low certainty evidence).
Mixed LBP
Individual education: Engers 2008 narratively reported that when added to verbal information, it is uncertain whether an education booklet increases the likelihood of returning to work in the short or intermediate term (one trial judged as low quality, 294 participants, rated as very low certainty evidence).
Advice: Dahm 2010 reported that it is uncertain whether there is a difference between advice to rest for a short time and advice to rest for a long time in the short term (one trial judged as low quality, 189 participants, rated as very low certainty evidence). The review reported that a subgroup analysis suggested that two days of bed rest resulted in fewer days off work (mean: 3.1 days) than seven days of bed rest (mean 5.6 days) in patients who were employed (about one‐third of the sample).
Assistive materials
One Cochrane review, judged at moderate confidence, included eight trials with 1361 participants and investigated the effects of lumbar support on LBP (Van Duijvenbode 2008).
Pain intensity
Chronic LBP
Van Duijvenbode 2008 narratively reported:
When added to muscle strengthening program, it is uncertain whether lumbar support and muscle strengthening program reduces pain intensity in the short term or intermediate term (one trial judged as low quality, 63 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between flexible corset lumbar support and semi‐rigid corset lumbar support (one trial judged as low quality, 79 participants) or lumbar support without spine support (one trial judged as low quality, 19 participants, rated as very low certainty evidence) in pain intensity in the short term.
Mixed LBP
Van Duijvenbode 2008 narratively reported:
There may be no difference between lumbar support and other interventions (traction, manipulation and exercises, physiotherapy and manipulation, electrical therapy modality and manipulation) on pain intensity in the short term (three trials mostly judged as low quality, 954 participants, rated as low certainty evidence).
There may be no difference between traction added to physiotherapy and physiotherapy alone on pain intensity in the short term (0‐100 pain intensity scale, MD 0.0, 95% CI ‐7.6 to 7.6, one trial, 42 participants, low certainty evidence) or the intermediate term (0‐100 pain intensity scale, MD 5.0, 95% CI ‐5.7 to 15.7, one trial, 39 participants, low certainty evidence).
Function
Chronic LBP
Van Duijvenbode 2008 narratively reported:
When added to a muscle strengthening program, it is uncertain whether lumbar support improves function in the short term and intermediate term (one trial judged as low quality, 63 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between flexible corset lumbar support and semi‐rigid corset lumbar support (one trial judged as low quality, 79 participants, rated as very low certainty evidence) or lumbar support without spine support (one trial judged as low quality, 19 participants, rated as very low certainty evidence) in function in the short term.
Mixed LBP
Van Duijvenbode 2008 narratively reported that it is uncertain whether there is a difference between lumbar support and other interventions (electrical therapy modality and manipulation) in function in the short term (one trial judged as low quality, 164 participants, rated as very low certainty evidence).
Minor outcomes
Participant ratings of improvement
Mixed LBP
Van Duijvenbode 2008 narratively reported that there may be no difference between lumbar support and other interventions (traction, manipulation and exercises, physiotherapy and manipulation, electrical therapy modality and manipulation) on global impression of improvement in the short term (two trials judged as low quality, 790 participants, rated as low certainty evidence) or in the intermediate term (two trials judged as low quality, 790 participants, rated as low certainty evidence).
Workplace participation
Mixed LBP
Van Duijvenbode 2008 narratively reported that there may be no difference between lumbar support and other interventions (traction, manipulation and exercises) on return to work in the short term (one trial judged as low quality, 334 participants, rated as low certainty evidence).
Electrophysical modalities
Three Cochrane reviews investigated the effects of electrophysical modalities on LBP (Khadilkar 2008, Yousefi‐Nooraie 2008, Ebadi 2020). Khadilkar 2008, judged at high confidence, included four trials with 585 participants and investigated the effects of TENS on LBP. Yousefi‐Nooraie 2008, judged at low confidence, included seven trials with 384 participants and investigated the effects of low‐level laser therapy for low back pain. Ebadi 2020, judged at low confidence, included 10 trials with 1025 participants and investigated the effects of therapeutic ultrasound for chronic LBP.
Major outcomes
Pain intensity
Chronic pain
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported:
When added to exercise and sham low‐level laser therapy, low‐level laser therapy may not reduce pain intensity in the short‐term (post‐treatment) (0 to 100 pain intensity scale, MD ‐6.4 favouring laser therapy, 95% CI ‐15.7 to 2.9, I2=0%, two trials, 61 participants, rated as low certainty).
When added to exercise, it is uncertain whether low‐level laser therapy reduces pain intensity in the short term (one trial judged as low quality, 75 participants, rated as very low certainty evidence).
Therapeutic ultrasound: Ebadi 2020 reported:
It is uncertain whether there is a difference between therapeutic ultrasound and electrical stimulation in the short term (0 to 100 pain intensity scale, MD 2.5 favouring electrical stimulation, 95% CI ‐3.7 to 8.7, one trial, 39 participants, very low certainty evidence).
It is uncertain whether there is a difference between therapeutic ultrasound and phonophoresis (0 to 100 pain intensity scale, MD 0.1 favouring ultrasound, 95% CI ‐0.7 to 0.9, one trial, 40 participants, very low certainty evidence).
It is uncertain whether there is a difference between therapeutic ultrasound and laser therapy in the short term (0 to 100 pain intensity scale, MD 0.4, 95% CI 0.2 to 8.7, one trial, 150 participants, very low certainty evidence).
Compared to osteopathic manual intervention, therapeutic ultrasound probably reduces pain intensity (participants reporting at least 30% of pain reduction) in the short term (RR 0.75, 95% CI 0.59 to 0.95, moderate certainty evidence).
Compared to spinal manipulation, therapeutic ultrasound may provide less reduction in pain intensity (medium effect) in the short term (0 to 100 pain intensity scale, MD ‐16.5, 95% CI ‐27.6 to ‐5.5, one trial, 112 participants, low certainty evidence), but there may be no difference in the intermediate term (0 to 100 pain intensity scale, MD ‐15.1, 95% CI ‐32.8 to 2.6, one trial, 73 participants, low certainty evidence).
When added to exercise, it is uncertain whether therapeutic ultrasound reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐21.1, 95% CI ‐27.6 to ‐14.5, I2=34%, two trials, 70 participants, very low certainty evidence).
Function
Chronic pain
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported:
When added to exercise and sham low‐level laser therapy, low‐level laser therapy may not improve function in the short‐term (post‐treatment) (SMD ‐0.05 favouring laser therapy, 95% CI ‐0.56 to 0.45, I2=0%, two trails, 61 participants, rated as low certainty evidence).
When added to exercise, low‐level laser therapy may not improve function in the short term (one trial judged as low quality, 75 participants, rated as low certainty evidence).
Therapeutic ultrasound: Ebadi 2020 reported:
It is uncertain whether there is a difference between therapeutic ultrasound and electrical stimulation in function in the short term (0 to 100 function scale, MD 1.9 favouring electrical stimulation, 95% CI 0.0 to 3.8, one trial, 39 participants, very low certainty evidence).
It is uncertain whether there is a difference between therapeutic ultrasound and phonophoresis in function in the short term (0 to 100 function scale, MD 1.3 favouring phonophoresis, 95% CI ‐0.5 to 3.1, one trial, 40 participants, very low certainty evidence).
Compared to spinal manipulation, therapeutic ultrasound may provide less improvement in function (small effect) in the short term (0 to 100 function scale, MD ‐7.8, 95% CI ‐13.2 to ‐2.4, one trial, 112 participants, low certainty evidence), but there may be no difference in the intermediate term (0 to 100 pain intensity scale, MD ‐5.2, 95% CI ‐13.1 to 2.7, one trial, 73 participants, low certainty evidence).
When added to exercise, it is uncertain whether therapeutic ultrasound improves function in the short term (0 to 100 function scale, MD ‐0.4, 95% CI ‐3.1 to 2.3, I2=0%, two trials, 79 participants, very low certainty evidence).
Safety
Chronic pain
Low‐level laser therapy: Yousefi‐Nooraie 2008 reported that when added to exercise with sham low‐level laser therapy, low‐level laser therapy may not increase the risk of experiencing adverse effects in the short term (one two judged as high quality, 78 participants, low certainty evidence).
Minor outcomes
Quality of life
Chronic pain
Therapeutic ultrasound: Ebadi 2020 reported:
It is uncertain whether there is a difference between therapeutic ultrasound and phonophoresis in quality of life in the short term (0 to 100 quality of life scale, MD ‐10.8 favouring ultrasound, 95% CI ‐21.8 to 0.3, one trial, 40 participants, very low certainty).
When added to exercise, it is uncertain whether therapeutic ultrasound improves the quality of life in the short term (0 to 100 quality of life scale, MD ‐2.5 favouring phonophoresis, 95% CI ‐9.5 to 4.5, I2=0%, two trials, 79 participants, very low certainty).
Exercise therapies
Seven Cochrane reviews investigated the effects of exercise on LBP (Almeida 2023; Hayden 2021a; Macedo 2016; Saragiotto 2016; Wieland 2022; Yamato 2015; Hayden 2005). Hayden 2005, judged at high confidence, included 17 trials with 2357 participants and investigated the effects of all types of exercises in acute and subacute participants. Hayden 2021a, judged at high confidence, included 249 trials with 24486 participants and investigated the effects of all types of exercises in participants with chronic LBP. Yamato 2015, judged at high confidence, included 10 trials with 510 participants and investigated the effects of Pilates exercise on LBP. Macedo 2016 judged at low confidence, included three trials with 197 participants and investigated the effects of motor control exercise for acute LBP. Almeida 2023, judged at high confidence, included five trials with 538 participants and investigated the effects of McKenzie exercise for acute and subacute participants. Saragiotto 2016, judged as high confidence, included 29 trials with 2431 investigating the effects of motor control exercise for chronic LBP. Wieland 2022, judged as high confidence, included 21 trials with 2223 participants and investigated the effects of Yoga exercise for chronic LBP.
Major outcomes
Pain intensity
Acute/subacute LBP
Exercises: Hayden 2005 reported:
For acute LBP, there may be no difference between exercise and other conservative treatments in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐0.3 favouring exercise, 95% CI ‐0.7 to 0.1, I2=0, seven trials, most judged as low quality, 606 participants, rated as moderate certainty evidence), in the intermediate term (0 to 100 pain intensity scale, MD ‐0.3 favouring exercise, 95% CI ‐7.8 to 7.2, I2=0, two trials judged as low quality, 195 participants, rated as low certainty evidence) or in the long term (0 to 100 pain intensity scale, MD ‐0.8 favouring exercise, 95% CI ‐6.4 to 4.8, I2=0, two trials judged as low quality, 195 participants, rated as low certainty evidence).
For subacute LBP, there may be no difference between exercise and other conservative treatments in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐0.7 favouring exercise, 95% CI ‐2.8 to 1.4, I2=0, three trials judged as low quality, 350 participants, rated as low certainty evidence), the intermediate term (0 to 100 pain intensity scale, MD ‐1.5 favouring exercise, 95% CI ‐7.4 to 4.3, I2=32, three trials judged as low quality, 385 participants, rated as low certainty evidence) or the long term (0 to 100 pain intensity scale, MD ‐2.0 favouring exercise, 95% CI ‐5.2 to 1.2, one trial judged as low quality, 187 participants, rated as low certainty evidence).
Motor control: Macedo 2016 reported:
For acute/subacute LBP, there is probably no difference between motor control exercise and other exercises in pain intensity in the short term (0 to 100 pain intensity scale, MD 5.7 favouring other exercises, 95% CI ‐3.3 to 14.8, I2=0, two trials, 89 participants, moderate certainty evidence) or in the intermediate term (0 to 100 pain intensity scale, MD ‐1.2 favouring other exercises, 95% CI ‐18.2 to 15.8, one trial, 33 participants, low certainty evidence).
For acute/subacute LBP, there may be no difference between motor control exercise and spinal manipulation in pain intensity in the short term (0 to 100 pain intensity scale, MD 9.0 favouring spine manipulation, 95% CI ‐1.6 to 19.5, one trial, 58 participants, low certainty evidence).
McKenzie: Almeida 2023 reported:
For acute/subacute LBP, there may be no difference between McKenzie exercise and manual therapy in pain intensity in the short term (0 to 100 pain intensity scale MD ‐8.7 favouring McKenzie, 95% CI ‐27.4 to 10, I2=83%, three trials, 298 participants, low certainty evidence), but may result in a small increase (worse) in pain intensity (0 to 100 pain intensity scale MD 7, 95% CI 0.7 to 13, one trial, 235 participants, low certainty evidence).
Chronic LBP
Exercises: Hayden 2021a reported:
Compared to other conservative treatments, exercise may provide a small reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐9.1, 95% CI ‐12.6 to ‐5.6, I2=93%, 64 trials, 6295 participants, low certainty evidence), in the intermediate term (0 to 100 pain intensity scale, MD ‐8.1, 95% CI ‐12.3 to ‐3.9, I2=90%, 38 trials, 4078 participants, rated as low certainty evidence), and in the intermediate to long term (closest to 12 months follow‐up) (0 to 100 pain intensity scale, MD ‐5.2, 95% CI ‐11.3 to ‐0.8, I2=92%, 19 trials, 2478 participants, rated as low certainty evidence. The conservative treatments involved education [32%], multimodal non‐exercise physical therapy [30%], manual therapy [15%], electrical therapy [6%], and psychotherapy [6%]).
Compared to mixed physical therapies, exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.4, 95% CI ‐15.2 to ‐5.3, I2=66%, 21 trial, 1869 participants, rated as low certainty evidence).
Compared to education, exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐12.1, 95% CI ‐19.4 to ‐5, I2=94%, 26 trials, 2039 participants, rated as low certainty), but its beneficial effect is uncertain when compared to only back school education (0 to 100 pain intensity scale, MD ‐14, 95% CI ‐22.7 to ‐5.6, I2=66%, three trials, 241 participants, rated as very low certainty evidence).
Compared to non‐exercise physiotherapy, exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.4, 95% CI ‐15.2 to ‐5.6, I2=89%, 21 trials, 1869 participants, rated as low certainty evidence).
Compared to anti‐inflammatories, exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐14.8, 95% CI ‐24.4 to ‐5.2, two trials, I2=0%, 112 participants, rated as low certainty evidence).
Compared to psychological therapy, exercise may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐18.7, 95% CI ‐30.8 to ‐6.6, I2=23%, three trials, 134 participants, rated as low certainty evidence). However, it is uncertain whether there is a difference between exercise and relaxation in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3.5 favouring exercise, 95% CI ‐13.1 to 6.1, one trial, 68 participants, rated as very low certainty evidence).
There may be no difference between exercise and manual therapy in pain intensity in the short term (0 to 100 pain intensity scale, MD 1 favouring manual therapy, 95% CI ‐3.1 to 5.1, I2=73%, 12 trials, 1632 participants, rated as low certainty evidence).
There may be no difference between exercise and electrical therapy in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.3, 95% CI ‐23.9 to 3.2, I2=91%, three trials, 200 participants, rated as very low certainty evidence).
Motor control: Saragiotto 2016 reported:
Compared to other exercises, motor control exercise may provide a small reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐7.5, 95% CI ‐10.5 to ‐4.5, I2=42%, 13 trials, 872 participants, low certainty evidence), but there is no difference in the intermediate term (0 to 100 pain intensity scale, MD ‐3 favouring motor control, 95% CI ‐7 to 1, I2=0%, six trials, 588 participants, high certainty evidence) or in the long term (0 to 100 pain intensity scale, MD ‐2.7 favouring motor control, 95% CI ‐7 to 1.5, I2=0%, six trials, 643 participants, high certainty evidence).
Compared to exercise combined with electrophysical modalities, motor control exercise may provide a large reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐30.2, 95% CI ‐35.3 to ‐25, I2=0%, two trials, 68 participants, low certainty evidence) and the intermediate term (0 to 100 pain intensity scale, MD ‐19.4, 95% CI ‐36.8 to ‐1.9, I2=94%, two trials, 68 participants, very low certainty evidence).
There probably be no difference between exercise and manual therapy in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐4.4 favouring motor control, 95% CI ‐9.5 to 0.8, I2=0%, three trials, 284 participants, moderate certainty evidence), in the intermediate term (0 to 100 pain intensity scale, MD ‐7.1 favouring motor control, 95% CI ‐14.2 to 0.1, I2=0%, four trials, 485 participants, moderate certainty evidence), or in the long term (0 to 100 pain intensity scale, MD ‐3.7 favouring motor control, 95% CI ‐9.3 to 1.9, I2=0%, four trials, 406 participants, moderate certainty evidence).
It is uncertain whether there is a difference between exercise and telerehabilitation in pain intensity in the intermediate term (0 to 100 pain intensity scale, MD ‐10 favouring motor control, 95% CI ‐32.4 to 12.4, one trial, 48 participants, very low certainty evidence).
Pilates: Yamato 2015 narratively reported that compared to other exercises, Pilates may provide a small reduction in pain intensity in the short term (three trials, 166 participants, low certainty evidence). The review reported that two trials demonstrated a significant effect in favour of Pilates (0 to 100 pain intensity scale, point estimate ranged from ‐6.6 to ‐20), and one trial did not find a significant difference (0 to 100 pain intensity scale, point estimate of ‐5, 95% CI ‐11.9 to 3.9). For the intermediate term, the review reported conflicting results (two trials, 151 participants, low certainty evidence).
Yoga: Wieland 2022 reported:
Compared to other exercises, it is uncertain whether Yoga exercise reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐12.5, 95% CI ‐18.3 to ‐6.6, I2=36%, three trials, 201 participants, very low certainty evidence), in the intermediate (0 to 100 pain intensity scale, MD ‐6.4 favouring Yoga exercise, 95% CI ‐21.6 to 8.8, I2=93%, three trials, 331 participants, very low certainty evidence) or in the long term (0 to 100 pain intensity scale, MD 3.0 favouring other exercises, 95% CI ‐4.3 to 10.2, one trial, 199 participants, very low certainty evidence).
Compared to other intensive exercises, intensive Yoga may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, MD ‐14.5, 95% CI ‐22.9 to ‐6.1, one trial, 80 participants, low certainty evidence).
