Nonoperative treatment for lumbar spinal stenosis with neurogenic claudication

Abstract

Background

Lumbar spinal stenosis with neurogenic claudication is one of the most commonly diagnosed and treated pathological spinal conditions. It frequently afflicts the elderly population.

Objectives

To systematically review the evidence for the effectiveness of nonoperative treatment of lumbar spinal stenosis with neurogenic claudication.

Search methods

CENTRAL, MEDLINE, CINAHL, and Index to Chiropractic Literature (ICL) databases were searched up to June 2012.

Selection criteria

Randomized controlled trials published in English, in which at least one arm provided data on nonoperative treatments

Data collection and analysis

We used the standard methodological procedures expected by The Cochrane Collaboration. Risk of bias in each study was independently assessed by two review authors using the 12 criteria recommended by the Cochrane Back Review Group (Furlan 2009). Dichotomous outcomes were expressed as relative risk, continuous outcomes as mean difference or standardized mean difference; uncertainty was expressed with 95% confidence intervals. If possible a meta‐analysis was performed, otherwise results were described qualitatively. GRADE was used to assess the quality of the evidence.

Main results

From the 8635 citations screened, 56 full‐text articles were assessed and 21 trials (1851 participants) were included. There was very low‐quality evidence from six trials that calcitonin is no better than placebo or paracetamol, regardless of mode of administration or outcome assessed. From single small trials, there was low‐quality evidence for prostaglandins, and very low‐quality evidence for gabapentin or methylcobalamin that they improved walking distance. There was very low‐quality evidence from a single trial that epidural steroid injections improved pain, function, and quality of life, up to two weeks, compared with home exercise or inpatient physical therapy. There was low‐quality evidence from a single trial that exercise is of short‐term benefit for leg pain and function compared with no treatment. There was low and very low‐quality evidence from six trials that multimodal nonoperative treatment is less effective than indirect or direct surgical decompression with or without fusion. A meta‐analysis of two trials comparing direct decompression with or without fusion to multimodal nonoperative care found no significant difference in function at six months (mean difference (MD) ‐3.66, 95% CI ‐10.12 to 2.80) and one year (MD ‐6.18, 95% CI ‐15.03 to 2.66), but at 24 months a significant difference was found favouring decompression (MD ‐4.43, 95% CI ‐7.91 to ‐0.96).

Authors' conclusions

Moderate and high‐quality evidence for nonoperative treatment is lacking and thus prohibits recommendations for guiding clinical practice. Given the expected exponential rise in the prevalence of lumbar spinal stenosis with neurogenic claudication, large high‐quality trials are urgently needed.

Plain language summary

Non‐surgical treatment for spinal stenosis with leg pain

Review question

We reviewed the evidence on the effectiveness of non‐surgical treatments for people with leg pain caused by pressure on the nerves in the spine.

Background

If the spinal canal becomes narrow and presses on nerves (spinal stenosis) it can cause pain in the legs or buttocks (neurogenic claudication). Spinal stenosis is treated with a variety of non‐surgical methods, including painkillers and other medications, injections into the spine, exercise, physiotherapy and similar treatments. We wanted to discover whether using non‐surgical methods was better or worse than other alternatives.

Study characteristics

We included 21 randomised controlled trials that compared non‐surgical treatments to placebo, no treatment, or to surgery. All the participants had leg pain and a confirmed diagnosis of lumbar stenosis. There were 1851 people, with an average age of 50 years, equally divided between men and women. The follow‐up time ranged from one week to six years. The evidence was current to June 2012.

Key results

Overall, the review suggests that surgery is more effective at relieving pain than non‐surgical treatments.

Medicines taken by mouth. One small trial each of prostaglandins (compared to another medicine), gabapentin (compared to placebo), and vitamin B1 (compared to a variety of treatments) suggested improvements in pain and walking distance. Some digestive problems were reported with both medicines in the prostaglandins trial; and some people in the gabapentin trial reported dizziness or sleepiness.

Epidural injections. Two small trials showed short‐term improvements in pain and quality of life (up to two weeks) and two showed no difference compared with placebo injections. No trial reported bad reactions or problems.

Calcitonin injections. Six small trials indicated that calcitonin is no better than paracetemol or placebo. A number of people reported feeling sick or developed a rash.

Mixed approaches compared to surgery. Five trials compared results from surgery with results from a variety of non‐surgical treatments. One trial found that after two years there was no difference between treatments as far as pain was concerned. The other four found that surgery improved pain more than non‐surgical treatments over different time periods, but not necessarily walking ability. Between 5% and 18% suffered unwanted side effects from surgery, some serious.

Physical therapy. Four small trials, all including some form of exercise, failed to demonstrate improved walking ability after physical therapy. One trial each suggested that exercise is better than no treatment for leg pain, and that treadmill walking and stationary cycling produce similar, limited results.

Quality of evidence

The findings of all the trials in this study were based on low or very low‐quality evidence. The studies were poorly designed, or they failed to give enough information about what was done. This means that we cannot be confident that the findings are reliable, and further research may well come to a different conclusion.

Background

Description of the condition

Lumbar spinal stenosis (LSS) is characterized by bilateral or unilateral buttock, thigh, or calf discomfort; pain; or weakness precipitated by walking and prolonged standing (Comer 2009). The pathophysiology is thought to be compression or ischemia, or both, of the lumbosacral nerve roots due to narrowing of the lateral and central vertebral canals, usually as a consequence of osteoarthritic thickening of the articulating facet joints, infolding of the ligamentum flavum, and degenerative bulging of the intervertebral discs (Comer 2009; Katz 2008). Neurogenic claudication can have a significant impact on functional ability, quality of life, and independence in the elderly. Those afflicted have greater walking limitations than individuals with knee or hip osteoarthritis (Winter 2010). Lumbar spinal stenosis is the most common reason for spine surgery among individuals older than 65 years (Deyo 2010). The incidence of new cases of neurogenic claudication is expected to rise dramatically over the next 20 years when an estimated 23% to 25% of the population will be older than 65 years (Government of Canada 2006). This will have a significant impact on healthcare resources in the near future.

Description of the intervention

Most patients who seek care for neurogenic claudication are treated nonoperatively. A course of conservative treatment is also recommended prior to surgical intervention (AHRQ 2001). However, what constitutes effective conservative or nonoperative treatment is unknown (AHRQ 2001; Chou 2007). The purpose of this review was to systematically evaluate the clinical effectiveness of nonoperative treatments of lumbar spinal stenosis with neurogenic claudication.

Various nonoperative treatments are available to patients with neurogenic claudication including epidural injections (with steroid or anaesthetic, or both), oral medications (such as non‐steroidal anti‐inflammatory drugs (NSAIDs), analgesics, muscle relaxants, prostaglandins, and neuropathic drugs), vitamin B12, nasal or intramuscular calcitonin, physical therapies (such as exercise therapy, orthoses, electrical modalities, and traction), and manual therapy. Patients receiving nonoperative care often receive a combination of these treatments.