It is uncertain whether there is a difference between intensive Yoga and Qigong exercise in pain intensity in short to intermediate term (0 to 100 pain intensity scale, MD ‐1.5, 95% CI ‐5.0 to 8.8, one trial, 111 participants, very low certainty evidence) or long term (0 to 100 pain intensity scale, MD 7.9, 95% CI 0 to 15.8, one trial, 107 participants, very low certainty evidence).
When added to another exercise, it is uncertain whether Yoga exercise reduces pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3.2, 95% CI ‐13.8 to 7.3, one trial, 24 participants, very low certainty evidence).
Function
Acute/subacute LBP
Exercises: Hayden 2005 reported:
For acute LBP, there may be no difference between exercise and other conservative treatments in function in the short term (0 to 100 pain intensity scale, MD ‐1.3 favouring exercise, 95% CI ‐5.5 to 2.8, I2=79, six trials, most judged as low quality, 684 participants, rated as low certainty evidence), in the intermediate term (0 to 100 pain intensity scale, MD ‐0.4 favouring exercise, 95% CI ‐4.3 to 3.5, I2=0, two trials judged as low quality, 195 participants, rated as low certainty evidence) or in the long term (0 to 100 pain intensity scale, MD ‐0.5 favouring exercise, 95% CI ‐4.1 to 3.1, I2=0, two trials judged as low quality, 195 participants, rated as low certainty evidence).
For subacute LBP, there may be no difference between exercise and other conservative treatments in the short term (0 to 100 function scale, MD 1.3 favouring other treatments, 95% CI ‐1.9 to 4.4, I2=14%, two trials, 321 participants, rated as low certainty evidence), in the intermediate term (0 to 100 function scale, MD ‐1.1 favouring exercise, 95% CI ‐6.9 to 4.7, three trials, 385 participants, rated as low certainty evidence) or in the long term (0 to 100 function scale, MD ‐1.4 favouring exercise, 95% CI ‐5.8 to 3.1, one trial, 187 participants, rated as low certainty evidence).
Motor control: Macedo 2016 reported:
There may be no difference between motor control exercise and other exercises in function in the short term (0 to 100 function scale, MD ‐0.8 favouring motor control, 95% CI ‐8.7 to 7.0, I2=15, two trials, 116 participants, moderate certainty evidence), in intermediate‐term (0 to 100 function scale, MD ‐6.7 favouring motor control, 95% CI ‐22.8 to 9.4, one trial, 33 participants, low certainty evidence) or in the long term (0 to 100 function scale, MD 5.7 favouring other exercises, 95% CI ‐1.4 to 12.7, one trial, 83 participants, low certainty evidence).
There may be no difference between motor control exercise and spinal manipulation in function in the short term (0 to 100 function scale, MD 4 favouring spine manipulation, 95% CI ‐3.4 to 11.4, one trial, 95 participants, low certainty evidence) or the intermediate term (0 to 100 function scale, MD 3.7 favouring spine manipulation, 95% CI ‐4.1 to 11.5, one trial, 95 participants, low certainty evidence).
McKenzie: Almeida 2023 reported:
For acute/subacute LBP, there may be no difference between McKenzie exercise and manual therapy in function in the short term (0 to 100 function scale MD ‐5 favouring McKenzie, 95% CI ‐15 to 5, I2=63%, three trials, 298 participants, low certainty evidence), or in the intermediate term (0 to 100 function scale MD 4.3 favouring manual therapy, 95% CI ‐0.7 to 9.3, one trial, 235 participants, low certainty evidence).
For acute/subacute LBP, there may be no difference between McKenzie exercise and other interventions (massage plus health care advice) in function for acute/subacute LBP in the short term (0 to 100 function scale MD ‐4favouring McKenzie, 95% CI ‐15.4 to 23.4, one trial, 30 participants, very low certainty evidence), or in the intermediate term (0 to 100 function scale favouring massage plus health care advice MD 10, 95% CI ‐8.9 to 28.9, one trial, 25 participants, very low certainty evidence)
Chronic LBP
Exercise: Hayden 2021a reported:
Compared to other conservative treatments, exercise probably provides little to a small improvement in function in the short term (0 to 100 function scale, MD ‐4.1, 95% CI ‐6.0 to ‐2.2, I2=82%, 52 trials, 6004 participants, moderate certainty evidence) and the intermediate term (0 to 100 function scale, MD ‐2.2, 95% CI ‐4.5 to ‐0.1, I2=77%, 29 trials, 3531 participants). However, there may be no difference between exercise and other conservative treatments in function in the long term (term (0 to 100 function scale, MD 0.3 favouring other conservative treatments, 95% CI ‐2.2 to 2.7, I2=71%, 15 trials, 2816 participants, rated as low certainty evidence).
There may be no difference between exercise and mixed physical therapies in function in the short term (0 to 100 pain intensity scale, MD ‐4.3 favouring exercises, 95% CI ‐9 to 0.4, I2=89%, 13 trials, 1074 participants, rated as low certainty evidence).
Compared to education, exercise may provide a small improvement in function in the short term (0 to 100 function scale, MD ‐8, 95% CI ‐10.8 to ‐5.3, I2=78%, 19 trials, 2598 participants). However, there may be no difference between exercise and back school education in the short term (0 to 100 function scale, MD ‐1.7, 95% CI ‐6.3 to 2.9, I2=46%, two trials, 209 participants, rated as low certainty evidence).
There may be no difference between exercise and non‐exercise physiotherapy in function in the short term (0 to 100 function scale, MD ‐4.3 favouring exercises, 95% CI ‐9 to 0.4, I2=89%, 12 trials, 1074 participants, rated as low certainty evidence).
Compared to anti‐inflammatories, exercise may provide a medium reduction in function in the short term (0 to 100 function scale, MD ‐13.7, 95% CI ‐21.9 to ‐5.5, two trials, 142 participants, rated as low certainty evidence).
It is uncertain whether there is a difference between exercise and psychological therapy (0 to 100 function scale, MD ‐7.5 favouring exercise, 95% CI ‐25.3 to 10.3, 1 trial, 28 participants, rated as very low certainty evidence) or relaxation (0 to 100 function scale, MD 4.6 favouring relaxation, 95% CI ‐4.7 to 13.9, 1 trial, 68 participants, rated as very low certainty evidence) in the short term.
There may be no difference between exercise and manual therapy in function in the short term (0 to 100 function scale, MD 1.5 favouring manual therapy, 95% CI ‐1.5 to 4.4, I2=64%, 12 trials, 1637 participants, rated as low certainty evidence).
It is uncertain whether there is a difference between exercise and electrical therapy in the short term (0 to 100 function scale, MD ‐3.8 favouring exercise, 95% CI ‐10.1 to 2.4, I2=71%, four trials, 248 participants, rated as very low certainty evidence).
Motor control: Saragiotto 2016 reported:
Compared to other exercises, motor control exercise may result in little to no difference in function in the short term (0 to 100 function scale, MD ‐4.8, 95% CI ‐7 to ‐2.7, I2=49%, 11 trials, 794 participants, low certainty evidence), but there is no difference between motor control exercise and other exercises in the intermediate term (0 to 100 function scale, MD ‐2.9 favouring motor control, 95% CI ‐6.9 to 1.2, I2=0%, six trials, 588 participants, high certainty evidence) or the long term (0 to 100 function scale, MD ‐0.7 favouring motor control, 95% CI ‐4.9 to 3.5, I2=0%, four trials, 570 participants, high certainty evidence).
Compared to a combination of exercise and electrophysical agents, it is uncertain whether motor control exercise improves function in the short term (0 to 100 function scale, MD ‐20.8, 95% CI ‐28.0 to ‐13.5, one trial, 38 participants, very low certainty evidence) or the intermediate term (0 to 100 function scale, MD ‐11.5, 95% CI ‐29.6 to ‐2.3, one trial, 38 participants, low certainty evidence).
There probably no difference between exercise and manual therapy in function in the short term (0 to 100 function scale, MD ‐2.8 favouring motor control, 95% CI ‐6.6 to 1.0, I2=0%, three trials, 282 participants, moderate certainty evidence), in the intermediate term (0 to 100 function scale, MD ‐3.3 favouring motor control, 95% CI ‐6.7 to 0.4, I2=0%, four trials, 485 participants, moderate certainty evidence), or in the long term (0 to 100 function scale, MD ‐3.4 favouring motor control, 95% CI ‐7.9 to 1.0, I2=0%, four trials, 406 participants, moderate certainty evidence).
It is uncertain whether there is a difference between exercise and telerehabilitation in function in the intermediate term (0 to 100 function scale, MD 12.5 favouring telerehabilitation, 95% CI ‐16.3 to 42.4, one trial, 48 participants, very low certainty evidence).
Pilates: Yamato 2015 reported that is probably no difference between Pilates and other exercises in function in the short term (0 to 100 function scale, MD ‐3.3 favouring Pilates, 95% CI ‐6.8 to 0.2, two trials, 149 participants, moderate certainty evidence), or in the intermediate term (0 to 100 function scale, MD ‐0.9 favouring Pilates, 95% CI ‐5.0 to 3.2, two trials, 151 participants, moderate certainty evidence).
Yoga: Wieland 2022 reported:
It is uncertain whether there is a difference between Yoga exercise and other exercises in function in the short term (SMD ‐0.04 favouring Yoga, 95% CI ‐0.32 to 0.23, I2=42%, four trials, 395 participants, very low certainty evidence), in the intermediate term (SMD ‐0.08 favouring Yoga exercise, 95% CI ‐0.40 to 0.23, I2=93%, three trials, 331 participants, low certainty evidence) or in the long term (SMD ‐0.02, 95% CI ‐0.29 to 0.26, one trial, 200 participants, very low certainty evidence).
Compared to other intensive exercises, intensive Yoga may result in a large improvement in function in the short term (0 to 100 function scale, SMD ‐1.3, 95% CI ‐1.7 to ‐0.8, one trial, 80 participants, low certainty evidence).
It is uncertain whether there is a difference between intensive Yoga and Qigong exercise in function in the short term (SMD ‐0.24 favouring Yoga, 95% CI ‐0.61 to 0.14, one trial, 111 participants, very low certainty evidence) and in intermediate‐term (SMD ‐0.21 favouring Yoga, 95% CI ‐0.59 to 0.17, one trial, 107 participants, very low certainty evidence).
It is uncertain whether there is a difference between Yoga added to another exercise and that exercise alone on function in the short term (SMD ‐0.6 favouring Yoga, 95% CI ‐1.4 to 0.2, one trial, 24 participants, very low certainty evidence).
Safety
Chronic LBP
Motor control: Saragiotto 2016 narratively reported:
There may be no difference between motor control exercise and other exercises in the risk of experiencing adverse effects in the short term (five trials with low risk of bias in the randomization and allocation domains but one with unclear allocation, 763 participants, rated as low certainty evidence). Three trials did not report any adverse events related to the interventions, and two reported similar adverse events between the interventions.
There may be no difference between motor control exercise and manual therapy in safety in the short term (two trials, one with low risk of bias in the randomisation and allocation and another with unclear allocation, 452 participants, rated as low certainty evidence).
It is uncertain whether there is a difference between motor control exercise and a combination of exercise with electrophysical modalities in safety in the short term (one trial with high risk of bias for allocation, 30 participants, rated as very low certainty evidence).
Pilates: Yamato 2015 narratively reported that it is uncertain whether there is a difference between Pilates and general exercises in safety in the short term. The review reported that two participants experienced shoulder pain and one knee pain in the Pilates group in one trial with low risk of bias for randomisation and allocation, 86 participants, rated as very low certainty evidence) while two participants experienced back spasms and two had worsened back pain in the other exercise group. The two participants with worsened back pain from the general exercise group ceased the exercise program because of their pain.
Yoga: Wieland 2022 reported that there is probably no difference between Yoga exercises and other exercises in the risk of experiencing adverse events in the long term (RR 0.93, 95% CI 0.56 to 1.53, I2=0%, five trials, 640 participants, rated as moderate certainty evidence).
Minor outcomes
Participant ratings of improvement
Chronic LBP
Motor control: Saragiotto 2016 reported:
There probably be no difference between motor control exercise and other exercises in participant rantings of improvement in the short term (‐5 to 5 participant rantings of improvement scale, MD 0.88 favouring motor control, 95% CI ‐0.88 to 2.63, I2=93%, two trials, 332 participants, moderate certainty evidence), in the intermediate term (‐5 to 5 participant rantings of improvement scale, MD 0.3 favouring motor control, 95% CI ‐0.22 to 0.82, I2=0%, two trials, 332 participants, moderate certainty evidence), or in the long term (‐5 to 5 participant rantings of improvement scale, MD 0.53 favouring motor control, 95% CI ‐0.03 to 1.08, I2=0%, two trials, 332 participants, moderate certainty evidence).
There may be no difference between motor control exercise and manual therapy in participant rantings of improvement in the short term (‐5 to 5 participant rantings of improvement scale, MD 0.5 favouring motor control, 95% CI ‐0.12 to 1.12, one trial, 160 participants, low certainty evidence), in the intermediate term (‐5 to 5 participant rantings of improvement scale, MD 0.2 favouring motor control, 95% CI ‐0.59 to 0.98, one trial, 160 participants, low certainty evidence), and in the long term (‐5 to 5 participant rantings of improvement scale, MD 0.6 favouring motor control, 95% CI ‐0.24 to 1.44, one trial, 160 participants, low certainty evidence).
Compared to the combination of exercise and electrophysical modalities, motor control exercise may provide a small increase in participant rantings of improvement in the short term (‐5 to 5 participant rantings of improvement scale, MD 1.85 favouring motor control, 95% CI 1.09 to 2.61, one trial, 38 participants, low certainty evidence), and in the intermediate term (‐5 to 5 participant rantings of improvement scale, MD 1.67 favouring motor control, 95% CI 0.89 to 2.45, one trial, 38 participants, low certainty evidence).
Quality of life
Chronic LBP
Motor control: Saragiotto 2016 reported:
There may be no difference between motor control exercise and other exercises in physical‐health‐related quality of life in the short term (0 to 100 physical‐health related quality of life scale, MD 0, 95% CI ‐3.8 to 3.8, one trial, 172 participants, low certainty evidence), or the intermediate term (0 to 100 physical‐health related quality of life scale, MD 1.4 favouring motor control, 95% CI ‐2.6 to 5.4, one trial, 172 participants, low certainty evidence), and there probably be no difference in the long term (0 to 100 physical‐health related quality of life, MD 0, 95% CI ‐3.1 to 3.3, one trial, 172 participants, moderate certainty evidence).
There may be no difference between motor control exercise and other exercises in mental‐health‐related quality of life in the short term (0 to 100 mental‐health related quality of life scale, MD 0.2 favouring motor control, 95% CI ‐3.3 to 3.7, one trial, 172 participants, low certainty evidence), or the intermediate term (0 to 100 mental‐health related quality of life scale, MD ‐2 favouring other exercises, 95% CI ‐5.3 to 1.3, one trial, 172 participants, low certainty evidence), and there probably be no difference in the long term (0 to 100 mental‐health related quality of life scale, MD ‐0.8 favouring other exercises, 95% CI ‐3.3 to 1.8, one trial, 269 participants, moderate certainty evidence).
Compared to the combination of exercise and electrophysical modalities, motor control exercise may provide a small improvement in physical‐health‐related quality of life in the short term (0 to 100 physical‐health related quality of life scale, MD 8.4, 95% CI 2.7 to 14.1, one trial, 38 participants, low certainty evidence) and the intermediate term (0 to 100 physical‐health related quality of life scale, MD 8, 95% CI 2.3 to 13.8, one trial, 38 participants, low certainty evidence).
There may be no difference between motor control exercise and the combination of exercise and electrophysical modalities in mental‐health‐related quality of life in the short term (0 to 100 mental‐health related quality of life scale, MD 2.4 favouring motor control, 95% CI ‐2.1 to 7.1, one trial, 38 participants, low certainty evidence) or i the intermediate term (0 to 100 mental‐health related quality of life scale, MD 1.6 favouring motor control, 95% CI ‐3 to 6.2, one trial, 38 participants, low certainty evidence).
It is uncertain whether there is a difference between motor control exercise and telerehabilitation in physical‐health‐related quality of life in the short term (0 to 100 mental‐health related quality of life scale, MD 0, 95% CI ‐9.0 to 9.0, one trial, 48 participants, very low certainty evidence) or the intermediate term (0 to 100 mental‐health related quality of life scale, MD ‐5 favouring combination of exercise and electrophysical modalities, 95% CI ‐16.3 to 6.3, one trial, 48 participants, very low certainty evidence).
Yoga: Wieland 2022 reported:
Compared to other exercises, it is uncertain whether Yoga exercise improves the physical‐health‐related quality of life (SMD 1.3 Yoga exercise, 95% CI 0.75 to 1.9, one trial, 54 participants, very low certainty evidence) or in the intermediate term (SMD 1.0 Yoga exercise, 95% CI 0.36 to 1.71, one trial, 219 participants, very low certainty evidence).
Compared to other exercises, it is uncertain whether Yoga exercise improves the mental‐health‐related quality of life (SMD 1.03 Yoga exercise, 95% CI ‐0.44 to 2.51, one trial, 219 participants, very low certainty evidence) and the intermediate term (SMD 1.33 Yoga exercise, 95% CI 0.74 to 1.92, one trial, 54 participants, very low certainty evidence).
Compared to other intensive exercises, intensive Yoga exercise may provide a large improvement in physical‐health‐related quality of life in the short term (SMD 1.06 Yoga exercise, 95% CI 0.59 to 1.53, one trial, 80 participants, low certainty evidence), and may provide a large improvement in mental‐health related quality (SMD 0.87 Yoga exercise, 95% CI 0.41 to 1.33, one trial, 80 participants, low certainty evidence).
It is uncertain whether there is a difference between intensive Yoga exercise and Qigong exercise in quality of life in the short to intermediate term (SMD ‐0.1 favouring Qigong, 95% CI ‐0.4 to 1.3, one trial, 111 participants, very low certainty evidence) or in the intermediate term (SMD ‐0.44 favouring Qigong, 95% CI ‐0.8 to ‐0.4, one trial, 107 participants, very low certainty evidence).
Heat and cold therapies
One Cochrane review, judged at low confidence, included five trials with 856 participants and investigated the effects of superficial heat or cold on LBP (French 2006).