How the intervention might work

The mechanism of action of the various treatments differs. Epidural steroid injections, calcitonin, and NSAIDS aim to reduce nerve root inflammation (Fukusaki 1988; Podichetty 2004). Analgesics, epidural anaesthetic injections, and prostaglandins act to block pain transmission and increase neural blood flow (Fukusaki 1988; Matsudaira 2009). Vitamin B12 is thought to increase nerve root blood flow (Waikakul 2000) and neuropathic medications act to block pain transmission. Physical and manual therapies aim to reduce pain and maximize function by improving lumbar spine and lower extremity flexibility, muscular strength, and endurance (Whitman 2006).

Why it is important to do this review

Lumbar spinal stenosis causing neurogenic claudication is a leading cause of pain, disability, and lost independence in people over the age of 65 years. With the aging population the number of people who suffer from this condition is expected to grow exponentially. The vast majority of people with lumbar spinal stenosis causing neurogenic claudication receive nonoperative treatments. However, what constitutes effective nonoperative treatment is unknown (AHRQ 2001; Chou 2007). This review is an update of a previous review on this topic (Ammendolia 2012).

Objectives

To assess the effects of nonoperative treatments for lumbar spinal stenosis with neurogenic claudication.

Methods

Criteria for considering studies for this review

Types of studies

Studies were included if they were randomized controlled trials (RCTs) published in English, at least one arm of the trial provided data on the effectiveness of a nonoperative treatment, and at least 80% of participants had neurogenic claudication with lumbar spinal stenosis confirmed by imaging. Studies evaluating participants with radiculopathy due to disc lesions were excluded. Studies with mixed populations were included only if separate data for participants with neurogenic claudication due to lumbar spinal stenosis were provided.

Types of participants

Neurogenic claudication was defined as buttock or leg pain or aching, numbness, tingling, weakness, or fatigue with or without back pain precipitated by standing or walking. Studies evaluating participants with radiculopathy due to disc lesions were excluded.

Types of interventions

Any study that included at least one treatment arm that employed nonoperative treatment, including but not limited to oral and injection‐based medications, manual therapies, and physical therapies.

Nonoperative treatments were compared to placebo, no treatment, other nonoperative treatments, or surgical interventions.

Types of outcome measures

Studies were included if they measured at least one of the following outcomes: walking ability, pain intensity, function, quality of life, and global improvement. Common outcome measures used in this patient population included the following.

The Zurich Claudication Questionnaire (ZCQ), also known as the Swiss Spinal Stenosis questionnaire, is a validated self report, condition specific questionnaire for patients with neurogenic claudication (Comer 2011). It measures symptom intensity, function, and satisfaction with treatment. Walking ability is often measured using subscales of the Oswestry Disability Index (Fairbank 1980) or the Zurich Claudication Questionnaire (Stucki 1996), or objectively using the Self Paced Walking Test, Treadmill Walking Test, and distance walked. Pain intensity is usually measured using a visual analog scale (VAS), verbal response scale (VRS), McGill pain questionnaire (Melzack 2005), or numerical pain rating scale (NPRS). Function in patients with neurogenic claudication can be measured using the Oswestry Disability Index or the Roland Morris Disability Questionnaire (Brouwer 2004). Quality of life can be measured using the Short Form‐36 (SF‐36) questionnaire (Ware 1995). A subjective global rating of change or improvement can also be employed, typically in the form of a Likert scale.

Outcomes were categorized according to these follow‐up periods: immediate (up to one week), short‐term (between one week and three months), intermediate (between three months and one year), and long‐term (one year or longer).

Search methods for identification of studies

Electronic searches

An electronic search was performed by an experienced librarian from the Cochrane Back Review Group in CENTRAL (2012, Issue 6), MEDLINE (1966 to June 2012), EMBASE (1980 to June 2012), CINAHL (1982 to June 2012), and Index to Chiropractic Literature (1985 to June 2012). The terms “spinal stenosis”, “lumbar spinal stenosis”, “neurogenic claudication”, “lumbar radicular pain”, “cauda equina”, and “spondylosis” were combined with a highly sensitive search strategy to identify RCTs. See Appendix 1 for more details.

Data collection and analysis

Selection of studies

Two review authors (CA and KS) independently screened all titles and abstracts identified by the search strategy. The full‐texts of articles deemed to be potentially relevant were independently assessed by two review authors, who made the final decision for inclusion. A third review author was consulted if consensus was not reached.

Data extraction and management

Two review authors independently performed data extraction using a standardized form. The data included characteristics of patients and treatments, and outcomes. Safety data (adverse effects and complications) were also collected when available.

Assessment of risk of bias in included studies

Two review authors independently assessed risk of bias. Risk of bias was assessed using the 12‐item criteria recommended by the Cochrane Back Review Group (Furlan 2009). See Appendix 2. The criteria were scored as 'high', 'low', or 'unclear' and were reported in the 'Risk of bias' table. A study was considered at low risk of bias if it fulfilled six or more of the 12 criteria, including clearly described and appropriate randomization and allocation concealment, and with no severe flaws. A severe flaw was defined a priori as a serious methodological deficiency not captured by the 12‐item criteria that significantly increases the risk of bias, such as very high dropout or crossover rates. Discrepancies in risk of bias scoring and data extraction were discussed during a consensus meeting.

Data synthesis

Dichotomous outcomes were analysed by calculating the relative risk (RR). Continuous outcomes were analysed by calculating the mean difference (MD) when the same instrument was used to measure outcomes, or the standardized mean difference (SMD) when different instruments were used to measure the outcomes. The uncertainty was expressed with 95% confidence intervals (95% CI). The outcome measures from the individual trials were combined through meta‐analysis where possible (clinical comparability of population, intervention, and outcomes between trials) using a fixed‐effect model unless there was significant statistical heterogeneity, in which case a random‐effects model was used. A P value less than 0.05 of the Chi test indicates significant statistical heterogeneity.

If a meta‐analysis was not possible, the results from clinically comparable trials were described qualitatively in the text.

Regardless of whether there were sufficient data available to use quantitative analyses to summarize the data, we assessed the overall quality of the evidence for each outcome. To accomplish this, we used the GRADE approach as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and adapted in the updated Cochrne Back Review Group (CBRG) method guidelines (Furlan 2009). Factors that may decrease the quality of the evidence are: study design and risk of bias, inconsistency of results, indirectness (not generalisable), imprecision (sparse data), and other factors (for example reporting bias). The quality of the evidence for a specific outcome was reduced by a level according to the performance of the studies against the five factors.

High quality evidence: there are consistent findings among at least 75% of RCTs with low risk of bias; consistent, direct and precise data; and no known or suspected publication biases. Further research is unlikely to change either the estimate or our confidence in the results.

Moderate quality evidence: one of the domains is not met. Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

Low quality evidence: two of the domains are not met. Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

Very low quality evidence: three of the domains are not met. We are very uncertain about the results.

No evidence: no RCTs were identified that addressed this outcome.

Results

Description of studies

A total of 21 RCTs met the inclusion criteria and were included in the review (Figure 1). Table 1 describes the characteristics of the included trials. A total of 1851 participants (926 men and 925 women) were randomized to 1 of 23 comparison groups. Nineteen trials were conducted at tertiary care centers and two at medical or rehabilitation clinics (Goren 2010; Whitman 2006). The mean age of participants was more than 50 years in all but two trials, in which the mean age was just less than 50 years (Cuckler 1985; Zahaar 1991). The duration of symptoms varied considerably among the studies, with a mean range of 12 weeks to 15 years. Follow‐up periods also varied significantly. Three of the four studies evaluating epidural injections provided follow‐up data up to a week after the injections (Cuckler 1985; Fukusaki 1988; Zahaar 1991), whereas all studies comparing multimodal nonoperative care with surgery provided long‐term follow‐up (at least two‐year) outcome data (Amundsen 2000; Malmivaara 2007; Weinstein 2007; Weinstein 2008; Zucherman 2004).