Major outcomes
Pain intensity
Acute/subacute LBP
French 2006 reported:
Compared to acetaminophen, it is uncertain whether heated back wrap reduces pain intensity in the short term (0 to 5 pain intensity scale, MD 1.05 heated back wrap, 95% CI 0.62 to 1.48, one trial judged as low quality, 219 participants, rated as very low certainty evidence).
When added to McKenzie exercise, it is uncertain whether heated back wrap reduces pain intensity in the short term (0 to 5 pain intensity scale, MD 0.3 favouring heated back wrap, 95% CI ‐0.41 to 1.01, one trial judged as high quality, 50 participants, rated as very low certainty evidence).
Compared to an education booklet, it is uncertain whether heated back wrap reduces pain intensity in the short term (0 to 5 pain intensity scale, MD 1.1 favouring heated back wrap, 95% CI 0.55 to 1.65, one trial judged as high quality, 51 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between heated back wrap and education booklet in the short term (0 to 5 pain intensity scale, MD 0.8 favouring heated back wrap, 95% CI ‐0.03 to 1.63, one trial judged as high quality, 49 participants, rated as very low certainty evidence).
When added to exercise, it is uncertain whether heated back wrap reduces pain intensity in the short term (0 to 5 pain intensity scale, MD 1.6 favouring heated back wrap, 95% CI 0.89 to 2.31, one trial judged as high quality, 50 participants, rated as very low certainty evidence).
Function
Acute/subacute LBP
French 2006 reported:
Compared to acetaminophen, heated back wrap may provide a small improvement in function in the short term (up to four days, but not on day seven) (0 to 24 disability scale, MD 2, 95% CI 0.86 to 3.14, one trial judged as low quality, 226 participants, rated as low certainty evidence).
Compared to ibuprofen, heated back wrap may provide a small improvement in function in the short term (up to four days, but not on day seven) (0 to 24 disability scale, MD 2.2, 95% CI 1.1 to 3.2, one trial judged as low quality, 219 participants, rated as low certainty evidence).
When added to McKenzie exercise, heated back wrap may not improve function in the short term (0 to 24 pain intensity scale, MD ‐0.9 favouring heated back wrap, 95% CI ‐2.84 to 1.04, one trial judged as high quality, 50 participants, rated as low certainty evidence).
Compared to an education booklet, it is uncertain whether heated back wrap improves function in the short term (0 to 24 disability scale, MD ‐2.3, favouring heated back wrap, 95% CI ‐4.24 to ‐0.36, one trial judged as high quality, 51 participants, rated as very low certainty evidence).
When added to exercise, it is uncertain whether heated back wrap provides little or no improvement in function in the short term (0 to 24 disability scale, MD ‐1.2 favouring heated back wrap, 95% CI ‐3.14 to ‐0.74, one trial judged as high quality, 50 participants, rated as very low certainty evidence).
Compared to an education booklet alone, heated back wrap combined with exercise may provide a small improvement in function in the short term compared to education booklet alone (0 to 24 disability scale, MD ‐2.6, 95% CI ‐4.54 to ‐0.66, one trial judged as high quality, 50 participants, rated as low certainty evidence).
Herbal therapies
Oltean 2014, judged at low confidence, included 14 trials with 2050 participants and investigated the effects of herbal interventions on LBP.
Pain intensity
Chronic LBP
Herbal interventions: Oltean 2014 narratively reported:
It is uncertain whether there is a difference between procumbens extracts (60 mg harpagoside per day) and rofecoxib (12.5 mg per day) in the short term (one trial with unclear allocation concealment, 88 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between salicin (240mg daily, S. alba) and rofecoxib (12.5 mg per day) in the short term (one trial with high risk of bias in blinding and incomplete outcome data, 228 participants, rated as very low certainty evidence).
Pain intensity
Mixed LBP
Herbal interventions: Oltean 2014 narratively reported that it is uncertain whether there is a difference between frutescens gel (Capsici oleoresin) and homoeopathic treatment (Spiroflor SLR homoeopathic gel) in the short term (one trial, 161 participants, rated as very low certainty evidence).
Manual therapies
Five Cochrane reviews investigated the effects of manual therapies on LBP (Rubinstein 2011, Rubinstein 2012, Wegner 2013, Franke 2015, Furlan 2015). Two reviews, judged at high confidence, investigated the effects of spinal manipulation therapy for acute LBP (20 trials with 2674 participants) (Rubinstein 2012) and chronic LBP (26 trials with 6070 participants) (Rubinstein 2011). Wegner 2013, judged at low confidence, included 32 trials with 2762 participants and investigated the effects of traction for LBP. Furlan 2015, judged at high confidence, included 25 trials with 3096 participants and investigated the effects of massage for LBP. Franke 2015, judged at low confidence, included 12 trials with 500 participants and investigated the effects of muscle energy technique for LBP.
Pain intensity
Acute/subacute LBP
Spinal manipulation: Rubinstein 2012 reported:
There probably be no difference between spinal manipulation and other interventions (combination of physical therapy modalities) on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐1.5, 95% CI ‐4.9 to 1.8, I2=0%, three trials, 606 participants, moderate certainty evidence), and there may be no difference in the intermediate term (0 to 100 pain intensity scale, MD ‐2, 95% CI ‐11.3 to 7.3, I2=0%, two trials, 548 participants, low certainty evidence).
It is uncertain whether there is a difference between spinal thrust manipulation and spinal mobilisation in pain intensity in the short term (high risk of bias, 272 participants, rated as very low certainty evidence). The review reported that one trial suggested that side‐lying and supine thrust techniques may provide a small reduction in pain intensity compared to spinal mobilisation, and another trial reported that there may be no difference between the techniques.
Muscle energy therapy: Franke 2015 reported:
It is uncertain whether there is a difference between muscle energy therapy and other interventions (passive mobilisation, electrophysical modality) on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐10.7 favouring muscle energy therapy, 95% CI ‐32.5 to 11.1, I2= 85%, two trials, 88 participants, very low certainty evidence).
There may be no difference between muscle energy therapy added to other interventions (electrophysical modality) on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐3 favouring muscle energy therapy, 95% CI ‐11.3 to 5.3, one trial, 40 participants, low certainty evidence).
There may be no difference between muscle energy therapy added to other interventions (hot moist) and other therapies (positional release therapy) added to that same intervention (hot moist) on pain intensity in the short term (0 to 100 pain intensity scale, MD 5.2 favouring positional release therapy, 95% CI ‐3.0 to 13.4, one trial, 60 participants, low certainty evidence).
Chronic LBP
Muscle energy therapy: Franke 2015 reported:
There may be no difference between muscle energy therapy and other interventions (passive mobilisation) on pain intensity in the short term (0 to 100 pain intensity scale, MD ‐9.7 favouring muscle energy therapy, 95% CI 20.2 to 0.8, one trial, 30 participants, low certainty evidence).
When added to electrophysical modalities (interferential therapy), muscle energy therapy may result in a large reduction in pain intensity (0 to 100 pain intensity scale, MD ‐34.1 95% CI ‐38.4 to ‐29.7, one trial, 30 participants, low certainty evidence).
Mixed LBP
Spinal manipulation: Rubinstein 2011 reported:
Compared to other interventions (NSAIDs, exercises, home exercises, back school, other manual therapies, usual medical care), spinal manipulation provides little to no difference in pain intensity in the short term (0‐100 pain scale, MD ‐2.8, 95% CI ‐5.2 to ‐0.3, I2=27%, 10 trials, 1405 participants, high certainty evidence) and the intermediate term (0‐100 pain scale, MD ‐3.1, 95% CI ‐5.4 to ‐0.7, I2=0%, seven trials, 1105 participants, high certainty evidence).
When added to other interventions (myofascial therapy and exercise), spinal manipulation may provide a small reduction in pain intensity in the short term (0‐100 pain intensity scale, MD ‐5.9, 95% CI ‐10.9 to ‐0.1, I2=0%, three trials, 228 participants, low certainty evidence), and no evidence of a difference between spinal manipulation and other interventions alone on pain intensity in the short term (0‐100 pain intensity scale, MD ‐6.8, 95% CI ‐14.1 to 0.5, I2=0%, two trials, 143 participants, low certainty evidence).
Massage: Furlan 2015 reported that compared to other interventions, it is uncertain whether massage reduces pain intensity in the short term (SMD ‐0.37, 95% CI ‐0.62 to ‐0.13, I2=68%, 12 trials, 964 participants, very low certainty evidence) or the intermediate term (SMD ‐0.4, 95% CI ‐0.80 to ‐0.01, I2=85%, five trials, 757 participants, very low certainty evidence).
Function
Acute/subacute LBP
Spinal manipulation: Rubinstein 2012 reported:
There probably be no difference between spinal manipulation and other interventions (combination of physical therapy modalities) on function in the short term (SMD ‐0.1, 95% CI ‐0.26 to 0.05, I2=0%, three trials, 681 participants, moderate certainty evidence), and there may be no difference in the intermediate term (SMD ‐0.09, 95% CI ‐0.3 to 0.1, I2=51%, two trials, 548 participants, low certainty evidence).
There may be no difference between spinal manipulation added to other interventions (electrophysical modality) and other interventions on function in the short term (SMD ‐0.09 95% CI ‐0.39 to 0.21, I2=37%, three trials, 322 participants, low certainty evidence).
Muscle energy therapy: Franke 2015 reported:
There may be no difference between muscle energy therapy added to other interventions (hot moist) and other therapies (positional release therapy) added to that same intervention (hot moist) on function in the short term (0 to 100 function scale, MD 6 favouring positional release therapy, 95% CI ‐0.5 to 12.4, one trial, 60 participants, low certainty evidence).
There may be no difference between muscle energy therapy and other interventions (passive mobilisation) on function in the short term (0 to 100 function scale, MD 0.9 95% CI ‐6.3 to 8.1, one trial, 50 participants, low certainty evidence).
When added to electrophysical modalities (interferential therapy), muscle energy therapy may provide a medium improvement in function (0 to 100 function scale, MD ‐17.6 95% CI ‐27.1 to ‐8.2, one trial, 40 participants, low certainty evidence).
Chronic LBP
Muscle energy therapy: Franke 2015 reported:
There may be no difference between muscle energy therapy and other interventions (electrophysical modalities) on function in the short term (0 to 100 function scale, MD ‐4.1, 95% CI ‐9.5 to 1.3, one trial, 30 participants, low certainty evidence).
When added to other interventions (electrophysical modalities), muscle energy therapy may provide a large improvement in function in the short term (0 to 100 function scale, MD ‐22, 95% CI ‐27.4 to ‐16.5, one trial, 30 participants, low certainty evidence).
There may be no difference between muscle energy therapy added to other interventions (exercise) and other interventions (mobilisation) added to that same intervention (exercise) on function in the short term (SMD ‐18 95% CI ‐0.43 to 0.08, I2=0%, six trials, 232 participants, low certainty evidence).
Mixed LBP
Spinal manipulation: Rubinstein 2011 reported:
Compared to other interventions (NSAIDs, exercises, home exercises, back school, other manual therapies, usual medical care), spinal manipulation provides little to no difference in function in the short term (SMD ‐0.17, 95% CI ‐0.29 to ‐0.06, I2=3%, six trials, 1402 participants, high certainty evidence) and the intermediate term (SMD ‐0.12, 95% CI ‐0.23 to ‐0.00, I2=0%, five trials, 1313 participants, high certainty evidence).
When added to other interventions, spinal manipulation may provide a small improvement in function in the short term (SMD ‐0.4, 95% CI ‐0.73 to ‐0.07, I2=0%, two trials, 165 participants, low certainty evidence), but there may be no difference in the intermediate term (SMD ‐0.3, 95% CI ‐0.64 to 0.03, I2=0%, two trials, 142 participants, low certainty evidence).
Massage: Furlan 2015 reported that it is uncertain whether there is a difference between massage and other interventions in the short term (SMD ‐0.24, 95% CI ‐0.62 to 0.13, I2=78%, six trials, 618 participants, very low certainty evidence) and the intermediate term (SMD ‐0.21, 95% CI ‐0.60 to 0.17, I2=81%, four trials, 618 participants, very low certainty evidence).
Quality of life
Mixed LBP
Spinal manipulation: Rubinstein 2011 reported that there is no difference between spinal manipulation and other interventions (NSAIDs, exercises, home exercises, back school, other manual therapies, usual medical care) in the short term (SMD ‐0.08 favouring other interventions, 95% CI ‐0.29 to 0.13, I2=0%, three trials, 361 participants, high certainty evidence) and the intermediate term (SMD ‐1 favouring other interventions, 95% CI ‐1.75 to ‐0.24, one trial judged as low quality, 31 participants, rated as low risk of bias).
Participant ratings of improvement
Acute/subacute LBP
Spinal manipulation: Rubinstein 2012 reported:
There may be no difference between spinal manipulation and other interventions (combination of physical therapy modalities) on participant ratings of improvement or recovery in the short term (RR 1.06 95% CI 0.94 to 1.21, I2=0%, two trials, 117 participants, low certainty evidence).
There may be no difference between spinal manipulation added to other interventions (combination of exercise and posture advice) on participant ratings of improvement or recovery compared to other interventions alone (RR 1.15 95% CI 0.6 to 2.19, I2=78, two trials, 225 participants, rated as low certainty evidence).
Mixed LBP
Spinal manipulation: Rubinstein 2011 reported that compared to other interventions (NSAIDs, exercises, home exercises, back school, other manual therapies, usual medical care), spinal manipulation probably increases the chance of perceiving recovery with the intervention in the short term (RR 1.2, 95% CI 1.04 to 1.37, I2=0%, three trials, 370 participants, moderate certainty evidence), but there may be no difference in the intermediate term (RR 1.05, 95% CI 0.81 to 1.38, one trial, 112 participants, low certainty evidence.
Workplace participation
Acute/subacute LBP
Spinal manipulation: Rubinstein 2012 reported:
It is uncertain whether there is a difference between spinal manipulation and other interventions (combination of physical therapy modalities) and other interventions alone on return to work in the short (RR 1.01 95% CI 0.91 to 1.12, one trial, 311 participants, very low certainty evidence) or in the intermediate term (RR 1.07 95% CI 0.98 to 1.16, one trial, 313 participants, very low certainty evidence).
It is uncertain whether there is a difference between spinal manipulation added to other interventions (electrophysical modality) and other interventions alone on return to work in the short term (RR 1.21 95% CI 0.99 to 1.47, one trial, 92 participants, very low certainty evidence).
Mixed LBP
Spinal manipulation: Rubinstein 2011 reported no evidence of a difference between spinal manipulation and other interventions (NSAIDs, exercises, home exercises, back school, other manual therapies, usual medical care) in the short term (RR 1.1, 95% CI 0.91 to 1.35, one trial, 71 participants, low certainty evidence) and the intermediate term (RR 1.09, 95% CI 0.98 to 1.21, three trials, 389 participants, low certainty evidence).
Multimodal therapies
Three Cochrane reviews investigated the effects of multimodal therapies on LBP (Walker 2010, Schaafsma 2013, Parry 2019). Walker 2010, judged at low confidence, included 12 trials with 2887 participants and investigated the effects of chiropractic (combination of spinal manipulation, massage, thermotherapies, electrophysical therapies, exercise, nutritional advice, lifestyle modification and education) for LBP. Schaafsma 2013, judged at low confidence, included 25 trials with 4404 participants and investigated the effects of physical conditioning (combination of exercise, functional restoration, duties modification, education, and behavioural approaches) as part of a return‐to‐work strategy to reduce sickness absence for workers with LBP. Schaafsma 2013 reported separate results for light physical conditioning (fewer than five sessions) and intense physical conditioning (more than five sessions). Parry 2019, judged at low confidence, included 10 trials with 955 participants and investigated the effects of workplace interventions for increasing standing or walking (e.g., sit‐stand workstation, treadmill workstation, activity tracker, break‐reminding, policy modification, modification to office layout) in sedentary workers with LBP.
Pain intensity
Acute/subacute LBP
Chiropractic: Walker 2010 reported that compared to other interventions (the McKenzie approach plus education; ultrasound plus gentle massage; Back School Program; ice packs), chiropractic may provide a small reduction in pain intensity in the short term (SMD ‐0.25, 95% CI ‐0.46 to ‐0.04, I2=0%, three trials, 423 participants, rated as low certainty evidence), but there is no evidence of a difference at the intermediate term (0 to 100 pain intensity scale, MD ‐4.6 favouring chiropractic, 95% CI ‐11.8 to 2.6, one trial, 138 participants, rated as low certainty evidence).
Chronic LBP
Chiropractic: Walker 2010 reported there may be no difference between chiropractic and other interventions (NSAID; active and stretching exercises; medication and/ or injections) in the short term (0 to 100 pain intensity scale, MD ‐6.8 favouring chiropractic, 95% CI ‐14.3 to 0.7, I2=42%, three trials, 388 participants, rated as low certainty evidence) or in the intermediate term (0 to 100 pain intensity scale, MD ‐4.7 favouring chiropractic, 95% CI ‐11.6 to 2.2, one trial, 174 participants, rated as low certainty evidence).
Mixed LBP
Chiropractic: Walker 2010 reported:
When added to hot pack, chiropractic may provide a medium reduction in pain intensity in the short term (0 to 100 pain intensity scale, ‐14.4, 95% CI ‐20.2 to ‐0.9, one trial with high risk of bias, 217 participants, rated as low certainty evidence).
There probably be no difference between chiropractic with medical care and physiotherapy with medical care in pain intensity in the short term (0 to 100 pain intensity scale, 0, 95% CI ‐3.7 to 3.7, one trial with high risk of bias, 655 participants randomised, rated as moderate certainty evidence), there may be no difference in the intermediate term (0 to 100 pain intensity scale, ‐2.2 favouring chiropractic, 95% CI ‐6.5 to 2.1, one trial with high risk of bias, 655 participants randomised, rated as low certainty evidence), and there probably be no difference in the long term (0 to 100 pain intensity scale, ‐2.2 favouring chiropractic, 95% CI ‐6.9 to 2.5, one trial with high risk of bias, 655 participants, rated as moderate certainty evidence).
Compared to an unidentified control group, chiropractic may provide a small increase in the chance of experiencing partial or complete pain relief in the intermediate term (RR 1.12, 95% CI 1.04 to 1.21, one trial with high risk of bias, 450 participants, rated as low certainty evidence), but there is no evidence of a difference between chiropractic and an unidentified control group.
Function
Acute/subacute LBP
Chiropractic: Walker 2010 reported that compared to other interventions (the McKenzie approach plus education; ultrasound plus gentle massage; Back School Program; ice packs), chiropractic may provide a small reduction in pain intensity in the short term (SMD ‐0.36, 95% CI ‐0.7 to ‐0.02, I2=65%, four trials, 545 participants, rated as low certainty evidence), but there may be no difference between chiropractic and other interventions in the intermediate term (0 to 24 disability scale, MD ‐0.75, 95% CI ‐2.07 to 0.57, one trial, 137 participants, rated as low certainty evidence).
Chronic LBP
Chiropractic: Walker 2010 reported that there probably be no difference between chiropractic and other interventions (NSAID, active and stretching exercises, medication and/or injections) in the short term (0 to 24 disability scale, MD ‐0.67 chiropractic, 95% CI ‐1.62 to 0.28, I2=0%, three trials, 409 participants, rated as moderate certainty evidence), and there may be no difference in the intermediate term (0 to 24 disability scale, MD ‐0.13 chiropractic, 95% CI ‐1.48 to 1.22, one trial, 173 participants, rated as low certainty evidence).
Mixed LBP
Chiropractic: Walker 2010 reported that it is uncertain whether there is a difference between chiropractic and other interventions (i.e., massage and medical care with physiotherapy) in function in the short term (0 to 100 pain intensity scale, ‐3.1, 95% CI ‐10.4 to 4.2, I2=0%, two trials, 761 participants, rated as very low certainty evidence) or in the intermediate term (0 to 100 pain intensity scale, ‐7.7, 95% CI ‐16.5 to 1, I2=0%, two trials, 872 participants, rated as very low certainty evidence).
Quality of life
Chronic LBP
Chiropractic: Walker 2010 reported there may be no difference between chiropractic and other interventions (NSAID, active and stretching exercises, medication and/or injections) in the short term (0 to 100 quality of life scale, MD ‐0.04 favouring chiropractic, 95% CI ‐1.75 to 1.68, two trials, 378 participants, low certainty evidence) or in the intermediate term (0 to 100 quality of life scale, MD ‐0.4 favouring chiropractic, 95% CI ‐4.9 to 4.0, one trial, 132 participants, low certainty evidence).
Participant ratings of improvement
Mixed LBP
Chiropractic: Walker 2010 reported that it is uncertain whether there is a difference between chiropractic and other interventions (i.e., physiotherapy) in treatment satisfaction in the short term (RR 1.05, 95% CI% 0.57 to 1.94, one trial with high risk of bias, 19 participants, rated as very low certainty evidence) and intermediate‐term (RR 1.20, 95% CI% 0.69 to 2.09, one trial with high risk of bias, 19 participants, rated as very low certainty evidence).
Workplace participation
Acute/subacute LBP
Physical conditioning: Schaafsma 2013 reported:
There may be no difference between light physical conditioning and back book education in workplace participation (sick leave) in the short term (OR 0.36, 95% CI 0.13 to 1.03, one trial, 80 participants, low certainty evidence).
There may be no difference between light physical conditioning added to brief clinical intervention and brief clinical intervention alone (examination by a physician, relevant imaging and examinations were ordered, and treatment options were discussed) in workplace participation (sick leave) in the intermediate term (SMD 0.21 favouring brief clinical intervention, 95% CI 0 to 0.42, one trial, 351 participants, low certainty evidence).
There may be no difference between light physical conditioning and intense physical conditioning in workplace participation (sick leave) in the intermediate term (SMD ‐0.24 favouring light physical conditioning, 95% CI ‐0.52 to 0.04, one trial, 196 participants, low certainty evidence).
Compared to multidisciplinary exercise treatment, intense physical conditioning may decrease the likelihood of experiencing sick leave in the short term (OR 0.41 95% CI 0.22 to 0.78, one trial 173 participants, low certainty evidence) and the intermediate term (OR 0.72 95% CI 0.4 to 0.99, one trial, 149 participants, low certainty evidence).
Chronic LBP
Physical conditioning: Schaafsma 2013 reported:
There may be no difference between intensive physical conditioning and exercise in workplace participation in the short term (OR 1 favouring exercise, 95% CI 0.37 to 2.7, one trial, 136 participants, low certainty evidence), in the intermediate (SMD 0.19 favouring intensive physical conditioning, 95% CI ‐0.63 to 0.24, I2=20%, two trials, 114 participants, low certainty evidence), and it is uncertain whether intensive physical conditioning improves workplace participation (return to work) in the long term (SMD ‐0.62, 95% CI ‐1.21 to ‐0.04, one trial, 52 participants, very low certainty evidence).
There may be no difference between intensive physical conditioning and cognitive behavioural therapy in workplace participation in the long term (SMD ‐0.47 favouring intensive physical conditioning, 95% CI ‐1.36 to 0.42, I2=52%, low certainty evidence).
There may be no difference between light physical conditioning and intense physical conditioning in workplace participation in the intermediate term (SMD 0.1 favouring light physical conditioning, 95% CI ‐0.28 to 0.48, one trial, 109 participants, low certainty evidence) or in the long term (SMD 0.2 favouring light physical conditioning, 95% CI ‐0.18 to 0.58, one trial, 109 participants, low certainty evidence).
Multidisciplinary therapies
Two Cochrane reviews investigated the effects of multidisciplinary biopsychosocial interventions on LBP (Kamper 2014, Marin 2017). Kamper 2014, judged at moderate confidence, included 41 trials with 6858 participants and investigated the effects of multidisciplinary biopsychosocial interventions on chronic LBP. Marin 2017, judged at low confidence, included nine trials with 981 participants and investigated the effects of multidisciplinary biopsychosocial interventions on subacute LBP.
Pain intensity
Acute/subacute LBP
Marin 2017 reported that it is uncertain whether there is a difference between multidisciplinary therapies and other interventions on pain intensity in the short term (SMD ‐0.09 favouring multidisciplinary therapies, 95% CI ‐0.5 to 0.33, I2=44%, two trials, 165 participants, very low certainty evidence), or in the intermediate term (SMD ‐0.64 favouring multidisciplinary therapies, 95% ‐1.85 to 0.57, I2=92%, two trials, 162 participants, very low certainty evidence), and there may be no difference in the long term (SMD ‐0.14 favouring multidisciplinary therapies, 95% CI ‐0.36 to 0.07, I2=0%, two trials, 336 participants, low certainty evidence).
Chronic LBP
Kamper 2014 reported:
Compared to physical interventions, multidisciplinary therapies may provide a small improvement in function in the short term (SMD ‐0.39, 95% CI ‐0.68 to ‐0.10, I2=88%, 13 trials, 1878 participants, low certainty evidence) and may provide a medium improvement in function in the long term (SMD ‐0.68 favouring multidisciplinary intervention, 95% CI 1.19 to ‐0.16, I2=94%, ten trials, 1169 participants, low certainty evidence). However, there may be no difference between multidisciplinary therapies and physical interventions in the intermediate (SMD ‐0.21favouring multidisciplinary intervention, 95% CI ‐0.48 to 0.06, I2=52%, nine trials, 511 participants, low certainty evidence).
There may be no difference between multidisciplinary therapies and surgery in function in the long term (SMD ‐0.25 favouring surgery, 95% CI ‐0.08 to 0.57, I2=61%, two trials, 423 participants, low certainty evidence).
Function
Acute/subacute LBP
Marin 2017 reported that there may be no difference between multidisciplinary therapies and other interventions on function in the short term (SMD 0, 95% CI ‐0.34 to 0.34, I2=17%, two trials, 165 participants, low certainty evidence), there may be no difference in the intermediate term (SMD ‐0.49 favouring multidisciplinary therapies, 95% CI ‐1.50 to 0.51, I2=89%, 162 participants, very low certainty evidence), and there may be no difference in the long term (SMD ‐0.03, 95% CI ‐0.24 to 0.18, I2=0%, two trials, 345 participants, low certainty evidence).
Chronic LBP
Kamper 2014 reported:
Compared to physical interventions, multidisciplinary therapies may provide a small improvement in function in the short term (SMD ‐0.39, 95% CI ‐0.68 to ‐0.10, I2=88%, 13 trials, 1878 participants, low certainty evidence) and may provide a medium improvement in function in the long term (SMD ‐0.68 favouring multidisciplinary intervention, 95% CI 1.19 to ‐0.16, I2=94%, ten trials, 1169 participants, low certainty evidence). However, no evidence of a difference between multidisciplinary therapies and physical interventions in the intermediate (SMD ‐0.21favouring multidisciplinary intervention, 95% CI ‐0.48 to 0.06, I2=52%, nine trials, 511 participants, low certainty evidence).
No evidence of a difference between multidisciplinary therapies and surgery in function in the long term (SMD ‐0.25 favouring surgery, 95% CI ‐0.08 to 0.57, I2=61%, two trials, 423 participants, low certainty evidence).
Safety
Chronic LBP
Kamper 2014 reported that multidisciplinary therapies may decrease the likelihood of experiencing adverse events in the long term compared to surgery (OR 28.25, 95% CI 3.77 to 211.9, two trials, 385 participants, low certainty evidence).
Workplace participation
Acute/subacute LBP
Marin 2017 reported that :
It is uncertain whether there is a difference between multidisciplinary therapies and usual care on workplace participation (return to work rate) in the long term (SMD ‐0.25, 95% CI ‐0.98 to ‐0.47, I2=78%, very low certainty evidence).
It is uncertain whether there is a difference between multidisciplinary therapies and usual care on workplace participation (time to return to work) in the long term (one trial with high risk of bias, 351 participants, very low certainty evidence)
Chronic LBP
Other interventions
Kamper 2014 reported:
There probably be no difference between multidisciplinary therapies and physical interventions in workplace participation in the short term (OR 1.6 favouring multidisciplinary intervention, 95% CI 0.92 to 2.78, I2=23, three trials, 379 participants, moderate certainty evidence), but multidisciplinary therapies probably increase workplace participation (being at work) in the intermediate term (OR 2.1, 95% CI 1.1 to 4.1, I2=0, three trails, 221 participants, moderate certainty evidence), and in the long term (OR 1.87, 95% CI 1.39 to 2.53, I2=0, three trails, 1006 participants, moderate certainty evidence).
There may be no difference between multidisciplinary therapies and physical interventions in workplace participation compared to surgery in the long term (OR 0.67 favouring surgery, 95% CI 0.31 to 1.45, I2=0, one trial, 133 participants, low certainty evidence).
Quality of life
Chronic LBP
Kamper 2014 reported:
There may be no difference between multidisciplinary therapies and physical interventions in quality of life in the short term (0 to 100 quality of life scale, MD ‐0.04 favouring physical interventions, 95% CI ‐0.3 to 0.3, I2=63%, three trials, 568 participants, rated as low certainty evidence) or in the intermediate term (0 to 100 quality of life scale, MD 0.2 favouring multidisciplinary intervention, 95% CI ‐0.1 to 0.5, I2=47%, two trials, 342 participants, rated as low certainty evidence).
Compared to surgery, the quality of life of the multidisciplinary therapies in the long term was not clearly described in the review.
Neuroreflexology
Urrutia 2004, judged at low confidence, included three trials with 273 participants and investigated the effects of neuroreflexotherapy (i.e. temporary implantation of several epidermal devices into trigger points in the back and into referred tender points in the ear) on LBP. The review did not report data for neuroreflexotherapy compared to other interventions.
Psychological therapies
One Cochrane review, judged at low confidence, included 31 trials with 3438 participants and investigated the effects of psychological therapies on LBP (Henschke 2010). The review included several psychological therapies such as operant approaches (dealing with external factors associated with pain), respondent approaches (dealing with muscle tension through progressive relaxation or biofeedback), cognitive approaches (dealing with thoughts, feelings, beliefs, or a combination of the three), and combined approaches (combining cognitive and behavioural approaches such as cognitive behavioural therapy).
Pain intensity
Chronic LBP
Henschke 2010 reported:
One trial investigated the effect of two types of respondent approaches. The review narratively reported that it is uncertain whether there is a difference between EMG biofeedback and relaxation in pain intensity in the short term (one trial with high risk of bias, 24 participants, rated as very low certainty evidence).
There probably be no difference between cognitive approaches and operant approaches in pain intensity in the short term (SMD 0.41favouring operant approach, 95% CI ‐0.63 to 1.45, I2=70%, two trials, 93 participants, moderate certainty evidence) or the intermediate term (SMD 0.35 favouring operant approach, 95% CI ‐0.64 to 1.35, I2=40%, two trials, 82 participants, moderate certainty evidence).
It is uncertain whether there is a difference between cognitive approaches and respondent approach (progressive relaxation) in pain intensity in the short term (one trial with high risk of bias, 47 participants, rated as very low certainty evidence).
It is uncertain whether there is a difference between cognitive approaches and combined approaches in pain intensity in the short term (SMD ‐0.24 favouring combined approaches, 95% CI ‐1.36 to 0.8, I2=75%, two trials, 61 participants, very low certainty evidence), in the intermediate term (SMD ‐0.3 favouring combined approaches, 95% CI ‐2.59 to 1.98, I2=90%, two trials, 44 participants, very low certainty evidence), and in the long term (SMD ‐0.89 favouring combined approaches, 95% CI ‐3.64 to 1.87, I2=91%, two trials, 48 participants, very low certainty evidence)
There probably be no difference between operant approaches and combined approaches in pain intensity in the short term (SMD ‐0.15 favouring combined approaches, 95% CI ‐0.46 to 0.16, I2=0%, three trials, 161 participants, moderate certainty evidence), in the intermediate term (SMD ‐0.23 favouring combined approaches, 95% CI ‐0.57 to 0.11, I2=0%, three trials, 161 participants, moderate certainty evidence), or in the long term (SMD ‐0.31 favouring combined approaches, 95% CI ‐0.65 to 0.03, I2=0%, three trials, 140 participants, moderate certainty evidence).
There may be no difference between respondent approaches (progressive relaxation) and combined approaches in pain intensity in the short term (SMD 0.09 favouring respondent approaches, 95% CI ‐0.31 to 0.5, I2=0%, three trials, 97 participants, low certainty evidence) or in the intermediate term (SMD 0.47 favouring respondent approaches, 95% CI ‐0.42 to 1.35, I2=65%, three trials, 62 participants, low certainty evidence).
There may be no difference between psychological therapies (combined approaches, progressive muscle relaxation) and exercise in pain intensity in the short term (SMD ‐2.31 favouring combined approaches, 95% CI ‐6.3 to 1.7, I2=0%, two trials, 146 participants, low certainty evidence), there probably be no difference in the intermediate term (SMD 1.18 favouring exercise, 95% CI ‐3.16 to 5.53, I2=0%, two trials, 146 participants, moderate certainty evidence) or in the long term (SMD 0.14 favouring exercise, 95% CI ‐4.4 to 4.67, I2=0%, two trials, 136 participants, moderate certainty evidence).
When added to physiotherapy, psychological therapies may not reduce pain intensity in the short term (0 to 6 pain intensity scale, MD ‐0.13 favouring psychological intervention, 95% CI ‐1 to 0.75, I2=73%, two trials, 59 participants, low certainty evidence).
When added to physiotherapy and education, psychological therapies may not reduce pain intensity in the short term (0 to 6 pain intensity scale, MD ‐0.11 favouring psychological intervention, 95% CI ‐0.67 to 0.44, I2=0%, two trials, 45 participants, low certainty evidence).
When added to inpatient rehabilitation, psychological therapies probably do not reduce pain intensity in the short term (SMD ‐0.14 favouring psychological intervention, 95% CI ‐0.34 to 0.05, I2=0%, two trials, 405 participants, moderate certainty evidence).
Function
Chronic LBP
Henschke 2010 reported:
There may be no difference between combined approaches and cognitive approaches in function in the short term (SMD ‐2 favouring combined approaches, 95% CI ‐10 to 5.9, I2=62%, two trials, 61 participants, low certainty evidence), and it is unclear of its beneficial effect in the intermediate term (SMD ‐3.2 favouring combined approaches, 95% CI ‐16.44 to 10, I2=83%, two trials, 47 participants, very low certainty evidence), and the long term (SMD ‐2.23 favouring combined approaches, 95% CI ‐12.5 to 8.1, I2=79%, two trials, 47 participants, very low certainty evidence).
There may be no difference between combined approaches and operant approaches in function in the short term (SMD 0.21 favouring operant approaches, 95% CI ‐0.24 to 0.67, I2=0%, two trials, 77 participants, low certainty evidence), in the intermediate term (SMD ‐0.23 favouring combined approaches, 95% CI ‐1 to 0.55, I2=46%, two trials, 61 participants, low certainty evidence), or in the long term (SMD ‐0.51 favouring combined approaches, 95% CI ‐1.5 to 5.6, I2=69%, two trials, 66 participants, low certainty evidence).
There may be no difference between combined approaches and respondent approaches (progressive relaxation) in function in the short term (SMD 0.38 favouring respondent approaches, 95% CI ‐0.02 to 0.78, I2=0%, three trials, 97 participants, low certainty evidence), or in the intermediate term (SMD 0.13 favouring respondent approaches, 95% CI ‐0.81 to 1.07, I2=0%, two trials, 62 participants, low certainty evidence).
There may be no difference between psychological therapies (combined approaches, progressive muscle relaxation) and surgical stabilisation in the long term (0 to 100 function scale, MD 2.4 favouring surgical stabilisation, 95% CI ‐1.9 to 6.6, I2=0%, two trials, 345 participants, low certainty evidence).
There may be no between psychological therapies (combined approaches, progressive muscle relaxation) and physiotherapy with back education in the short term (0 to 136 function scale, MD ‐6.3 favouring psychological therapies, 95% CI ‐12.7 to 0.2, I2=0%, two trials, 59 participants, low certainty evidence), or in the intermediate term (0 to 136 function scale, MD ‐0.9 favouring psychological therapies, 95% CI ‐6.7 to 4.8, I2=24%, two trials, 51 participants, low certainty evidence).