1.

Selection process of included articles.

1. Nonoperative interventions for lumbar spinal stenosis with neurogenic claudication: a summary of GRADE assessment and outcomes.

Comparisons and trials	Risk of bias	GRADE assessment and outcomes/measures (walking ability, pain, function, and quality of life) at each follow‐up point
		Consistency	Directness	Precision	Selective reporting	Immediate	Short‐term	Intermediate	Long‐term	Quality of the evidence
Calcitonin
Calcitonin injection versus placebo injection
Eskola 1992	High High	No No	Yes Yes	No No	Yes Yes		= TWT = VAS	= TWT = VAS	= TWT = VAS	+000 +000
Porter 1983	High	No	Yes	No	Yes		? distance walked	? distance walked		+000
Porter 1988	High High	No No	Yes Yes	No No	Yes Yes		= distance walked = VAS			+000 +000
Calcitonin nasal spray versus placebo nasal spray
Podichetty 2004	High High High High	No No No No	Yes Yes Yes Yes	No No No No	Yes		= distance walked = time walked = SF‐36 = VAS			+000 +000 +000 +000
Tafazal 2007	High High High High High	No No No No No	Yes Yes Yes Yes Yes	No No No No No			= Shuttle walk = VAS leg = VAS back = ODI = global			+000 +000 +000 +000 +000
Calcitonin nasal spray plus physical therapy versus paracetamol plus physical therapy
Sahin 2009	High High High	No No No	Yes Yes Yes	No No No			= distance walked = VAS = RMDI			+000 +000 +000
Oral medication
Oral prostaglandin versus etodolac (NSAID)
Matsudaira 2009	Low Low Low Low Low	No No No No No	Yes Yes Yes Yes Yes	No No No No No	Yes		> distance walked ? SF‐36 = LBP > Leg pain > global			++00 +000 ++00 ++00 ++00
Methylocobalamin (vitamin B12) plus conservative care versus conservative care
Waikakul 2000	High	No	Yes	No	Yes			> distance walked	> distance walked	+000
Gabapentin versus placebo
Yaski 2007	High High High	No No No	Yes Yes Yes	No No No				> distance walked = VAS (1‐2 months) > VAS (3 months)	> distance walked > VAS	+000 +000 +000
Physical therapy
Exercise plus ultrasound versus exercise plus sham ultrasound
Goren 2010	Low Low Low Low	No No No No	Yes Yes Yes Yes	No No No No			= TWT = VAS back = VAS leg = ODI			++00 ++00 ++00 ++00
Exercise plus ultrasound versus no treatment
Goren 2010	Low Low Low Low	No No No No	Yes Yes Yes Yes	No No No No			= TWT = VAS back > VAS leg > ODI			++00 ++00 ++00 ++00
Exercise plus sham ultrasound versus no treatment
Goren 2010	Low Low Low Low	No No No No	Yes Yes Yes Yes	No No No No			= TWT = VAS back > VAS leg > ODI			++00 ++00 ++00 ++00
Inpatient physical therapy versus home exercise program plus oral diclofenac
Koc 2009	High High High High	No No No No	Yes Yes Yes Yes	No No No No	Yes		= TWT > VAS > RMDI > NHP	= TWT = VAS = RMDI = NHP		+000 +000 +000 +000
Unweighted treadmill walking plus exercise versus cycling plus exercise
Pua 2007	Low Low Low Low Low	No No No No No	Yes Yes Yes Yes Yes	No No No No No			= distance walked = ODI = RMDI = VAS = global			++00 ++00 ++00 ++00 ++00
Manual therapy, exercise and unweighted treadmill versus flexion exercise, walking and sham ultrasound
Whitman 2006	High High High High	No No No No	Yes Yes Yes Yes	No No No No			= TWT > global = ODI = NPRS		= global	+000 +000 +000 +000
Epidural injection
Translaminar epidural steroid injections versus placebo injections
Cuckler 1985	High	No	Yes	No		= global			=global	+000
Translaminar epidural steroids plus epidural block versus placebo injections
Fukusaki 1988	High	No	Yes	No		> distance walked	= distance walked			+000
Translaminar epidural steroids plus epidural block versus epidural block injections
Fukusaki 1988	High	No	Yes	No		= distance walked	= distance walked			+000
Translaminar epidural block versus placebo
Fukusaki 1988	High	No	Yes	No		> distance walked	= distance walked			+000
Intralaminar epidural steroid plus epidural block versus home exercise program plus oral diclofenac
Koc 2009	High High High High	No No No No	Yes Yes Yes Yes	No No No No	Yes Yes Yes Yes		= TWT > VAS > RMDI > NHP	= TWT = VAS = RMDI = HNP		+000 +000 +000 +000
Intralaminar epidural steroid plus epidural block versus inpatient physical therapy
Koc 2009	High High High High	No No No No	Yes Yes Yes Yes	No No No No	Yes Yes Yes Yes		= TWT > VAS > RMDI > NHP	= TWT = VAS = RMDI = HNP		+000 +000 +000 +000
Caudal epidural steroids versus placebo injections
Zahaar 1991	High	No	Yes	No		= global			= global	+000
Multimodal nonoperative treatment
Indirect decompression using interspinous spacer (X‐Stop) versus multimodal nonoperative care for degenerative spondylolisthesis
Zucherman 2004, Anderson 2006	High	No	Yes	No	SR (short & intermediate)		> ZCQ	> ZCQ	> ZCQ	+000
Indirect decompression using interspinous spacer (X_Stop) versus multimodal nonoperative care
Zucherman 2004, 2005, Hsu 2006	High	No No	Yes Yes	No No	SR SR		> ZCQ > SF‐36	> ZCQ > SF‐36	> ZCQ > SF‐36	+000 +000
Direct decompression ± fusion versus multimodal nonoperative care for degenerative spondylolisthesis
Weinstein 2007	High High High High High	No No No No No	Yes Yes Yes Yes Yes	No No No No No			= SF‐36 = ODI = LBPBS = LPBI = SBS	= SF‐36 = ODI = LBPBS = LPBI = SBS	= SF‐36 = ODI = LBPBS = LPBI = SBS	+000 +000 +000 +000 +000
Direct decompression ± fusion versus multimodal nonoperative care
Amundsen 2000	High High	No No	Yes Yes	No No			?* pain severity	?* global	?* pain severity ? global	+000 +000
Malmivaara 2007	Low Low Low Low Low Low	No No No No No No	Yes Yes Yes Yes Yes Yes	No No No No No No				= TWT = SW > VAS leg walk > VAS leg > VAS  LB ODI (see Figure 4)	= TWT = SW > VAS leg walk > VAS leg > VAS  LB ODI (see Figure 4)	++00 ++00 ++00 ++00 ++00 +000
Weinstein 2008	High High High High High High	No No No No No No	Yes Yes Yes Yes Yes Yes	No No No No No No			= SF‐36 BP = SF‐36 PF = LBPBS = LPBI = SBS = ODI	= SF‐36 BP = SF‐36 PF = LBPBS = LPBI = SBS ODI (see Figure 4)	> SF‐36 BP** = SF‐36  PF = LBPBS = LPBI = SBS ODI (Figure 4)	+000 +000 +000 +000 +000 +000

Follow‐up points: Immediate = up to 1 week, short‐term = >1 week ‐ 3 months, intermediate = 3 months ‐ 1 year, long‐term = > 1 year

> statistically significant favouring intervention (first comparison), < statistically significant favouring control (second comparison), = no statistically significant difference between intervention and control groups

TWT = Treadmill Walking Test, VAS = Visual Analog Scale for Pain Intensity, RMDI = Roland‐Morris Back Disability Index, NHP = Nottingham Health Profile, Global = Patient Perceived Improvement, SR = Selective Reporting, ODI = Oswestry Back Disability Index, ? = insufficient data, LBP = Low back Pain Severity Scale, Leg pain = Leg Pain Severity Scale, ? SF‐36 = No data on overall score, improvement in some subscales, NPRS = Numeric Pain Rating Scale, SF‐36 BP = SF‐36 Bodily Pain Subscale, SF‐36‐ PF = SF‐36 Physical Function Subscale, LBPBS ‐ Low Back Pain Bothersome Scale, LPBI = Leg Pain Bothersome Index, SBS ‐ Stenosis Bothersome Scale, SW = Subjective Walking, VAS leg = Visual Analog Scale for Leg Pain, VAS LB = Visual Analog Scale for Low Back Pain, VAS leg walking = Visual Analog Scale for Leg pain while walking, ?* = no between group statistical comparisons, ** = SF‐36 BP significantly better at 2 years but not 4 years.

GRADE evidence: +000 = Very low quality evidence, ++00 = Low quality evidence, +++0 = Moderate quality evidence, ++++ = High quality evidence

Results of the search

We obtained 8635 records that were screened, and 56 articles were eligible for full‐text review.

Included studies

Types of studies: all studies were randomized controlled trials (RCTs)

Study population: all participants had clinical findings and imaging confirmed lumbar spinal stenosis with neurogenic claudication

Technique: type, practitioner, number and duration of treatments varied

Outcome measures, follow‐up and safety: there was a wide variation in outcomes measures used, follow‐up, and safety profiles

Excluded studies

Reasons for exclusion included publication duplication (n = 1), studies that were not RCTs (n = 16), studies that lacked imaging confirmation (n = 2), studies that did not have neurogenic claudication as an inclusion criterion (n = 9), articles not published in English (n = 3), and studies that did not report results for spinal stenosis patients separately (n = 4).

Risk of bias in included studies

Although 13 studies met six or more criteria for risk of bias, only four studies were considered to have low risk of bias (Goren 2010; Malmivaara 2007; Matsudaira 2009; Pua 2007). Specifically, among the remaining nine studies that met six or more criteria, seven failed to explicitly describe or use an appropriate randomization procedure, allocation concealment, or both (Cuckler 1985; Fukusaki 1988; Podichetty 2004; Tafazal 2007; Whitman 2006; Zahaar 1991; Zucherman 2004); and two had severe flaws due to high crossover rates, which made the intention‐to‐treat analyses uninterpretable (Weinstein 2007; Weinstein 2008). Other common sources of risk of bias included failure to blind the participants receiving the intervention or control (Amundsen 2000; Fukusaki 1988; Goren 2010; Koc 2009; Malmivaara 2007; Matsudaira 2009; Porter 1983; Sahin 2009; Waikakul 2000; Weinstein 2007; Weinstein 2008; Whitman 2006; Yaski 2007; Zucherman 2004), failure to blind the treating healthcare provider (Amundsen 2000; Fukusaki 1988; Goren 2010; Koc 2009; Malmivaara 2007; Matsudaira 2009; Porter 1983; Sahin 2009; Waikakul 2000; Weinstein 2007; Weinstein 2008; Whitman 2006; Yaski 2007; Zucherman 2004), and selective reporting (Eskola 1992; Koc 2009; Matsudaira 2009; Porter 1983; Porter 1988; Waikakul 2000). No studies reported co‐interventions. See Figure 2 and Figure 3.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

3.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Effects of interventions

Nineteen of the 23 comparisons were examined in a single trial, most with small sample sizes. It was possible to combine data from only two trials for one outcome in a meta‐analysis (Malmivaara 2007; Weinstein 2008) (Analysis 1.1). Heterogeneity in source population, intervention, and outcome instruments precluded pooling of data for the other trials. Table 1 summarizes the quality of the evidence for the outcomes for each comparison. The results below are reported on the basis of statistically significant differences between comparators for each outcome.

1.1. Analysis.

Comparison 1 Direct Decompression ± fusion versus multimodal nonoperative care for Oswestry Disability Index, Outcome 1 Oswestry Disability Index.

Oral medication

There was low‐quality evidence based on one trial (N = 79) that prostaglandins improved walking distance and leg pain in the short‐term compared with etodolac (an NSAID), with 5% of participants in both groups reporting gastrointestinal upset (Matsudaira 2009). A small trial evaluating gabapentin (N = 55) provided very low‐quality evidence for improved walking distance and pain intensity compared with placebo in the intermediate and long‐term follow‐up periods (Yaski 2007). This trial reported that some participants randomized to the gabapentin group (no data specified) experienced mild to moderate drowsiness or dizziness, or both. There was very low‐quality evidence from one trial (N = 152) that methylcobalamin (vitamin B12) plus conservative care improved walking distance in the intermediate and long‐term follow‐up compared with conservative treatment alone (Waikakul 2000). There were no reported adverse effects for methylocabalin.

Epidural injections

All four trials evaluating epidural injections provided very low‐quality evidence for all outcomes (Cuckler 1985; Fukusaki 1988; Koc 2009; Zahaar 1991). One trial (N = 53) comparing translaminar epidural block injections, with or without steroids, with placebo showed improved walking distance only immediately after the injection (Fukusaki 1988). Another small trial (N = 29)evaluating intralaminar epidural steroid injection plus epidural block compared with home exercise or inpatient physical therapy demonstrated improvements in pain intensity, function, and quality of life at two weeks follow‐up (Koc 2009). One trial evaluating caudal (Zahaar 1991) (N = 30) and another translaminar (Cuckler 1985) (N = 37) epidural steroid injections showed no difference in global improvement compared with placebo injections. Two trials did not mention adverse events (Cuckler 1985; Zahaar 1991), whereas the other two trials reported no complications after the injections (Fukusaki 1988; Koc 2009).

Calcitonin injections

There was very low‐quality evidence from six trials (N = 231) that calcitonin was no better than placebo or paracetamol, regardless of mode of administration or outcome assessed (Eskola 1992; Porter 1983; Porter 1988; Podichetty 2004; Sahin 2009; Tafazal 2007). Adverse effects of the calcitonin injections were reported as minor (nausea and rash) and were experienced among 40% to 89% of the participants.