Appendix 4. Reasons for excluding reviews
| Name of review | Reason for exclusion |
| Assendelft 2013 | Withdrawn review |
| Bleakley 2012 | Ineligible patient population |
| Choi 2010 | Ineligible outcomes |
| Costello 2015 | Ineligible patient population |
| Eccleston 2014 | Ineligible patient population |
| Eccleston 2017 | Ineligible outcomes |
| French 2010 | Ineligible outcomes |
| Gadsby 2007 | Withdrawn review |
| Hagen 2010 | Withdrawn review |
| Hayden 2019 | Ineligible study design |
| Henschke 2013 | Ineligible outcomes |
| Heymans 2004 | Updated review |
| Hilde 2007 | Withdrawn review |
| Jordan 2010 | Ineligible outcomes |
| Karjalainen 2000 | Ineligible patient population |
| Nnoaham 2014 | Ineligible patient population |
| O'Connell 2018 | Ineligible patient population |
| Sahar 2007 | Ineligible outcomes |
| Verbeek 2011 | Ineligible outcomes |
| Wieland 2017 | Updated review |
| Williams 2020 | Ineligible patient population |
Appendix 5. Characteristics of the interventions in the included systematic reviews
| Non‐pharmacological intervention class/ definition | Treatment duration | Number of sessions | Duration of the sessions |
| Acupuncture and dry needling |
Acute or subacute LBP Median (range): 2 (1 to 3) weeks Chronic LBP Median (range): 4 weeks (1 week to 8 months) |
Acute or subacute LBP Median (range) 4 (1 to 10) sessions Chronic LBP Median (range): 10 (1 to 28) sessions |
Acute or subacute LBP Median (range): 20 (20 to 30) minutes Chronic LBP Median (range): 25 (10 to 35) minutes |
| Advice and education |
Acute LBP Range: 1 session to 8 weeks Chronic LBP Median (range): 4 weeks (1 weeks to 6 months) |
Acute LBP Range: 1 to 4 sessions Chronic LBP Median (range): 4 (1 to 36 sessions) |
Advice Range: 45 to 90 minutes Education Median (range): 60 (5 minutes to 4 hours) |
| Lumbar support | Median (range): 4 (3 to 8) weeks | Not reported | Not reported |
| Electrical physical modalities |
Low laser therapy Median (range): 4 (1 day to 6 weeks) Therapeutic ultrasound Not reported TENS Median (range): 3 (2 to 4) weeks |
Low laser therapy Median (range): 12 (1 to 20) Therapeutic ultrasound 10 (6 to 18 weeks) TENS Median (range): 15 (10 to 60 sessions) |
Low laser therapy Median (range) of 8 points (6 to 10) with median (range) of 4 (1 to 10) minutes Therapeutic ultrasound Gray’s formula for calculation of the application time or ultrasound for 5 to 10 minutes with 1 MHz continuous ultrasound at intensities between 1 W/cmW and 2.5 W/cmW. TENS 50 minutes (20 minutes to 3 hours) |
| Exercise therapies | Most from 4 to 8 weeks (median), ranging from 1 week (for acute and subacute pain) to 30 months | Most from 12 to 17 sessions, ranging from 1 (for acute and subacute pain) to 45 sessions | Median (range): 60 minutes (20 to 90) minutes |
| Heat and cold therapies | Median (range): 4 days (1 day to 2 weeks) | Not reported | Range: 20 minutes to 8 hours at night |
| Herbal therapies | Median: 3 weeks | Everyday once a day, to everyday three times a day | Not reported |
| Manual therapies | Most from 1 to 5 weeks, ranging from 1 week to 5 months | Most from 6 to 10 sessions, ranging from 1 to 16 sessions | Most from 15 to 30 minutes, ranging from 5 to 60 minutes |
| Multimodal therapies | Range: 10 days to 12 months | Range: 3 to 20 sessions | Range: 10 to 60 minutes |
| Multidisciplinary therapies | Range: 2 to 18 weeks | Range: 1 to 26 sessions | Range: 1 to 6 hours |
| Neuroreflexotherapy | Range: 30 days to 60 days | 2 sessions (implementation and follow‐up) | 60 minutes |
| Psychological therapies | Median (range): 6 (1 to 10) weeks | Median (range): 8 (1 to 30) sessions | Median (range): 60 minutes (25 minutes to 8 hours) |
Appendix 6. Results reported in the overview of non‐pharmacological interventions for LBP stratified by comparisons and pain duration
| Interventions/comparison* | Placebo/sham | No treatment/usual care | Other interventions | ||||||
| Acute/ subacute | Chronic | Mixed | Acute/ subacute | Chronic | Mixed | Acute/ subacute | Chronic | Mixed | |
| Acupuncture | x | x | x | x | x | x | |||
| Advice to rest or stay active | x | x | |||||||
| Lumbar support | x | x | |||||||
| Back school | x | x | x | x | x | ||||
| Chiropractic | x | x | x | x | |||||
| Cold therapy | |||||||||
| Dry needling | |||||||||
| Exercise | x | x | x | x | x | x | |||
| Heat therapy | x | x | x | ||||||
| Herbal interventions | x | x | |||||||
| Individual patient education | x | x | x | x | x | x | |||
| Low‐level laser therapy | x | x | x | ||||||
| Massage | x | ||||||||
| McKenzie exercise | x | ||||||||
| Motor control exercise | x | x | x | ||||||
| Multidisciplinary therapies | x | x | x | x | |||||
| Muscle energy technique | x | x | |||||||
| Neuroreflexotherapies | x | x | x | x | x | ||||
| Physical conditioning with lifestyle modification | x | x | x | x | |||||
| Pilates | x | x | |||||||
| Psychological therapies | x | x | |||||||
| Spinal manipulation | x | x | |||||||
| TENS | x | ||||||||
| Traction | x | ||||||||
| Ultrasound | x | x | |||||||
| Workplace intervention to increase standing and walking | x | ||||||||
| Yoga | x | x | |||||||
*Missing fields mean the absence of a comparison for a particular intervention identified in the overview.
Appendix 7. Overview of reviews table with all outcomes (except pain intensity) comparing acupuncture and dry needling with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Acupuncture and dry needling for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Chronic | Acupuncture vs sham | SMD ‐0.38, 95% CI ‐0.69 to ‐0.07 | 67 | 957 (3 RCTs) | Moderate | Downgraded for inconsistency |
| Function (0 to 70 Pain Disability Index) |
Chronic | Acupuncture vs sham | MD ‐3.28, 95% CI ‐6.39 to ‐0.17 | 0 | 285 (2 RCTs) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Acupuncture vs no treatment | SMD ‐0.39, 95%CI ‐0.72 to ‐0.06 | 0 | 144 (3 RCTs) | Moderate | Downgraded for imprecision |
| Function | Chronic | Acupuncture vs usual care | MD 9.4, 95% CI 6.15 to 12.65 | ‐ | 734 (1 RCT) | Moderate | Downgraded for study limitations |
| Function | Chronic | Acupuncture added to usual care vs usual care alone | SMD ‐0.51, 95% CI ‐0.80 to ‐0.22 | 0 | 187 (2 RCTs) | Low | Downgraded for study limitations and imprecision |
| Safety | Chronic | Acupuncture vs sham | RR 0.68, 95% CI 0.42 to 1.10 | 0 | 465 (4 RCTs) | Low | Downgraded for study limitations and imprecision *Rated as moderate according to the Overview GRADE judgement |
| Participant ratings of improvement (6‐point improvement scale) |
Chronic | Acupuncture vs usual care | MD ‐0.3, 95% CI ‐0.50 to ‐0.10 | ‐ | 730 (1 RCT) | Moderate | Downgraded for study limitations |
| Quality of life (physical) | Chronic | Acupuncture vs usual care | MD 4.2, 95% CI 2.82 to 5.58 | ‐ | 731 (1 RCT) | Moderate | Downgraded for study limitations |
| Quality of life (mental) |
Chronic | Acupuncture vs usual care | MD 1.9, 95% CI 0.25 to 3.55 | ‐ | 731 (1 RCT) | Moderate | Downgraded for study limitations |
| Quality of life | Chronic | Acupuncture added to usual care vs usual care alone | MD 7, 95% CI 0.95 to 13.05 | ‐ | 143 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Participant ratings of improvement (6‐point improvement scale) |
Chronic | Acupuncture vs sham | MD ‐0.3, 95% CI ‐0.49 to ‐0.01 | ‐ | 747 (1 RCT) | Moderate | Downgraded for study limitations |
| Quality of life | Chronic | Acupuncture vs sham | SMD 0.24, 95% CI 0.03 to 0.45 | 0 | 1068 (3 RCTs) | Low | Downgraded for study limitations and inconsistency |
| Quality of life (physical) |
Chronic | Acupuncture vs no treatment | MD 5.1, 95% CI 3.09 to 7.11 | 58 | 2837 (2 RCTs) | Low | Downgraded for study limitations and inconsistency |
| Quality of life (mental) |
Chronic | Acupuncture vs no treatment | MD 1.85, 95% CI 1.06 to 2.64 | 0 | 2837 (2 RCTs) | Low | Downgraded for study limitations and inconsistency |
| Workplace participation | Chronic | Acupuncture vs no treatment | SMD ‐0.56, 95% CI ‐0.64 to ‐0.49 | 0 | 2937 (2 RCTs) | Moderate | Downgraded for study limitations |
MD on a 0 to 100 scale unless otherwise indicated. We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; SMD: standardised mean difference
Appendix 8. Overview of reviews table with all outcomes (except pain intensity) comparing advice and education with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Advice and education for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Chronic | Back school vs no treatment | MD ‐3.38, 95% CI ‐6.7 to ‐0.05 | 0 | 426 (3 RCTs) | Very low | Downgraded for study quality, imprecision and publication bias |
| Function | Chronic | Back school vs usual medical care | MD ‐1.19, 95% CI ‐7.02 to 4.64 | 0 | 130 (1 RCT) | Very low | Downgraded for study quality, imprecision and publication bias |
MD on a 0‐100 scale unless otherwise indicated.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial
Appendix 9. Overview of reviews table with all outcomes (except pain intensity) comparing electrophysical modalities with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Electrophysical modalities for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Chronic | Therapeutic ultrasound vs placebo (detuned ultrasound) | SMD ‐0.29, 95% CI ‐0.51 to ‐0.07 | 0 | 325 (4 RCTs) | Low | Downgraded for study limitations and imprecision |
| Function | Mixed | Low laser therapy vs sham | MD ‐8.2, 95% CI ‐13.4 to 2.9 | ‐ | 56 (1 RCT) | Low | Downgraded imprecision |
| Quality of life | Chronic | Therapeutic ultrasound vs placebo (detuned ultrasound) | MD ‐2.71, 95% CI ‐9.85 to 4.44 | 29 | 267 (2 RCT) | Moderate | Downgraded imprecision |
| Participant ratings of improvement | Mixed | Low laser therapy vs sham | MD ‐9.5, 95% CI ‐20.9 to 1.9 | ‐ | 56 (1 RCT) | Low | Downgraded imprecision |
MD on a 0‐100 scale unless otherwise indicated. We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; SMD: standardised mean difference
Appendix 10. Overview of reviews table with all outcomes (except pain intensity) comparing exercise with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Exercise for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Acute | Exercise vs no treatment/usual care | MD ‐2.82, 95% CI ‐15.35 to 9.71 | 91 | 491 (3 RCTs) | Very low | Downgraded for study limitations, imprecision, and inconsistency |
| Function | Acute/subacute | McKenzie vs no treatment/usual care | MD ‐2.7 favouring McKenzie, 95% CI ‐7.5 to 2 | 0 | 328 (2 RCTs) | Low | Downgraded for study limitations and imprecision |
| Function | Subacute | Exercise (McKenzie) vs no treatment | MD ‐3.38, 95% CI ‐6.22 to ‐0.74 | ‐ | 194 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Function | Acute | Motor control added to usual care vs usual care | MD ‐2.4, 95% CI ‐4.87 to 0.07 | ‐ | 45 (1 RCT) | Very low | Downgraded for study limitations and imprecision |
| Function | Chronic | Exercise vs no treatment/usual care | MD ‐6.8, 95% CI ‐8.3 to ‐5.3 | 38 | 2942 (38 RCTs) | Moderate | Downgraded for study limitations |
| Function | Chronic | Motor control vs no treatment/usual care | MD ‐8.63, 95% CI ‐14.78 to ‐2.47 | 74 | 332 (5 RCTs) | Very low | Downgraded for study limitations, imprecision, and inconsistency |
| Function | Chronic | Pilates vs no treatment/usual care | MD ‐7.95, 95% CI ‐13.23 to ‐2.67 | 56 | 248 (5 RCTs) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Yoga vs no treatment/usual care | SMD ‐0.41, 95% CI ‐0.61 to ‐0.21 | 6 | 474 (8 RCTs) | Moderate | Downgraded for study limitations |
| Participant ratings of improvement (‐5 to 5 scale, 5 is complete improvement) |
Chronic | Motor control vs usual care | MD 1.3, 95% CI 0.3 to 2.3 | ‐ | 154 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Participant ratings of improvement (‐5 to 5 scale, 5 is complete improvement) |
Chronic | Pilates vs no treatment | MD 1.5, 95% CI 0.7 to 2.3 | ‐ | 86 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Quality of life (physical) |
Chronic | Yoga vs usual care/no treatment | SMD 0.5, 95% CI 0.05 to 0.95 | 0 | 81 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Quality of life (mental) |
Chronic | Yoga vs usual care/no treatment | SMD ‐0.15 95%, CI 1.24 to 0.93 | 67 | 81 (2 RCTs) | Very low | Downgraded for study limitations, imprecision, and inconsistency |
MD on a 0 to 100 scale unless otherwise indicated. We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; SMD: standardised mean difference
Appendix 11. Overview of reviews table with all outcomes (except pain intensity) comparing heat and cold therapy with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Heat and cold therapy for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function (0‐24 RMDQ) | Acute/subacute | Heated back wrap vs placebo (oral placebo) | MD ‐2.1, 95% CI ‐3.07 to ‐1.18 | 0 | 258 (2 RCTs) | Low | Downgraded for study limitations and imprecision |
MD on a 0 to 100 scale unless otherwise indicated.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RMDQ: Roland‐Morris Disability Questionnaire
Appendix 12. Overview of reviews table with all outcomes (except pain intensity) comparing manual therapy with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Manual therapy for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Acute/subacute | Massage vs no treatment/usual care (inert control) | SMD ‐0.5, 95% CI ‐1.06 to 0.06 | ‐ | 51 (1 RCT) | Very low | Downgraded for study quality and imprecision |
| Function | Acute/subacute | Spinal manipulation vs placebo (detuned electrical physical modality) | SMD ‐0.27, 95% CI ‐0.58 to 0.04 | ‐ | 178 (1 RCT) | Low | Downgraded for imprecision and inconsistency |
| Function | Acute/subacute | Spinal manipulation vs no treatment (inert control) | SMD ‐0.27, 95% CI ‐0.58 to 0.04 | ‐ | 178 (1 RCT) | Low | Downgraded imprecision and inconsistency |
| Function | Acute/subacute | Spinal manipulation vs sham manipulation | SMD ‐0.35, 95% CI ‐0.76 to 0.06 | ‐ | 94 (1 RCT) | Very low | Downgraded for study quality, imprecision, and inconsistency |
| Safety | Acute/subacute | Massage vs usual care | RD 0, 95% CI ‐0.07 to 0.07 | ‐ | 51 (1 RCT) | Low | Downgraded for study quality and imprecision |
| Participant ratings of improvement | Acute/subacute | Spinal manipulation vs placebo (detuned electrical physical modality) | RR 0.97, 95% CI 0.85 to 1.1 | ‐ | 239 (1 RCT) | Low | Downgraded imprecision and inconsistency |
| Function | Chronic | Spinal manipulation vs sham manipulation | SMD ‐0.45, 95% CI ‐0.97 to 0.06 | 55 | 55 (1 RCT) | Very low | Downgraded for study quality and imprecision |
| Function | Mixed | Massage vs no treatment (inactive control) | SMD ‐0.72, 95% CI ‐1.05 to ‐0.39 | 74 | 725 (6 RCTs) | Low | Downgraded for study quality and imprecision |
| Safety | Mixed | Massage vs no treatment (inactive control) | RD 0.06, 95% CI 0 to 0.11 | 72 | 624 (4 RCTs) | Very low | Downgraded for study quality and imprecision |
| Workplace participation | Chronic | Spinal manipulation vs placebo (detuned electrical physical modality) | RR 1.29 favouring spinal manipulation, 95% CI 1 to 1.65 | ‐ | 72 (1 RCT) | Very low | Downgraded for study quality and imprecision |
We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RD: risk difference; SMD: standardised mean difference
Appendix 13. Overview of reviews table with all outcomes (except pain intensity) comparing multimodal with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Multimodal for low back pain in adults compared to sham/placebo, no treatment or usual care ≤3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Workplace participation (sick leave) | Acute/subacute | Light physical conditioning vs usual care | SMD ‐0.18, 95% CI ‐0.45 to 0.10 | ‐ | 201 (1 RCT) | Very low | Downgraded for study limitations and imprecision |
| Workplace participation (return to work) | Acute/subacute | Intensive physical conditioning vs usual care | SMD ‐0.03, 95% CI ‐0.22 to 0.15 | 74 | 447 (3 RCTs) | Very low | Downgraded for study limitations, imprecision, and inconsistency |
| Function | Chronic | Multi‐component workplace interventions vs no treatment | MD ‐8.8, 95% CI ‐17.4 to ‐0.14 | ‐ | 27 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Function (0 to 24 RMDQ) |
Chronic | Sit‐stand workstation intervention vs no treatment | MD ‐0.4, 95% CI ‐2.7 to 1.90 | ‐ | 46 (1 RCT) | Low | Downgraded for study limitations and imprecision |
| Workplace participation (sick leave) | Chronic | Intensive physical conditioning vs usual care | SMD ‐0.23, 95% CI ‐0.42 to ‐0.03 | 0 | 1093 (5 RCTs) | Moderate | Downgraded for study limitations |
| Workplace participation (return to work) | Chronic | Light physical conditioning vs usual care | SMD ‐0.33 favouring light physical conditioning, 95% CI ‐0.67 to 0.02 | ‐ | 138 (1 RCT) | Low | Downgraded for imprecision and inconsistency |
MD on a 0 to 100 scale unless otherwise indicated. We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RMDQ: Roland‐Morris Disability Questionnaire; SMD: standardised mean difference
Appendix 14. Overview of reviews table with all outcomes (except pain intensity) comparing multidisciplinary therapies with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Multidisciplinary therapies for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of participants (trials) | Certainty of the evidence (GRADE) | Comments |
| Function | Acute/subacute | Multidisciplinary vs usual care | SMD ‐0.38, 95% CI ‐0.63 to ‐0.14 | 0 | 4 (272) | Very low | Downgraded for study limitations and imprecision |
| Function | Chronic | Multidisciplinary vs no treatment | SMD ‐0.49, 95% CI ‐0.76 to ‐0.22 | 0 | 3 (213) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Multidisciplinary vs usual care | SMD ‐0.41, 95% CI ‐0.62 to ‐0.19 | 58 | 9 (939) | Moderate | Downgraded for study limitations |
| Quality of life (physical) |
Chronic | Multidisciplinary vs usual care | MD 13.45, 95% CI ‐9.07 to 35.9 | 93 | 2 (144) | Very low | Downgraded for study limitations, imprecision, inconsistency |
| Quality of life (mental) |
Chronic | Multidisciplinary vs usual care | MD 15.25, 95% CI 2.05 to 28.44 | 73 | 2 (144) | Very low | Downgraded for study limitations, imprecision, inconsistency |
| Workplace participation | Chronic | Multidisciplinary vs usual care | OR 1.07, 95% CI 0.6 to 1.9 | 0 | 2 (373) | Moderate | Downgraded for study limitations |
MD on 0 to 100 scale unless otherwise indicated. We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial; SMD: standardised mean difference
Appendix 15. Overview of reviews table with all outcomes (except pain intensity) comparing psychological interventions with sham/placebo and no treatment/usual care at the short‐term follow‐up (≤ 3 months)
| Psychological interventions for low back pain in adults compared to sham/placebo, no treatment or usual care ≤ 3 months post‐intervention (short‐term) | |||||||
| Outcome | Low back pain duration | Intervention vs comparison | Relative effect, 95% CI | I2 (%) | Number of trials (participants) | Certainty of the evidence (GRADE) | Comments |
| Function | Chronic | Psychological interventions (progressive muscle relaxation and combination of behavioural approaches addressing fear, exercise, and self‐management) vs usual care | SMD ‐0.2, 95% CI ‐0.41 to 0.02 | 0 | 2 (330) | Moderate | Downgraded for imprecision |
| Function | Chronic | Cognitive approaches vs no treatment | SMD ‐0.15 favouring cognitive approaches, 95% CI ‐0.64 to 0.33 | 0 | 2 (68) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Combined approaches (cognitive behavioural) vs no treatment | SMD ‐0.15, 95% CI ‐0.64 to 0.33 | 49 | 4 (134) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Operant approaches vs no treatment | SMD ‐1.18, 95% CI ‐3.53 to 1.18 | 0 | 2 (87) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Respondent approaches (progressive relaxation) vs no treatment | SMD ‐0.88, 95% CI ‐1.36 to ‐0.39 | 0 | 3 (74) | Low | Downgraded for study limitations and imprecision |
| Function | Chronic | Respondent approaches (EEG feedback) vs no treatment | SMD ‐0.17, 95% CI ‐1.56 to 1.22 | 78 | 2 (44) | Very low | Downgraded for study limitations, imprecision, and inconsistency |
We have ordered the interventions in the table based on the following criteria: first, by pain duration (acute/subacute, chronic, and mixed); second, by outcomes (function, safety, and then secondary outcomes); third, by intervention label in alphabetical order; and finally, by comparison group (sham/placebo, no treatment/usual care). The absence of comparison indicates that no quantitative data were reported.
CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; SMD: standardised mean difference
Differences between protocol and review
One additional review author, Saurab Sharma, joined the review team to support data extraction and quality assessment.
We did not use the planned definition of a 10‐point reduction in pain intensity and disability as the minimally important difference. Instead, we used definitions from the American College of Physicians and the American Pain Society to improve the interpretation of the size of the effect (Chou 2017), without making judgements on what would be considered minimally important by an individual.
Contributions of authors
AGC, BMW, and JHM conceived the idea for an overview of non‐pharmacological treatments for low back pain. RRNR and AGC conducted the search and screening. RRNR, AGC, MCF, EOH and SS conducted data extraction, AMSTAR 2 and GRADE assessments. RRNR wrote the first draft, including drafting the tables and results of this overview. All authors reviewed and provided feedback on the draft versions and approved the final version published.
Sources of support
Internal sources
-
Institute for Work & Health, Canada
(ADF)
External sources
-
Australian Government Research Training Program, Australia
Postgraduate Research Scholarship (MKB, EO)
-
National Health and Medical Research Council, Australia
Research Fellowship (CGM), Program Grant (HL, CGM), Centre for Research Excellence (CGM)
-
Neuroscience Research Australia (NeuRA), Australia
PhD Candidature Supplementary Scholarship (AGC, RRNR, MKB, EO)
-
University of New South Wales, Australia
Research Excellence Award (MKB)
-
University of New South Wales Prince of Wales Clinical School, Australia
Postgraduate Research Scholarship (AGC)
Declarations of interest
CGM has received competitive grants from government agencies and industry to support his research. As an invited speaker at conferences, he has had his expenses covered and also received small gifts, such as a box of chocolates or a bottle of wine. He has received honoraria for marking theses, reviewing grants and preparing talks. CGM has published multiple papers on low back pain, some of which were referenced in the overview, and is on the Editorial Board of the Cochrane Back and Neck Review group. Mitigation of conflict of interest: CGM was not involved in editorial decisions on this review.
ADF was co‐ordinating Editor of the Cochrane Back and Neck Review group until December 2020. Mitigation of conflict of interest: ADF was not involved in decisions to approve or reject this review. She has published multiple papers on low back pain, some of which were referenced in the review. ADF has a monetised YouTube channel for people with chronic pain, including low back pain, and there are videos about non‐pharmacological and non‐surgical interventions. The institutions where ADF works have received various grants from external organisations, including government and public institutions in Ontario, Canada and the UK.
MWvT is on the Editorial Board of the Cochrane Back and Neck Review group. MWvT was co‐ordinating Editor until September 2017. Mitigation of conflict of interest: MWvT was not involved in editorial decisions on this review. MWvT has published multiple papers on low back pain, some of which were referenced in the review. MWvT has no additional competing interest; all research funding comes from non‐profit governmental funding agencies, and all funding (including travel and stay expenses) was paid to VU University.
HL has consulted for Cancer Council Australia and has received funding from the Australian Health and Medical Research Council (grant no. APP1126767), and Center for Effective Global Action (CEGA) & Berkeley Initiative for Transparency in the Social Sciences (BITSS).
BMW has received speaker fees for talks on non‐pharmacological care for people with low back pain.
SS is supported by the John J. Bonica Postdoctoral Fellowship from the International Association for the Study of Pain.
RRNR, AGC, MCF, EOH, and JHM have no known conflicts of interest.
This research project did not receive any specific grant from funding agencies in the public, commercial or not‐for‐profit sectors.
New
References
References to included reviews
Almeida 2023
- Almeida MO, Narciso Garcia A, Menezes Costa LC, Van Tulder MW, Lin CC, Machado LA. The McKenzie method for (sub)acute non-specific low back pain. Cochrane Database of Systematic Reviews 2023, Issue 4. Art. No: CD009711. [DOI: 10.1002/14651858.CD009711.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Dahm 2010
- Dahm KT, Brurberg KG, Jamtvedt G, Hagen KB. Advice to rest in bed versus advice to stay active for acute low-back pain and sciatica. Cochrane Database of Systematic Reviews 2010, Issue 6. Art. No: CD007612. [DOI: 10.1002/14651858.CD007612.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Ebadi 2020
- Ebadi S, Henschke N, Forogh B, Nakhostin Ansari N, Tulder MW, Babaei-Ghazani A, et al. Therapeutic ultrasound for chronic low back pain. Cochrane Database of Systematic Reviews 2020, Issue 7. Art. No: CD009169. [DOI: 10.1002/14651858.CD009169.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Engers 2008
- Engers A, Jellema P, Wensing M, Van der Windt DA, Grol R, Van Tulder MW. Individual patient education for low back pain. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No: CD004057. [DOI: 10.1002/14651858.CD004057.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Franke 2015
- Franke H, Fryer G, Ostelo RW, Kamper SJ. Muscle energy technique for non-specific low-back pain. Cochrane Database of Systematic Reviews 2015, Issue 2. Art. No: CD009852. [DOI: 10.1002/14651858.CD009852.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
French 2006
- French SD, Cameron M, Walker BF, Reggars JW, Esterman AJ. Superficial heat or cold for low back pain. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No: CD004750. [DOI: 10.1002/14651858.CD004750.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Furlan 2005
- Furlan AD, Van Tulder MW, Cherkin DC, Tsukayama H, Lao L, Koes BW, et al. Acupuncture and dry-needling for low back pain. Cochrane Database of Systematic Reviews 2005, Issue 1. Art. No: CD001351. [DOI: 10.1002/14651858.CD001351.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Furlan 2015
- Furlan AD, Giraldo M, Baskwill A, Irvin E, Imamura M. Massage for low-back pain. Cochrane Database of Systematic Reviews 2015, Issue 9. Art. No: CD001929. [DOI: 10.1002/14651858.CD001929.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hayden 2005
- Hayden JA, Van Tulder MW, Malmivaara A, Koes BW. Exercise therapy for treatment of non-specific low back pain. Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No: CD011674. [DOI: 10.1002/14651858.CD011674.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hayden 2021a
- Hayden JA, Ellis J, Ogilvie R, Malmivaara A, Van Tulder MW. Exercise therapy for chronic low back pain. Cochrane Database of Systematic Reviews 2021, Issue 9. Art. No: CD009790. [DOI: 10.1002/14651858.CD009790.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Henschke 2010
- Henschke N, Ostelo RW, Van Tulder MW, Vlaeyen JW, Morley S, Assendelft WJ, et al. Behavioural treatment for chronic low-back pain. Cochrane Database of Systematic Reviews 2010, Issue 7. Art. No: CD002014. [DOI: 10.1002/14651858.CD002014.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Kamper 2014
- Kamper SJ, Apeldoorn AT, Chiarotto A, Smeets RJ, Ostelo RW, Guzman J, et al. Multidisciplinary biopsychosocial rehabilitation for chronic low back pain. Cochrane Database of Systematic Reviews 2014, Issue 9. Art. No: CD000963. [DOI: 10.1002/14651858.CD000963.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Khadilkar 2008
- Khadilkar A, Odebiyi DO, Brosseau L, Wells GA. Transcutaneous electrical nerve stimulation (TENS) versus placebo for chronic low-back pain. Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No: CD003008. [DOI: 10.1002/14651858.CD003008.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Macedo 2016
- Macedo LG, Saragiotto BT, Yamato TP, Costa LO, Menezes Costa LC, Ostelo RW, et al. Motor control exercise for acute non-specific low back pain. Cochrane Database of Systematic Reviews 2016, Issue 2. Art. No: CD012085. [DOI: 10.1002/14651858.CD012085] [DOI] [PMC free article] [PubMed] [Google Scholar]
Marin 2017
- Marin TJ, Van Eerd D, Irvin E, Couban R, Koes BW, Malmivaara A, et al. Multidisciplinary biopsychosocial rehabilitation for subacute low back pain. Cochrane Database of Systematic Reviews 2017, Issue 6. Art. No: CD002193. [DOI: 10.1002/14651858.CD002193.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Mu 2020
- Mu J, Furlan AD, Lam WY, Hsu MY, Ning Z, Lao L. Acupuncture for chronic nonspecific low back pain. Cochrane Database of Systematic Reviews 2020, Issue 12. Art. No: CD013814. [DOI: 10.1002/14651858.CD013814] [DOI] [PMC free article] [PubMed] [Google Scholar]
Oltean 2014
- Oltean H, Robbins C, Van Tulder MW, Berman BM, Bombardier C, Gagnier JJ. Herbal medicine for low-back pain. Cochrane Database of Systematic Reviews 2014, Issue 12. Art. No: CD004504. [DOI: 10.1002/14651858.CD004504.pub4] [DOI] [PMC free article] [PubMed] [Google Scholar]
Parreira 2017
- Parreira P, Heymans MW, Van Tulder MW, Esmail R, Koes BW, Poquet N, et al. Back Schools for chronic non-specific low back pain. Cochrane Database of Systematic Reviews 2017, Issue 8. Art. No: CD011674. [DOI: 10.1002/14651858.CD011674.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Parry 2019
- Parry SP, Coenen P, Shrestha N, O'Sullivan PB, Maher CG, Straker LM. Workplace interventions for increasing standing or walking for decreasing musculoskeletal symptoms in sedentary workers. Cochrane Database of Systematic Reviews 2019, Issue 11. Art. No: CD012487. [DOI: 10.1002/14651858.CD012487.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Poquet 2016
- Poquet N, Lin CW, Heymans MW, Van Tulder MW, Esmail R, Koes BW, et al. Back schools for acute and subacute non-specific low-back pain. Cochrane Database of Systematic Reviews 2016, Issue 4. Art. No: CD008325. [DOI: 10.1002/14651858.CD008325.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Rubinstein 2011
- Rubinstein SM, Van Middelkoop M, Assendelft WJ, De Boer MR, Van Tulder MW. Spinal manipulative therapy for chronic low-back pain. Cochrane Database of Systematic Reviews 2011, Issue 2. Art. No: CD008112. [DOI: 10.1002/14651858.CD008112.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Rubinstein 2012
- Rubinstein SM, Terwee CB, Assendelft WJ, De Boer MR, Van Tulder MW. Spinal manipulative therapy for acute low-back pain. Cochrane Database of Systematic Reviews 2012, Issue 9. Art. No: CD008880. [DOI: 10.1002/14651858.CD008880.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Saragiotto 2016
- Saragiotto BT, Maher CG, Yamato TP, Costa LO, Menezes Costa LC, Ostelo RW, et al. Motor control exercise for chronic non-specific low-back pain. Cochrane Database of Systematic Reviews 2016, Issue 1. Art. No: CD012004. [DOI: 10.1002/14651858.CD012004] [DOI] [PMC free article] [PubMed] [Google Scholar]
Schaafsma 2013
- Schaafsma FG, Whelan K, Van der Beek AJ, Van der Es-Lambeek LC, Ojajarvi A, Verbeek JH. Physical conditioning as part of a return to work strategy to reduce sickness absence for workers with back pain. Cochrane Database of Systematic Reviews 2013, Issue 8. Art. No: CD001822. [DOI: 10.1002/14651858.CD001822.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Urrutia 2004
- Urrutia G, Burton AK, Morral A, Bonfill X, Zanoli G. Neuroreflexotherapy for non-specific low-back pain. Cochrane Database of Systematic Reviews 2004, Issue 2. Art. No: CD003009. [DOI: 10.1002/14651858.CD003009.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Van Duijvenbode 2008
- Van Duijvenbode IC, Jellema P, Van Poppel MN, Van Tulder MW. Lumbar supports for prevention and treatment of low back pain. Cochrane Database of Systematic Reviews 2008, Issue 2. Art. No: CD001823. [DOI: 10.1002/14651858.CD001823.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Walker 2010
- Walker BF, French SD, Grant W, Green S. Combined chiropractic interventions for low-back pain. Cochrane Database of Systematic Reviews 2010, Issue 4. Art. No: CD005427. [DOI: 10.1002/14651858.CD005427.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Wegner 2013
- Wegner I, Widyahening IS, Van Tulder MW, Blomberg SE, De Vet HC, Bronfort G, et al. Traction for low-back pain with or without sciatica. Cochrane Database of Systematic Reviews 2013, Issue 8. Art. No: CD003010. [DOI: 10.1002/14651858.CD003010.pub5] [DOI] [PMC free article] [PubMed] [Google Scholar]
Wieland 2022
- Wieland LS, Skoetz N, Pilkington K, Harbin S, Vempati R, Berman BM. Yoga for chronic non-specific low back pain. Cochrane Database of Systematic Reviews 2022, Issue 11. Art. No: CD010671. [DOI: 10.1002/14651858.CD010671.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Yamato 2015
- Yamato TP, Maher CG, Saragiotto BT, Hancock MJ, Ostelo RW, Cabral CM, et al. Pilates for low back pain. Cochrane Database of Systematic Reviews 2015, Issue 7. Art. No: CD010265. [DOI: 10.1002/14651858.CD010265.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Yousefi‐Nooraie 2008
- Yousefi-Nooraie R, Schonstein E, Heidari K, Rashidian A, Pennick V, Akbari-Kamrani M, et al. Low level laser therapy for nonspecific low-back pain. Cochrane Database of Systematic Reviews 2008, Issue 2. Art. No: CD005107. [DOI: 10.1002/14651858.CD005107.pub4] [DOI] [PMC free article] [PubMed] [Google Scholar]
References to excluded reviews
Assendelft 2013
- Assendelft WJ, Morton SC, Yu EI, Suttorp MJ, Shekelle PG. WITHDRAWN: Spinal manipulative therapy for low-back pain. Cochrane Database of Systematic Reviews 2013, Issue 1. Art. No: CD000447. [DOI: 10.1002/14651858.CD000447.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Bleakley 2012
- Bleakley C, McDonough S, Gardner E, Baxter GD, Hopkins JT, Davison GW. Cold-water immersion (cryotherapy) for preventing and treating muscle soreness after exercise. Cochrane Database of Systematic Reviews 2012, Issue 2. Art. No: CD008262. [DOI: 10.1002/14651858.CD008262.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Choi 2010
- Choi BK, Verbeek JH, Tam WW, Jiang JY. Exercises for prevention of recurrences of low-back pain. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No: CD006555. [DOI: 10.1002/14651858.CD006555.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Costello 2015
- Costello JT, Baker PR, Minett GM, Bieuzen F, Stewart IB, Bleakley C. Whole-body cryotherapy (extreme cold air exposure) for preventing and treating muscle soreness after exercise in adults. Cochrane Database of Systematic Reviews 2015, Issue 9. Art. No: CD010789. [DOI: 10.1002/14651858.CD010789.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Eccleston 2014
- Eccleston C, Fisher E, Craig L, Duggan GB, Rosser BA, Keogh E. Psychological therapies (Internet-delivered) for the management of chronic pain in adults. Cochrane Database of Systematic Reviews 2014, Issue 2. Art. No: CD010152. [DOI: 10.1002/14651858.CD010152.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Eccleston 2017
- Eccleston C, Fisher E, Thomas KH, Hearn L, Derry S, Stannard C, et al. Interventions for the reduction of prescribed opioid use in chronic non-cancer pain. Cochrane Database of Systematic Reviews 2017, Issue 11. Art. No: CD010323. [DOI: 10.1002/14651858.CD010323.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
French 2010
- French SD, Green S, Buchbinder R, Barnes H. Interventions for improving the appropriate use of imaging in people with musculoskeletal conditions. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No: CD006094. [DOI: 10.1002/14651858.CD006094.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Gadsby 2007
- Gadsby JG, Flowerdew MW. WITHDRAWN: Transcutaneous electrical nerve stimulation and acupuncture-like transcutaneous electrical nerve stimulation for chronic low back pain. Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No: CD000210. [DOI: 10.1002/14651858.CD000210.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hagen 2010
- Hagen KB, Hilde G, Jamtvedt G, Winnem M. WITHDRAWN: Bed rest for acute low-back pain and sciatica. Cochrane Database of Systematic Reviews 2010, Issue 6. Art. No: CD001254. [DOI: 10.1002/14651858.CD001254.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hayden 2019
- Hayden JA, Wilson MN, Riley RD, Iles R, Pincus T, Ogilvie R. Individual recovery expectations and prognosis of outcomes in non-specific low back pain: prognostic factor review. Cochrane Database of Systematic Reviews 2019, Issue 11. Art. No: CD011284. [DOI: 10.1002/14651858.CD011284.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Henschke 2013
- Henschke N, Maher CG, Ostelo RW, De Vet HC, Macaskill P, Irwig L. Red flags to screen for malignancy in patients with low-back pain. Cochrane Database of Systematic Reviews 2013, Issue 2. Art. No: CD008686. [DOI: 10.1002/14651858.CD008686.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Heymans 2004
- Heymans MW, Van Tulder MW, Esmail R, Bombardier C, Koes BW. Back schools for non-specific low-back pain. Cochrane Database of Systematic Reviews 2004, Issue 4. Art. No: CD000261. [DOI: 10.1002/14651858.CD000261.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hilde 2007
- Hilde G, Hagen KB, Jamtvedt G, Winnem M. WITHDRAWN: Advice to stay active as a single treatment for low-back pain and sciatica. Cochrane Database of Systematic Reviews 2007, Issue 2. Art. No: CD003632. [DOI: 10.1002/14651858.CD003632.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Jordan 2010
- Jordan JL, Holden MA, Mason EE, Foster NE. Interventions to improve adherence to exercise for chronic musculoskeletal pain in adults. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No: CD005956. [DOI: 10.1002/14651858.CD005956.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Karjalainen 2000
- Karjalainen K, Malmivaara A, Van Tulder M, Roine R, Jauhiainen M, Hurri H, et al. Multidisciplinary rehabilitation for fibromyalgia and musculoskeletal pain in working age adults. Cochrane Database of Systematic Reviews 2000, Issue 2. Art. No: CD001984. [DOI: 10.1002/14651858.CD001984] [DOI] [PMC free article] [PubMed] [Google Scholar]
Nnoaham 2014
- Nnoaham KE, Kumbang J. WITHDRAWN: Transcutaneous electrical nerve stimulation (TENS) for chronic pain. Cochrane Database of Systematic Reviews 2014, Issue 7. Art. No: CD003222. [DOI: 10.1002/14651858.CD003222.pub3] [DOI] [PubMed] [Google Scholar]
O'Connell 2018
- O'Connell NE, Marston L, Spencer S, DeSouza LH, Wand BM. Non-invasive brain stimulation techniques for chronic pain. Cochrane Database of Systematic Reviews 2018, Issue 3. Art. No: CD008208. [DOI: 10.1002/14651858.CD008208.pub4] [DOI] [PMC free article] [PubMed] [Google Scholar]
Sahar 2007
- Sahar T, Cohen MJ, Ne'eman V, Kandel L, Odebiyi DO, Lev I, et al. Insoles for prevention and treatment of back pain. Cochrane Database of Systematic Reviews 2007, Issue 4. Art. No: CD005275. [DOI: 10.1002/14651858.CD005275.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Verbeek 2011
- Verbeek JH, Martimo KP, Karppinen J, Kuijer PP, Viikari-Juntura E, Takala EP. Manual material handling advice and assistive devices for preventing and treating back pain in workers. Cochrane Database of Systematic Reviews 2011, Issue 6. Art. No: CD005958. [DOI: 10.1002/14651858.CD005958.pub3] [DOI] [PubMed] [Google Scholar]
Wieland 2017
- Wieland LS, Skoetz N, Pilkington K, Vempati R, D'Adamo CR, Berman BM. Yoga treatment for chronic non-specific low back pain. Cochrane Database of Systematic Reviews 2017, Issue 1. Art. No: CD010671. [DOI: 10.1002/14651858.CD010671.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Williams 2020
- Williams AC, Fisher E, Hearn L, Eccleston C. Psychological therapies for the management of chronic pain (excluding headache) in adults. Cochrane Database of Systematic Reviews 2020, Issue 8. Art. No: CD007407. [DOI: 10.1002/14651858.CD007407.pub4] [DOI] [PMC free article] [PubMed] [Google Scholar]
Additional references
Aas 2009
- Aas RW, Tuntland H, Holte KA, Røe C, Labriola M. Workplace interventions for low-back pain in workers. Cochrane Database of Systematic Reviews 2009, Issue 4. Art. No: CD008159. [DOI: 10.1002/14651858.CD008159] [CD008159] [DOI] [PMC free article] [PubMed] [Google Scholar]
Abbafati 2020
- Abbafati C, Machado DB, Cislaghi B, Salman OM, Karanikolos M, Mckee M, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;10258:1204-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ahern 2019
- Ahern M, Dean CM, Dear BF, Willcock SM, Hush JM. The experiences and needs of people seeking primary care for low-back pain in Australia. Pain Reports 2019;4(4):e756. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ahern 2020
- Ahern MM, Dean LV, Stoddart CC, Agrawal A, Kim K, Cook CE, et al. The effectiveness of virtual reality in patients with spinal pain: a systematic review and meta-analysis. Pain Practice 2020;20(6):656-75. [DOI] [PubMed] [Google Scholar]
Algafly 2007
- Algafly AA, George KP. The effect of cryotherapy on nerve conduction velocity, pain threshold and pain tolerance. British Journal of Sports Medicine 2007;41(6):365-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Almeida 2020
- Almeida L, Costa LO, Maher CG, Yamato TP, Fandim JV, Dear B, et al. Telerehabilitation for acute, subacute and chronic low back pain. Cochrane Database of Systematic Reviews 2020, Issue 8. Art. No: CD013704. [DOI: 10.1002/14651858.CD013704] [DOI] [Google Scholar]
Andersson 1995
- Andersson S, Lundeberg T. Acupuncture from empiricism to science: functional background to acupuncture effects in pain and disease. Medical Hypotheses 1995;45(3):271-81. [DOI] [PubMed] [Google Scholar]
Ashar 2021
- Ashar YK, Gordon A, Schubiner H, Uipi C, Knight K, Anderson Z, et al. Effect of pain reprocessing therapy vs placebo and usual care for patients with chronic back pain: a randomized clinical trial. JAMA Psychiatry 2022;79(1):13-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bagg 2018
- Bagg MK, Salanti SG, McAuley JH. Comparing interventions with network meta-analysis. Journal of Physiotherapy 2018;64:128-32. [DOI] [PubMed] [Google Scholar]
Bagg 2022
- Bagg MK, Wand BM, Cashin AG, Lee H, Hübscher M, Stanton TR, et al. Effect of graded sensorimotor retraining on pain intensity in patients with chronic low back pain: a randomized clinical trial. JAMA 2022;28(5):430-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Balagué 2012
- Balagué F, Mannion AF, Pellisé F, Cedraschi C. Non-specific low back pain Epidemiology and natural history. Lancet 2012;379(379):482-91. [DOI] [PubMed] [Google Scholar]
Balshem 2011
- Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, et al. GRADE guidelines: 3. Rating the quality of evidence. Journal of Clinical Epidemiology 2011;64:401-6. [DOI] [PubMed] [Google Scholar]
Bialosky 2009
- Bialosky JE, Bishop MD, Price DD, Robinson ME, George SZ. The mechanisms of manual therapy in the treatment of musculoskeletal pain: a comprehensive model. Manual Therapy 2009;14(5):531-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Binny 2019
- Binny J, Wong NLJ, Garga S, Lin C-WC, Maher CG, McLachlan AJ, et al. Transcutaneous electric nerve stimulation (TENS) for acute low back pain: systematic review. Scandinavian Journal of Pain 2019;19(2):225-33. [DOI] [PubMed] [Google Scholar]
Bishop 2015
- Bishop MD, Torres-Cueco R, Gay CW, Lluch-Girbés E, Beneciuk JM, Bialosky JE. What effect can manual therapy have on a patient's pain experience? Pain Management 2015;5(6):455-64. [DOI] [PMC free article] [PubMed] [Google Scholar]
Borrell‐Carrio 2004
- Borrell-Carrio F. The biopsychosocial model 25 years later: principles, practice, and scientific inquiry. Annals of Family Medicine 2004;2(6):576-82. [DOI] [PMC free article] [PubMed] [Google Scholar]
Braithwaite 2020
- Braithwaite FA, Walters JL, Moseley GL, Williams MT, McEvoy MP. Towards more homogenous and rigorous methods in sham-controlled dry needling trials: two Delphi surveys. Physiotherapy 2020;106:12-23. [DOI] [PubMed] [Google Scholar]
Buchbinder 2018
- Buchbinder R, Van Tulder M, Öberg B, Costa LM, Woolf A, Schoene M, et al. Low back pain: a call for action. Lancet 2018;391:2384-8. [DOI] [PubMed] [Google Scholar]
Burns 2016
- Burns JW. Mechanisms, mechanisms, mechanisms: it really does all boil down to mechanisms. Pain 2016;157(11):2393-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cashin 2021b
- Cashin AG, McAuley JH, Lamb SE, Lee H. Disentangling contextual effects from musculoskeletal treatments. Osteoarthritis and Cartilage 2021;29(3):297-9. [DOI] [PubMed] [Google Scholar]
Caspersen 1985
- Caspersen CJ, Powell KE, Christenson GM. Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Reports 1985;100(2):126-31. [PMC free article] [PubMed] [Google Scholar]
Chiarotto 2018
- Chiarotto A, Boers M, Deyo RA, Buchbinder R, Corbin TP, Costa LO, et al. Core outcome measurement instruments for clinical trials in nonspecific low back pain. Pain 2018;159(3):481-95. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chou 2017
- Chou R, Deyo R, Friedly J, Skelly A, Hashimoto R, Weimer M, et al. Nonpharmacologic therapies for low back pain: a systematic review for an American College of Physicians clinical practice guideline. Annals of Internal Medicine 2017;166:493-505. [DOI] [PubMed] [Google Scholar]
Chou 2018
- Chou L, Ranger TA, Peiris W, Cicuttini FM, Urquhart DM, Sullivan K, et al. Patients' perceived needs of health care providers for low back pain management: a systematic scoping review. Spine Journal 2018;18(4):691-711. [DOI] [PubMed] [Google Scholar]
Cieza 2020
- Cieza A, Causey K, Kamenov K, Hanson SW, Chatterji S, Vos T. Global estimates of the need for rehabilitation based on the Global Burden of Disease study 2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396(10267):2006-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
Clarke 2006
- Clarke J, Van Tulder M, Blomberg S, De Vet H, Van der Heijden G, Bronfort G. Traction for low back pain with or without sciatica: an updated systematic review within the framework of the Cochrane Collaboration. Spine 2006;31(14):1591-9. [DOI] [PubMed] [Google Scholar]
Coppieters 2008
- Coppieters MW, Butler DS. Do 'sliders' slide and 'tensioners' tension? An analysis of neurodynamic techniques and considerations regarding their application. Manual Therapy 2008;13(3):213-21. [DOI] [PubMed] [Google Scholar]
Costa 2012
- Costa LC, Maher C, Hancock M, McAuley J, Herbert R, Costa L. The prognosis of acute and persistent low back pain: a meta analysis. Canadian Medical Association Journal 2012;184(11):613-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
Covidence [Computer program]
- Covidence. Version accessed prior to 3 February 2025. Melbourne, Australia: Veritas Health Innovation, 2025. Available at covidence.org.
Da Silva 2017
- Da Silva T, Mills K, Brown BT, Herbert R, Maher C, Hancock M. Risk of recurrence of low back pain: A systematic review. Journal of Orthopaedic and Sports Physical Therapy 2017;47(5):305-13. [DOI] [PubMed] [Google Scholar]
Dieleman 2020
- Dieleman JL, Cao J, Chapin A, Chen C, Li Z, Liu A, et al. US health care spending by payer and health condition, 1996-2016. JAMA 2020;323(9):863-84. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dworkin 2005
- Dworkin RH, Turk DC, Farrar JT, Haythornthwaite JA, Jensen MP, Katz NP, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005;113:9-19. [DOI] [PubMed] [Google Scholar]
Edwards 2022
- Edwards RR, Schreiber KL, Dworkin RH, Turk DC, Baron R, Freeman R, et al. Optimizing and accelerating the development of precision pain treatments for chronic pain: IMMPACT review and recommendations. Journal of Pain 2022;22:00415-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Efthimiou 2016
- Efthimiou O, Debray TP, Valkenhoef G, Trelle S, Panayidou K, Moons KG, et al, GetReal Methods Review Group. GetReal in network meta-analysis: a review of the methodology. Research Synthesis Methods 2016;7(3):236-63. [DOI] [PubMed] [Google Scholar]
Foster 2018
- Foster NE, Anema JR, Cherkin D, Chou R, Cohen SP, Gross DP, et al. Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet 2018;391(10137):2368-83. [DOI] [PubMed] [Google Scholar]
Furlan 2009
- Furlan AD, Pennick V, Bombardier C, Van Tulder M, Editorial Board, Cochrane Back Review Group. 2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine 2009;34(18):1929-41. [DOI] [PubMed] [Google Scholar]
Furlan 2011
- Furlan AD, Yazdi F, Tsertsvadze A, Lao L, Sherman K, Ammendolia C, et al. Acupuncture for (sub)acute non‐specific low‐back pain. Cochrane Database of Systematic Reviews 2011;(8). [DOI: 10.1002/14651858.CD009265] [CD009265] [DOI]
Furlan 2015a
- Furlan AD, Malmivaara A, Chou R, Maher CG, Deyo RA, Schoene M, et al, Editorial Board of the Cochrane Back and Neck Group. 2015 updated method guideline for systematic reviews in the Cochrane Back and Neck Group. Cochrane Database of Systematic Reviews 2015;40(21):1660-73. [DOI] [PubMed] [Google Scholar]
Furlong 2020
- Furlong B, Aubrey-Bassler K, Etchegary H, Pike A, Darmonkow G, Swab M, et al. Patient education materials for non-specific low back pain and sciatica: a protocol for a systematic review and meta-analysis. BMJ Open 2020;10(9):e039530. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gao 2019
- Gao Y, Ge L, Ma X, Shen X, Liu M, Tian J. Improvement needed in the network geometry and inconsistency of Cochrane network meta-analyses: a cross-sectional survey. Journal of Clinical Epidemiology 2019;113:214-27. [DOI] [PubMed] [Google Scholar]
Garland 2020
- Garland EL, Brintz CE, Hanley AW, Roseen EJ, Atchley RM, Gaylord SA, et al. Mind-body therapies for opioid-treated pain: a systematic review and meta-analysis. JAMA Internal Medicine 2020;180(1):91-105. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gilanyi 2023
- Gilanyi YL, Wewege MA, Shah B, Cashin AG, Williams CM, Davidson SR, et al. Exercise increases pain self-efficacy in adults with nonspecific chronic low back pain: a systematic review and meta-analysis. Journal of Orthopaedic and Sports Physical Therapy 2023;0(6):1-8. [DOI] [PubMed] [Google Scholar]
Goldkuhle 2018
- Goldkuhle M, Narayan VM, Weigl A, Dahm P, Skoetz N. A systematic assessment of Cochrane reviews and systematic reviews published in high-impact medical journals related to cancer. BMJ Open 2018;8(3):e020869. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gopalakrishnan 2013
- Gopalakrishnan S, Ganeshkumar P. Systematic reviews and meta-analysis: Understanding the best evidence in primary healthcare. Journal Family Medicine and Primary Care 2013;2(1):9-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Grey 2020
- Grey A, Bolland MJ, Avenell A, Klein AA, Gunsalus CK. Check for publication integrity before misconduct. Nature 2020;577:167-9. [DOI] [PubMed] [Google Scholar]
Grooten 2022
- Grooten WJ, Boström C, Dedering Å, Halvorsen M, Kuster RP, Nilsson-Wikmar L, et al. Summarizing the effects of different exercise types in chronic low back pain - a systematic review of systematic reviews. BMC Musculoskeletal Disorders 2022;23(1):801. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hancock 2007
- Hancock MJ, Maher CG, Latimer J, Spindler MF, McAuley JH, Laslett M, et al. Systematic review of tests to identify the disc, SIJ or facet joint as the source of low back pain. European Spine Journal 2007;16(10):1539-50. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hansen 2010
- Hansen Z, Daykin A, Lamb SE. A cognitive-behavioural programme for the management of low back pain in primary care: a description and justification of the intervention used in the Back Skills Training Trial. Physiotherapy 2010;96(2):87-94. [DOI] [PubMed] [Google Scholar]
Hartvigsen 2018
- Hartvigsen J, Hancock MJ, Kongsted A, Louw Q, Ferreira ML, Genevay S, et al. What low back pain is and why we need to pay attention. Lancet 2018;391(10137):2356-67. [DOI] [PubMed] [Google Scholar]
Hayden 2021b
- Hayden JA, Ellis J, Ogilvie R, Stewart SA, Bagg MK, Stanojevic S, et al. Some types of exercise are more effective than others in people with chronic low back pain: a network meta-analysis. Journal of Physiotherapy 2021;67:252-62. [DOI: 10.1016/j.jphys.2021.09.004] [DOI] [PubMed] [Google Scholar]
Hennessy 2020
- Hennessy EA, Johnson BT. Examining overlap of included studies in meta-reviews: Guidance for using the corrected covered area index. Research Synthesis Methods 2020;11(1):134-45. [DOI] [PMC free article] [PubMed] [Google Scholar]
Henschke 2008
- Henschke N, Maher CG, Refshauge KM, Herbert RD, Cumming RG, Bleasel J, et al. Prognosis in patients with recent onset low back pain in Australian primary care: inception cohort study. BMJ (Clinical Research Ed.) 2008;337(jul07 1):a171. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2011
- Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al, Cochrane Bias Methods Group, Cochrane Statistical Methods Group. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ (Clinical Research Ed.) 2011;343:d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2024
- Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch V, editors. Cochrane Handbook for Systematic Reviews of Interventions Version 6.5 (updated August 2024). Cochrane, 2024. Available from https://training.cochrane.org/handbook.