Multimodal nonoperative treatment

Five trials compared multimodal nonoperative care with indirect or direct surgical decompression (Amundsen 2000; Malmivaara 2007; Weinstein 2007; Weinstein 2008; Zucherman 2004). In general, multimodal nonoperative treatment was used to simulate usual nonoperative care in the community and varied considerably within and across the trials. Nonoperative treatments included orthosis, rehabilitation, physical therapy, exercise, NSAIDs, analgesics, education, heat and cold applications, transelectrical nerve stimulation, ultrasound, and epidural injections. Details on the frequency or duration of nonoperative care were lacking.

There was very low‐quality evidence from one trial (N = 191) that indirect decompression using interspinous spacers (X‐Stop, St Francis Medical Technologies, Concord, CA) with or without grade 1 spondylolisthesis provided long‐term global improvement and improved quality of life compared with multimodal nonoperative care (Zucherman 2004). No complications were reported for participants receiving nonoperative care. Complications were reported in 11% of participants undergoing interspinous spacer implants; these included spinous process fracture, coronary ischemia, respiratory distress, hematoma, and death due to pulmonary edema.

There was very low‐quality evidence based on intention‐to‐treat analysis among randomized participants from one trial (N = 304) that direct surgical decompression with or without fusion for spondylolisthesis was no better than multimodal nonoperative care for all outcomes assessed (Weinstein 2007). At two‐year follow‐up, about 40% of participants crossed over in either direction. One trial (N = 94) provided low‐quality evidence that direct surgical decompression improved back and leg pain but not walking ability compared with nonoperative care at six months and two years (Malmivaara 2007). The intention‐to‐treat analysis in another trial (N = 289) with a high crossover rate (51% of participants assigned to nonoperative care received surgery) provided very low‐quality evidence that surgical decompression improved bodily pain at two years but not four years compared with nonoperative treatment (Weinstein 2008). A meta‐analysis was performed for one outcome, the Oswestry Disability Index, among randomized participants in two trials comparing direct decompression with or without fusion to multimodal nonoperative care (Malmivaara 2007; Weinstein 2008). There was no significant difference at six months (MD ‐3.66, 95% CI ‐10.12 to 2.80) and one year (MD ‐6.18, 95% CI ‐15.03 to 2.66). At 24 months, a significant difference was found favouring decompression (MD ‐4.43, 95% CI ‐7.91 to ‐0.96). See Figure 4. The quality of the evidence for each follow‐up period was very low. Longer‐term data were available but not combined due to larger differences in the follow‐up periods.

4.

Forest plot of comparison: 1 Direct decompression ± fusion versus multimodal nonoperative care for Oswestry Disability Index, outcome: 1.1 Oswestry Disability Index.

Perioperative complication rates reported for direct decompression with or without fusion ranged from 5.4% to 14%, with dural tears being the most commonly reported. Postoperative complications ranged from 8.2% to 18% of participants and included pulmonary edema, peridural hematoma, and sepsis.

Physical therapy

Four trials evaluated various physical therapy interventions, each of which included some form of exercise (Goren 2010; Koc 2009; Pua 2007; Whitman 2006). None of the trials demonstrated improved walking ability. There was low‐quality evidence from one small trial (N = 45) that exercise was better than no treatment for leg pain and function in the short term (Goren 2010). Another small trial (N = 68) provided low‐quality evidence that unweighted treadmill walking was no better than stationary cycling in the short term regardless of outcome (Pua 2007). The other two trials provided very low‐quality evidence for all outcomes. One trial (N = 29) demonstrated that inpatient physical therapy improved pain intensity, function, and quality of life in the short term compared with a home exercise program plus oral diclofenac (Koc 2009). The other trial (N = 68) showed short‐term global improvement using a combination of manual therapy, exercise, and unweighted treadmill walking compared with flexion exercises, walking, and sham ultrasound (Whitman 2006). Among the physical therapy trials, one patient reported a mild increase in symptoms with exercises (Pua 2007) and in another one patient developed angina pectoris (Koc 2009).

Discussion

Neurogenic claudication is an important and growing cause of pain, disability, and loss of independence in the elderly. The purpose of this review was to evaluate the effectiveness of nonoperative treatments for neurogenic claudication. Our findings suggest that the current evidence is of low and very low quality. This prohibits the ability to make any conclusions about the effectiveness of nonoperative treatment and suggests that future research is very likely to have an important impact on the estimates of the effect, and is likely to change the estimates found in our review. We found low or very low‐quality evidence from single, generally small trials that gabapentin, methylcobalamin, and prostaglandins may improve walking distance. Walking distance was also improved after the use of translaminar epidural block injections, with or without steroids, but only immediately after the injection. Benefits beyond two weeks were not seen with epidural injections regardless of dose, mode of administration, or outcome. Despite the lack of evidence, 25% of all epidural injections are administered for symptoms of lumbar spinal stenosis and their use is growing (Chou 2007). Calcitonin failed to show any benefit, whether administered by injection or nasal spray. Physical therapy is a recommended treatment for neurogenic claudication; however, current evidence has not established its role. What constitutes physical therapy varied considerably among the trials. A common denominator was exercise. Exercise was of short‐term benefit for leg pain and function compared with no treatment, but it is uncertain what the important components of an exercise program are and whether supervised exercise is more effective than a home‐based program. Among the nonoperative trials that reported statistically significant differences in outcomes, the effect sizes were small and unlikely to be of clinical significance. Larger effect sizes were seen favoring indirect decompression using interspinous spacers (X‐Stop) (Zucherman 2004) and direct decompression with or without fusion (Malmivaara 2007) compared with multimodal nonoperative treatment. However, the nonoperative care used in the surgical trials varied significantly, was typically unstructured, and often consisted of failed therapy (the surgical protocol required patients to have failed conservative care prior to surgery). The relationship between symptoms of neurogenic claudication (lower extremity pain numbness, tingling, burning, weakness, and heaviness) and standing or walking ability is unknown. This review found that participants who reported significantly improved back and leg pain (Koc 2009; Malmivaara 2007), back pain–related disability (Goren 2010; Malmivaara 2007), and global improvement (Koc 2009; Whitman 2006) did not have corresponding improvement in their ability to walk. This could be explained in part by the way walking ability was evaluated. The Treadmill Walking Test was used in a third of trials measuring walking distance. However, the Treadmill Walking Test has been found to underestimate walking ability (Tomkins 2009) and patients often refuse to walk on a treadmill because of their fear of falling (Whitman 2006). More valid methods of assessing walking ability include the Self Pace Walking Test (Tomkins 2009).

Summary of main results

We identified 21 RCTs including 1851 participants randomized to 23 different comparisons, in which at least one comparison was nonoperative treatment. Nonoperative treatments included calcitonin, epidural injections, oral medications, physical therapy, and multimodal nonoperative care.

Only four trials had a low risk of bias. There is low quality evidence that prostaglandins improve walking ability compared with etodolac (NSAID); exercise improves leg pain and function compared with no treatment; that unweighted treadmill walking provides similar improvements in pain, function, and walking ability compared with stationary cycling; and direct surgical decompression improves leg pain compared with multimodal nonoperative treatment. There is very low‐quality evidence that gabapentin and methylcobalamin improve walking ability compared with placebo and conservative treatment, respectively; calcitonin is no better than placebo or paracetamol; epidural steroid injections improve pain, function, and quality of life, up to two weeks, compared with home exercise or inpatient physical therapy; and indirect surgical decompression (interspinous spacers) improves quality of life and global recovery compared with multimodal nonoperative care.

Future studies should pay special attention to evidence‐based clinical criteria for neurogenic claudication and provide appropriate and clear descriptions of randomization and concealment of treatment allocation. Trials on epidural injections and oral medication should ensure participant, provider, and assessor blinding. Trials on physical therapies and surgery, in which participant blinding and provider blinding are not possible, should ensure that there is independent assessment of outcomes. All trials should include valid measures of walking ability, have high follow‐up rates, provide sufficient data on all primary outcomes, base conclusions on intention‐to‐treat analysis, and track and report co‐interventions. Adequate description of nonoperative treatments is also needed.

Overall completeness and applicability of evidence

Considerable variation in eligibility criteria exists among trials evaluating interventions for lumbar spinal stenosis with neurogenic claudication (Genevay 2010). This presents challenges with the selection, synthesis, and interpretation of the evidence on effective interventions for this population. A set of internationally agreed‐upon diagnostic criteria for neurogenic claudication due to lumbar spinal stenosis is needed. The findings in our review are concurrent with other recent systematic reviews evaluating exercise (Iwamoto 2010), calcitonin (Podichetty 2011), epidural injections (Chou 2009), oral medications (Chou 2007), interspinous spacers (Chou 2009b; Kabir 2010), and surgical decompression (Chou 2009b; Kovacs 2011). However, across these reviews there was variation on how the study population was defined.

Potential biases in the review process

The strengths of this review are the inclusion of all nonoperative interventions and the consistent inclusion and exclusion criteria for neurogenic claudication, which included the corroboration of a diagnosis of lumbar spinal stenosis with imaging. The use of these criteria to define the study population increases the likelihood that the presenting symptoms are caused by narrowing of the central or lateral foramina (Chou 2009b; Kovacs 2011; Suri 2010). However, a high degree of reliance on imaging alone can lead to an incorrect diagnosis because 20% of asymptomatic individuals older than 60 years have lumbar spinal stenosis on imaging (Boden 1990). Another strength of this review is the use of the rigorous methods recommended by The Cochrane Collaboration, the World Health Organization, and the Cochrane Back Review Group (Furlan 2009; Higgins 2011). This included the use of the GRADE method to analyse the quality of the evidence. To our knowledge, this is the first systematic review on this topic to use the GRADE method.

Limitations of this review include the potential for publication bias because only English articles were accepted. The definition of a severe flaw and the criteria used to assess risk of bias (low versus high) were arbitrary and therefore alternative definitions and criteria could have impacted the findings and conclusions of this review. The lack of high‐ or moderate‐quality evidence found in this review prohibits any recommendations for clinical practice. To resolve this uncertainty, more research is needed with special attention to evidence‐based clinical criteria for neurogenic claudication,and appropriate and clearly described methods of randomization and allocation concealment. Trials on epidural injections and oral medication should ensure blinding of all trial participants. Trials on physical interventions (physical therapy, manual therapy, surgery), in which participant blinding and provider blinding are not possible, should ensure that there is independent assessment of outcomes. All trials should use valid measures of walking ability, ensure that follow‐up rates are above 80%, provide sufficient data on all primary outcomes, base conclusions on intention‐to‐treat analysis, and track and report co‐interventions. Adequate description of nonoperative treatments is also needed.

Authors' conclusions

Implications for practice.

Lumbar spinal stenosis with neurogenic claudication is an important cause of disability in the elderly. Current evidence for nonoperative care is of low and very low‐quality and thus prohibits recommendations to guide clinical practice.

Implications for research.

Given the expected exponential rise in its prevalence, more high‐quality research is urgently needed.

Notes

There is no published protocol for this review. It was initially published in Spine (Ammendolia 2012) and then converted to a Cochrane Review. Although the authors conducted their review in accordance with the Cochrane Back Review Group method guidelines (Furlan 2009), they did not have access to complete risk of bias information at the time of publication. This information will be incorporated into the review when it is updated in two years' time.

Appendices

Appendix 1. Search strategies

MEDLINE

1 randomized controlled trial.pt.    
 2 controlled clinical trial.pt.    
 3 randomized.ab.    
 4 placebo.ab,ti.    
 5 drug therapy.fs.    
 6 randomly.ab,ti.    
 7 trial.ab,ti.    
 8 groups.ab,ti.    
 9 or/1‐8    
 10 (animals not (humans and animals)).sh.    
 11 9 not 10    
 12 exp Constriction, Pathologic/    
 13 limit 12 to yr="1976 ‐ 1982"  
 14 exp Lumbar Vertebrae/

15 limit 14 to yr="1966 ‐ 1982"      
 16 exp Spinal Canal/     
 17 limit 16 to yr="1966 ‐ 1982"  
 18 exp Spinal Diseases/   
 19 limit 18 to yr="1966 ‐ 1982" 
 20 exp Spinal Stenosis/
 21 spinal stenosis.mp.
 22 (lumbar adj5 stenosis).mp.
 23 (spin* adj5 stenosis).mp.
 24 neurogenic claudication.mp.
 25 exp Spinal Osteophytosis/
 26 exp Spondylosis/
 27 (lumb* adj5 spondyl*).mp.
 28 exp Cauda Equina/
 29 lumbar radicular pain.mp.
 30 27 or 25 or 28 or 21 or 26 or 17 or 20 or 15 or 22 or 24 or 13 or 19 or 23 or 29
 31 11 and 30

1   Clinical Article/ 
 2   exp Clinical Study/ 
 3   Clinical Trial/ 
 4   Controlled Study/ 
 5   Randomized Controlled Trial/ 
 6   Major Clinical Study/
 7   Double Blind Procedure/
 8   Multicenter Study/
 9   Single Blind Procedure/
 10 Phase 3 Clinical Trial/
 11 Phase 4 Clinical Trial/
 12 crossover procedure/
 13 placebo/
 14 or/1‐13
 15 allocat$.mp.
 16 assign$.mp.
 17 blind$.mp.
 18 (clinic$ adj25 (study or trial)).mp.
 19 compar$.mp.
 20 control$.mp.
 21 cross?over.mp.
 22 factorial$.mp.
 23 follow?up.mp.
 24 placebo$.mp.
 25 prospectiv$.mp.
 26 random$.mp.
 27 ((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).mp.
 28 trial.mp.
 29 (versus or vs).mp.
 30 or/15‐29
 31 14 and 30
 32 human/
 33 Nonhuman/
 34 exp ANIMAL/
 35 Animal Experiment/
 36 33 or 34 or 35
 37 32 not 36
 38 31 not 36
 39 37 and 38
 40 38 or 39
 41 exp vertebral canal stenosis/
 42 (spin* adj5 stenosis).mp.
 43 (lumbar adj5 stenosis).mp.
 44 (neurogenic adj2 claudication).mp.
 45 (Spin* adj2 Osteophytosis).mp.
 46 exp cauda equina/
 47 lumbar radicular pain.mp.
 48 (lumb* adj5 spondyl*).mp.
 49 exp spondylosis/
 50 spondylolisthesis/
 51 or/41‐50
 52 40 and 51