Ho 2022
- Ho EK, Chen L, Simic M, Ashton-James CE, Comachio J, Wang DX, et al. Psychological interventions for chronic, non-specific low back pain: systematic review with network meta-analysis. BMJ (Clinical Research Ed.) 2022;376:e067718. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hodges 2003
- Hodges PW. Core stability exercise in chronic low back pain. Orthopedic Clinics of North America 2003;34(2):245-54. [DOI] [PubMed] [Google Scholar]
Hoffman 2007
- Hoffman BM, Papas RK, Chatkoff DK, Kerns RD. Meta-analysis of psychological interventions for chronic low back pain. Health Psychology 2007;26(1):1-9. [DOI] [PubMed] [Google Scholar]
Hoffmann 2014
- Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: Template for Intervention Description and Replication (TIDieR) checklist and guide. BMJ (Clinical Research Ed.) 2014;348:g1687. [DOI] [PubMed] [Google Scholar]
Hohenschurz‐Schmidt 2022
- Hohenschurz-Schmidt D, Draper-Rodi J, Vase L, Scott W, McGregor A, Soliman N, et al. Blinding and sham control methods in trials of physical, psychological, and self-management interventions for pain (article I): a systematic review and description of methods. Pain 2022;164(3):469-84. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hunt 2018
- Hunt H, Pollock A, Campbell P, Estcourt L, Brunton G. An introduction to overviews of reviews: planning a relevant research question and objective for an overview. Systematic Reviews 2018;7(1):1-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
IJzelenberg 2011
- IJzelenberg W, Rubinstein SM, Hayden J, Koes BW, Van Tulder MW. Exercise therapy for acute non‐specific low‐back pain. Cochrane Database of Systematic Reviews 2011, Issue 10. Art. No: CD009365. [DOI: 10.1002/14651858.CD009365] [DOI] [PMC free article] [PubMed] [Google Scholar]
Jesus‐Moraleida 2016
- Jesus‐Moraleida FR, Silva JP, Pereira DS, Domingues Dias JM, Correa Dias R, Ferreira ML, et al. Exercise therapy for older adults with low‐back pain. Cochrane Database of Systematic Reviews 2016;(4). [DOI: 10.1002/14651858.CD012140] [CD012140] [DOI]
Kaptchuk 2006
- Kaptchuk TJ, Stason WB, Davis RB, Legedza AR, Schnyer RN, Kerr CE, et al. Sham device v inert pill: randomised controlled trial of two placebo treatments. BMJ (Clinical Research Ed.) 2006;332(7538):391-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Karlsson 2020
- Karlsson M, Bergenheim A, Larsson ME, Nordeman L, Van Tulder M, Bernhardsson S. Effects of exercise therapy in patients with acute low back pain: a systematic review of systematic reviews. Systematic Reviews 2020;9(1):182. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kent 2023
- Kent P, Haines T, O'Sullivan P, Smith A, Campbell A, Schutze R, et al. Cognitive functional therapy with or without movement sensor biofeedback versus usual care for chronic, disabling low back pain (RESTORE): a randomised, controlled, three-arm, parallel group, phase 3, clinical trial. Lancet 2023;401:1866-77. [DOI] [PubMed] [Google Scholar]
Koes 2007
- Koes BW, Van Tulder MW, Peul WC. Diagnosis and treatment of sciatica. BMJ (Clinical Research Ed.) 2007;334(7607):1313-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lao 1996
- Lao L. Acupuncture techniques and devices. Journal of Alternative and Complementary Medicine (New York, N.Y.) 1996;2(1):23-5. [DOI] [PubMed] [Google Scholar]
Lee 2016
- Lee H, Mansell G, McAuley JH, Kamper SJ, Hübscher M, Moseley GL, et al. Causal mechanisms in the clinical course and treatment of back pain. Best Practice & Research Clinical Rheumatology 2016;30(6):1074-83. [DOI] [PubMed] [Google Scholar]
Lewis 2012
- Lewis SE, Holmes PS, Woby SR, Hindle J, Fowler NE. Short-term effect of superficial heat treatment on paraspinal muscle activity, stature recovery, and psychological factors in patients with chronic low back pain. Archives of Physical Medicine and Rehabilitation 2012;93(2):367-72. [DOI] [PubMed] [Google Scholar]
Lim 2019
- Lim YZ, Chou L, Au RT, Seneviwickrama KM, Cicuttini FM, Briggs AM, et al. People with low back pain want clear, consistent and personalised information on prognosis, treatment options and self-management strategies: a systematic review. Journal of Physiotherapy 2019;65(3):124-35. [DOI] [PubMed] [Google Scholar]
Liu 2013
- Liu H, Yao K, Zhang J, Li L, Wu T, Brox JI, et al. Sling exercise therapy for chronic low‐back pain. Cochrane Database of Systematic Reviews 2013, Issue 9. Art. No: CD010689. [DOI: 10.1002/14651858.CD010689] [DOI] [Google Scholar]
Machado 2012
- Machado L, Van Tulder MW, Lin CW, Clare H, Hayden JA. The McKenzie method for chronic non‐specific low‐back pain. Cochrane Database of Systematic Reviews 2012, Issue 3. Art. No: CD009712. [DOI: 10.1002/14651858.CD009712] [DOI] [Google Scholar]
Maher 2017
- Maher C, Underwood M, Buchbinder R. Non-specific low back pain. Lancet 2017;389(10070):736-47. [DOI] [PubMed] [Google Scholar]
McClurg 2016
- McClurg D, Pollock A, Campbell P, Hazelton C, Elders A, Hagen S, et al. Conservative interventions for urinary incontinence in women: an Overview of Cochrane systematic reviews. Cochrane Database of Systematic Reviews 2016, Issue 9. Art. No: CD012337. [DOI: 10.1002/14651858.CD012337] [DOI] [PMC free article] [PubMed] [Google Scholar]
Meader 2014
- Meader N, King K, Llewellyn A, Norman G, Brown J, Rodgers M, et al. A checklist designed to aid consistency and reproducibility of GRADE assessments: development and pilot validation. Systematic Reviews 2014;3(1):82. [DOI] [PMC free article] [PubMed] [Google Scholar]
Melzack 1981
- Melzack R. Myofascial trigger points: relation to acupuncture and mechanisms of pain. Archives of Physical Medicine and Rehabilitation 1981;62(3):114-7. [PubMed] [Google Scholar]
Milne 2001
- Milne S, Welch V, Brosseau L, Saginur M, Shea B, Tugwell P, et al. Transcutaneous electrical nerve stimulation (TENS) for chronic low back pain. Cochrane Database of Systematic Reviews 2001, Issue 2. Art. No: CD003008. [DOI: 10.1002/14651858.CD003008] [DOI] [PubMed] [Google Scholar]
Moore 2015
- Moore GF, Audrey S, Barker M, Bond L, Bonell C, Hardeman W, et al. Process evaluation of complex interventions: Medical Research Council guidance. BMJ (Clinical Research Ed.) 2015;350:h1258. [DOI] [PMC free article] [PubMed] [Google Scholar]
Moseley 2009
- Moseley AM, Elkins MR, Herbert RD, Maher CG, Sherrington C. Cochrane reviews used more rigorous methods than non-Cochrane reviews: survey of systematic reviews in physiotherapy [Cochrane reviews used more rigorous methods than non-Cochrane reviews: survey of systematic reviews in physiotherapy]. Journal of Clinical Epidemiology 2009;62(10):1021-30. [DOI] [PubMed] [Google Scholar]
Moseley 2015
- Moseley GL, Butler DS. Fifteen years of explaining pain: the past, present, and future. Journal of Pain 2015;16(9):807-13. [DOI] [PubMed] [Google Scholar]
NICE 2016
- National Institute for Health and Care Excellence (NICE). Low back pain and sciatica in over 16s: assessment and management. NICE guideline [NG59]; November 2016; last updated December 2020. https://www.nice.org.uk/guidance/ng59 (accessed prior to 5 March 2025).
NICE 2021
- National Institute for Health and Care Excellence (NICE). Chronic pain (primary and secondary) in over 16s: assessment of all chronic pain and management of chronic primary pain. NICE guideline [NG193]; April 2021. https://www.nice.org.uk/guidance/ng193 (accessed prior to 5 March 2025). [PubMed]
Nuesch 2010
- Nuesch E, Trelle S, Reichenbach S, Rutjes AW, Tschannen B, Altman DG, et al. Small study effects in meta-analyses of osteoarthritis trials: meta-epidemiological study [Small study effects in meta-analyses of osteoarthritis trials: meta-epidemiological study]. BMJ (Clinical Research Ed.) 2010;341:c3515. [DOI] [PMC free article] [PubMed] [Google Scholar]
O'Connell 2010
- O'Connell NE, Wand BM, Marston L, Spencer S, DeSouza LH. Non-invasive brain stimulation techniques for chronic pain. Cochrane Database of Systematic Reviews 2010, Issue 9. Art. No: CD008208. [DOI: 10.1002/14651858.CD008208.pub5] [DOI] [PubMed] [Google Scholar]
O'Connell 2013
- O'Connell NE, Wand BM, McAuley J, Marston L, Moseley GL. Interventions for treating pain and disability in adults with complex regional pain syndrome. Cochrane Database of Systematic Reviews 2013, Issue 4. Art. No: CD009416. [DOI: 10.1002/14651858.CD009416.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Odebiyi 2013
- Odebiyi DO, Henschke N, Ferreira ML, Tella A. Transcutaneous electrical nerve stimulation (TENS) for chronic low-back pain. Cochrane Database of Systematic Reviews 2013, Issue 4. Art. No: CD010500. [DOI: 10.1002/14651858.CD010500] [CD010500] [DOI] [Google Scholar]
Oliveira 2018
- Oliveira CB, Maher CG, Pinto RZ, Traeger AC, Lin CW, Chenot JF, et al. Clinical practice guidelines for the management of non-specific low back pain in primary care: an updated overview. European Spine Journal 2018;27(11):2791-803. [DOI] [PubMed] [Google Scholar]
Owen 2020
- Owen PJ, Miller CT, Mundell NL, Verswijveren SJ, Tagliaferri SD, Brisby H, et al. Which specific modes of exercise training are most effective for treating low back pain? Network meta-analysis. British Journal of Sports Medicine 2020;54:1279-87. [DOI] [PMC free article] [PubMed] [Google Scholar]
Parker 2022
- Parker L, Boughton S, Lawrence R, Bero L. Experts identified warning signs of fraudulent research: a qualitative study to inform a screening tool. Journal of Clinical Epidemiology 2022;151:1-17. [DOI] [PubMed] [Google Scholar]
Pascoe 2022
- Pascoe SC, Spoonemore SL Jr, Young JL, Rhon DI. Proposing six criteria to improve reproducibility of "usual care" interventions in back pain trials: a systematic review. Journal of Clinical Epidemiology 2022;149:227-35. [DOI] [PubMed] [Google Scholar]
Pieper 2014
Pincus 2013
- Pincus T, Holt N, Vogel S, Underwood M, Savage R, Walsh DA, et al. Cognitive and affective reassurance and patient outcomes in primary care: a systematic review. Pain 2013;154(11):2407-16. [DOI] [PubMed] [Google Scholar]
Poder 2019
- Poder TG, Beffarat M, Benkhalti M, Ladouceur G, Dagenais P. A discrete choice experiment on preferences of patients with low back pain about non-surgical treatments: identification, refinement and selection of attributes and levels. Patient Preference and Adherence 2019;13:933-40. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pérez‐Bracchiglione 2022
- Pérez-Bracchiglione J, Meza N, Bangdiwala SI, Niño de Guzmán E, Urrútia G, Bonfill X, et al. Graphical Representation of Overlap for OVErviews: GROOVE tool. Research Synthesis Methods 2022;13(3):381-8. [DOI] [PubMed] [Google Scholar]
Qaseem 2020
- Qaseem A, McLean RM, O'Gurek D, Batur P, Lin K, Kansagara DL, et al, Clinical Guidelines Committee of the American College of Physicians, Commission on Health of the Public and Science of the American Academy of Family Physicians. Nonpharmacologic and pharmacologic management of acute pain from non-low back, musculoskeletal injuries in adults: a clinical guideline from the American College of Physicians and American Academy of Family Physicians. Annals of Internal Medicine 2020;173:739-48. [DOI: 10.7326/M19-3602] [DOI] [PubMed] [Google Scholar]
Rainville 2004
- Rainville J, Hartigan C, Martinez E, Limke J, Jouve C, Finno M. Exercise as a treatment for chronic low back pain. Spine Journal 2004;4(1):106-15. [DOI] [PubMed] [Google Scholar]
Rubinstein 2019
- Rubinstein SM, De Zoete A, Van Middelkoop M, Assendelft WJ, De Boer MR, Van Tulder MW. Benefits and harms of spinal manipulative therapy for the treatment of chronic low back pain: systematic review and meta-analysis of randomised controlled trials. BMJ (Clinical Research Ed.) 2019;364:l689. [DOI] [PMC free article] [PubMed] [Google Scholar]
Santesso 2020
- Santesso N, Glenton C, Dahm P, Garner P, Akl EA, Alper B, et al, GRADE Working Group. GRADE guidelines 26: informative statements to communicate the findings of systematic reviews of interventions. Journal of Clinical Epidemiology 2020;119:126-35. [DOI] [PubMed] [Google Scholar]
Schulz 2010
- Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMC Medicine 2010;8(18):1-9. [DOI: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
Shea 2017
- Shea BJ, Reeves BC, Wells G, Thuku M, Hamel C, Moran J, et al. AMSTAR 2: A critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ (Clinical Research Ed.) 2017;358:j4008. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sluka 2018
- Sluka KA, Frey-Law L, Hoeger Bement A. Exercise-induced pain and analgesia? Underlying mechanisms and clinical translation. Pain 2018;159:S91-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Smith 2019
- Smith BE, Hendrick P, Bateman M, Holden S, Littlewood C, Smith TO, et al. Musculoskeletal pain and exercise - Challenging existing paradigms and introducing new. British Journal of Sports Medicine 2019;53(14):907-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sridhar 2021
- Sridhar S, Haas R, Docking S, Buchbinder R, O'Connor D. Models of care for managing non-specific low back pain. Cochrane Database of Systematic Reviews 2021, Issue 12. Art. No: CD015083. [DOI: 10.1002/14651858.CD015083] [DOI] [PMC free article] [PubMed] [Google Scholar]
Traeger 2015
- Traeger AC, Hübscher M, Henschke N, Moseley GL, Lee H, McAuley JH. Effect of primary care based education on reassurance in patients with acute low back pain systematic review and meta-analysis. JAMA Internal Medicine 2015;175(5):733-43. [DOI] [PubMed] [Google Scholar]
Useem 2015
- Useem J, Brennan A, LaValley M, Vickery M, Ameli O, Reinen N, et al. Systematic differences between Cochrane and non-Cochrane meta-analyses on the same topic: a matched pair analysis. PloS One 2015;10(12):e0144980. [DOI] [PMC free article] [PubMed] [Google Scholar]
Van Middelkoop 2011
- Van Middelkoop M, Rubinstein SM, Kuijpers T, Verhagen AP, Ostelo R, Koes BW, et al. A systematic review on the effectiveness of physical and rehabilitation interventions for chronic non-specific low back pain. European Spine Journal 2011;20(1):19-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
Van Tulder 1997
- Van Tulder MW, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions. Spine 1997;22(18):2128-56. [DOI] [PubMed] [Google Scholar]
Van Tulder 2003
- Van Tulder M, Furlan A, Bombardier C, Bouter L, Editorial Board of the Cochrane Collaboration Back Review Group. Updated method guidelines for systematic reviews in The Cochrane Collaboration Back Review Group. Spine 2003;28(12):1290-9. [DOI] [PubMed] [Google Scholar]
Van Tulder 2006
- Van Tulder M, Becker A, Bekkering T, Breen A, Real MT, Hutchinson A, et al. Chapter 3. European guidelines for the management of acute nonspecific low back pain in primary care. European Spine Journal 2006;15:169-91. [DOI: 10.1007/s00586-006-1071-2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Vance 2014
- Vance CG, Dailey DL, Rakel BA, Sluka KA. Using TENS for pain control: the state of the evidence. Pain Management 2014;4(3):197-209. [DOI] [PMC free article] [PubMed] [Google Scholar]
Widerström 2019
- Widerström B, Rasmussen-Barr E, Boström C. Aspects influencing clinical reasoning and decision-making when matching treatment to patients with low back pain in primary healthcare. Musculoskeletal Science Practice 2019;41:6-14. [DOI] [PubMed] [Google Scholar]
Wood 2019
- Wood L, Hendrick PA. A systematic review and meta-analysis of pain neuroscience education for chronic low back pain: short-and long-term outcomes of pain and disability. European Journal of Pain (London, England) 2019;23(2):234-49. [DOI] [PubMed] [Google Scholar]
Wun 2021
- Wun A, Kollias P, Jeong H, Rizzo RR, Cashin AG, Bagg MK, et al. Why is exercise prescribed for people with chronic low back pain? A review of the mechanisms of benefit proposed by clinical trialists. Musculoskeletal Science and Practice 2021;51:102307. [DOI] [PubMed] [Google Scholar]
References to other published versions of this review
Cashin 2021a
- Cashin AG, Rizzo RR, Wand BM, O'Connell NE, Lee H, Bagg MK, et al. Non-pharmacological and non-surgical treatments for low back pain in adults: an overview of Cochrane Reviews. Cochrane Database of Systematic Reviews 2021, Issue 8. Art. No: CD014691. [DOI: 10.1002/14651858.CD014691] [DOI] [PMC free article] [PubMed] [Google Scholar]