CENTRAL

ID Search

#1 MeSH descriptor: [Spinal Stenosis] explode all tre
 #2 (spin* near/5 stenosis)
 #3 lumb* near/5 stenosis
 #4 neurogenic claudication
 #5 MeSH descriptor: [Spinal Osteophytosis] explode all trees
 #6 MeSH descriptor: [Spondylosis] explode all trees
 #7 lumb* near/5 spondyl*
 #8 MeSH descriptor: [Cauda Equina] explode all trees
 #9 lumbar radicular pain
 #10 #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 in Trials

CINAHL 

S39 S28 AND S38
 S38 S29 or S30 or S31 or S32 or S33 or S34 or S35 or S36 or S37
 S37 lumb* W5 spondyl*
 S36 (MH "Spondylolisthesis") or (MH "Spondylolysis")
 S35 "lumbar radicular pain"
 S34 (MH "Cauda Equina")
 S33 (MH "Spinal Osteophytosis")
 S32 "neurogenic claudication"
 S31 lumb* W5 stenosis
 S30 spin* W5 stenosis
 S29 (MH "Spinal Stenosis")
 S28 S26 NOT S27
 S27 (MH "Animals")
 S26 S7 or S12 or S19 or S25
 S25 S20 or S21 or S22 or S23 or S24
 S24 volunteer*
 S23 prospectiv*
 S22 control*
 S21 followup stud*
 S20 follow‐up stud*
 S19 S13 or S14 or S15 or S16 or S17 or S18
 S18 (MH "Prospective Studies+")
 S17 (MH "Evaluation Research+")
 S16 (MH "Comparative Studies")
 S15 latin square
 S14 (MH "Study Design+")
 S13 (MH "Random Sample")
 S12 S8 or S9 or S10 or S11
 S11 random*
 S10 placebo*
 S9 (MH "Placebos")
 S8 (MH "Placebo Effect")
 S7 S1 or S2 or S3 or S4 or S5 or S6
 S6 triple‐blind
 S5 single‐blind
 S4 double‐blind
 S3 clinical W3 trial
 S2 "randomi?ed controlled trial*"
 S1 (MH "Clinical Trials+")

ICL

S1 Publication Type:Clinical Trial
 S2 Publication Type:Controlled Clinical Trial
 S3 Publication Type:Randomized Controlled Trial
 S5 All Fields:random* OR All Fields:placebo* OR All Fields:sham
 S6 All Fields:versus OR All Fields:vs
 S7 All Fields:"clinical trial" OR All Fields:"controlled trial"
 S8 All Fields:double‐blind OR All Fields:"double‐blind"
 S9 All Fields:single‐blind OR All Fields:"single blind"
 S10 Publication Type:Clinical Trial OR , Publication Type:Controlled Clinical Trial OR , Publication Type:Randomized Controlled Trial OR All Fields:random* OR All Fields:placebo* OR All Fields:sham OR All Fields:versus OR All Fields:vs OR All Fields:"clinical trial" OR All Fields:"controlled trial" OR All Fields:double‐blind OR All Fields:"double‐blind" OR All Fields:single‐blind OR All Fields:"single blind"
 S11 Subject:"Spinal Stenosis" OR All Fields:"spinal stenosis"
 S12 Subject:"Spinal Osteophytosis" OR Subject:"Spondylosis" OR Subject:"Spondylolisthesis"
 S13 Subject:"Cauda Equina" OR All Fields:"lumbar radicular pain"
 S14 Subject:"Spinal Stenosis" OR All Fields:"spinal stenosis" OR Subject:"Spinal Osteophytosis" OR Subject:"Spondylosis" OR Subject:"Spondylolisthesis" OR Subject:"Cauda Equina" OR All Fields:"lumbar radicular pain"
 S15 Publication Type:Clinical Trial OR , Publication Type:Controlled Clinical Trial OR , Publication Type:Randomized Controlled Trial OR All Fields:random* OR All Fields:placebo* OR All Fields:sham OR All Fields:versus OR All Fields:vs OR All Fields:"clinical trial" OR All Fields:"controlled trial" OR All Fields:double‐blind OR All Fields:"double‐blind" OR All Fields:single‐blind OR All Fields:"single blind" AND Subject:"Spinal Stenosis" OR All Fields:"spinal stenosis" OR Subject:"Spinal Osteophytosis" OR Subject:"Spondylosis" OR Subject:"Spondylolisthesis" OR Subject:"Cauda Equina" OR All Fields:"lumbar radicular pain"

PEDro

Abstract and title: stenosis
 AND
 Body part: lumbar spine, sacroiliac joint or pelvis
 AND
 Method: clinical trialBottom of Form

Appendix 2. Criteria for risk of bias assessment for RCTs (Higgins 2011)

Random sequence generation (selection bias)

Selection bias (biased allocation to interventions) due to inadequate generation of a randomized sequence

There is a low risk of selection bias if the investigators describe a random component in the sequence generation process such as: referring to a random number table, using a computer random number generator, coin tossing, shuffling cards or envelopes, throwing dice, drawing of lots, minimization (minimization may be implemented without a random element, and this is considered to be equivalent to being random).

There is a high risk of selection bias if the investigators describe a non‐random component in the sequence generation process, such as: sequence generated by odd or even date of birth, date (or day) of admission, hospital or clinic record number; or allocation by judgment of the clinician, preference of the participant, results of a laboratory test or a series of tests, or availability of the intervention.

Allocation concealment (selection bias)

Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment

There is a low risk of selection bias if the participants and investigators enrolling participants could not foresee assignment because one of the following, or an equivalent method, was used to conceal allocation: central allocation (including telephone, web‐based and pharmacy‐controlled randomization); sequentially numbered drug containers of identical appearance; or sequentially numbered, opaque, sealed envelopes.

There is a high risk of bias if participants or investigators enrolling participants could possibly foresee assignments and thus introduce selection bias, such as allocation based on: using an open random allocation schedule (e.g. a list of random numbers); assignment envelopes were used without appropriate safeguards (e.g. if envelopes were unsealed or non‐opaque or not sequentially numbered); alternation or rotation; date of birth; case record number; or other explicitly unconcealed procedures.

Blinding of participants

Performance bias due to knowledge of the allocated interventions by participants during the study

There is a low risk of performance bias if blinding of participants was ensured and it was unlikely that the blinding could have been broken; or if there was no blinding or incomplete blinding, but the review authors judge that the outcome is not likely to be influenced by lack of blinding.

Blinding of personnel/care providers (performance bias)

Performance bias due to knowledge of the allocated interventions by personnel/care providers during the study

There is a low risk of performance bias if blinding of personnel was ensured and it was unlikely that the blinding could have been broken; or if there was no blinding or incomplete blinding, but the review authors judge that the outcome is not likely to be influenced by lack of blinding.

Blinding of outcome assessor (detection bias)

Detection bias due to knowledge of the allocated interventions by outcome assessors

There is low risk of detection bias if the blinding of the outcome assessment was ensured and it was unlikely that the blinding could have been broken; or if there was no blinding or incomplete blinding, but the review authors judge that the outcome is not likely to be influenced by lack of blinding, or:

• for patient‐reported outcomes in which the patient was the outcome assessor (e.g. pain, disability): there is a low risk of bias for outcome assessors if there is a low risk of bias for participant blinding (Boutron 2005);

• for outcome criteria that are clinical or therapeutic events that will be determined by the interaction between patients and care providers (e.g. co‐interventions, length of hospitalization, treatment failure), in which the care provider is the outcome assessor: there is a low risk of bias for outcome assessors if there is a low risk of bias for care providers (Boutron 2005);

• for outcome criteria that are assessed from data from medical forms: there is a low risk of bias if the treatment or adverse effects of the treatment could not be noticed in the extracted data (Boutron 2005).

Incomplete outcome data (attrition bias)

Attrition bias due to amount, nature or handling of incomplete outcome data

There is a low risk of attrition bias if there were no missing outcome data; reasons for missing outcome data were unlikely to be related to the true outcome (for survival data, censoring unlikely to be introducing bias); missing outcome data were balanced in numbers, with similar reasons for missing data across groups; for dichotomous outcome data, the proportion of missing outcomes compared with the observed event risk was not enough to have a clinically relevant impact on the intervention effect estimate; for continuous outcome data, the plausible effect size (difference in means or standardized difference in means) among missing outcomes was not enough to have a clinically relevant impact on observed effect size, or missing data were imputed using appropriate methods (if drop‐outs are very large, imputation using even 'acceptable' methods may still suggest a high risk of bias) (van Tulder 2003). The percentage of withdrawals and drop‐outs should not exceed 20% for short‐term follow‐up and 30% for long‐term follow‐up and should not lead to substantial bias (these percentages are commonly used but arbitrary, not supported by literature) (van Tulder 2003).

Selective reporting (reporting bias)

Reporting bias due to selective outcome reporting

There is low risk of reporting bias if the study protocol is available and all of the study's pre‐specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre‐specified way, or if the study protocol is not available but it is clear that the published reports include all expected outcomes, including those that were pre‐specified (convincing text of this nature may be uncommon).

There is a high risk of reporting bias if not all of the study's pre‐specified primary outcomes have been reported; one or more primary outcomes is reported using measurements, analysis methods or subsets of the data (e.g. subscales) that were not pre‐specified; one or more reported primary outcomes were not pre‐specified (unless clear justification for their reporting is provided, such as an unexpected adverse effect); one or more outcomes of interest in the review are reported incompletely so that they cannot be entered in a meta‐analysis; the study report fails to include results for a key outcome that would be expected to have been reported for such a study.

Group similarity at baseline (selection bias)

Bias due to dissimilarity at baseline for the most important prognostic indicators. 

There is low risk of bias if groups are similar at baseline for demographic factors, value of main outcome measure(s), and important prognostic factors (examples in the field of back and neck pain are duration and severity of complaints, vocational status, percentage of patients with neurological symptoms) (van Tulder 2003).

Co‐interventions (performance bias)

Bias because co‐interventions were different across groups

There is low risk of bias if there were no co‐interventions or they were similar between the index and control groups (van Tulder 2003).

Compliance (performance bias)

Bias due to inappropriate compliance with interventions across groups

There is low risk of bias if compliance with the interventions was acceptable, based on the reported intensity/dosage, duration, number and frequency for both the index and control intervention(s). For single‐session interventions (e.g. surgery), this item is irrelevant (van Tulder 2003).

Intention‐to‐treat analysis

There is low risk of bias if all randomized patients were reported and analysed in the group to which they were allocated by randomization.   

Timing of outcome assessments (detection bias)

Bias because important outcomes were not measured at the same time across groups

There is low risk of bias if all important outcome assessments for all intervention groups were measured at the same time (van Tulder 2003).

Other bias

Bias due to problems not covered elsewhere in the table

There is a low risk of bias if the study appears to be free of other sources of bias not addressed elsewhere (e.g. study funding).

Data and analyses

Comparison 1. Direct Decompression ± fusion versus multimodal nonoperative care for Oswestry Disability Index.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Oswestry Disability Index	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.1 6 Months	2	349	Mean Difference (IV, Random, 95% CI)	‐3.66 [‐10.12, 2.80]
1.2 12 Months	2	340	Mean Difference (IV, Random, 95% CI)	‐6.17 [‐15.02, 2.67]
1.3 24 Months	2	315	Mean Difference (IV, Random, 95% CI)	‐4.43 [‐7.91, ‐0.96]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Amundsen 2000.

Methods	Randomized Controlled Trial
Participants	100 subjects, 54 male, 46 female, median age of 59 (males were 1.5 years higher than females). Median back pain duration was 14 years, median duration of sciatica was 2 years Setting: Neurology department in a hospital in Norway
Interventions	1) Surgery – partial or total laminectomy, medial facetecomy, discectomy, and/or removal of osteophytes from the vertebral margins or facet joints.  No fusions. (n=13) 2) Conservative therapy ‐ "lumbar orthosis use for 1 month worn during the day for all activities plus instruction and back school.” (n=18)
Outcomes	1) VAS 2) Verbal Rating Scale 3) Subjective Change (Better, Worse, or Unchanged) 4) Work Status 5) Subjective rating from evaluating physician and study team (Excellent, Fair, Unchanged, Worse)   Follow‐up: 6 months, 1, 4 and 10 years
Notes	VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	Low risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Low risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Low risk
Was the timing of the outcome assessment similar in all groups?	Unclear risk	Information currently not available

Cuckler 1985.

Methods	Randomized Controlled Trial
Participants	73 subjects in total, 37 with spinal stenosis, 36 with acute herniated nucleus pulposus, 37 males, 36 female, average age of 48.5 years in the experimental group and 49.5 years in the placebo group. Experimental group average 36.6 months in symptom duration, placebo group averaged 29.4 months Setting: Orthopedic surgery department in the United States
Interventions	1) Steroid group ‐ 2mL of sterile water containing 80mg of methylprednisolone acetate combined with 5mL of 1% procaine was injected into the epidural space in the region between the 3rd and 4th lumbar vertebrae with the patient in the lateral decubitus position lying on the side of the painful limb (n=42, 20 with stenosis) 2) Placebo group ‐ 2mL of saline combined with 5mL of 1% procaine was injected into the epidural space in the region between the 3rd and 4th lumbar vertebrae with the patient in the lateral decubitus position lying on the side of the painful limb. (n=31, 17 with stenosis)   All patients were advised to take mild analgesics (aspirin or acetaminophen) during the post‐injection period. Second injection given if less than 50% improvement after 24 hours ‐ considered treatment failure
Outcomes	1) Subjective percentage of improvement with 75% required to be considered a treatment improvement, if less than 50% after 24 hours was considered a treatment failure 2) Re‐injection Rates 3) Surgery Rates   Follow‐up: 24 hours, every 3 months up to 30 months, averaging 20.2 months in the steroid group and 21.5 months in the control group
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	High risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	High risk
Was the timing of the outcome assessment similar in all groups?	High risk

Eskola 1992.

Methods	Randomized Controlled Trial
Participants	39 subjects with an average of 6 years of pain, average age of 56.6 years of age, 20 males and 19 females   Setting: Orthopedic hospital in Finland
Interventions	1)100IU Calcitonin injection every other day for 4 weeks. (n=20) 2) Placebo treatment (Miacalcic Sandoz 100IU) every other day for 4 weeks. (n=19)
Outcomes	1) VAS 2) Treadmill Test 3) Coping with ADLs 4) Digitest Ergojump Contact Test 5) Blood Tests   Follow up: I, 3, 4, 6, and 12 months
Notes	ADLs = Activities of Daily Living, VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	High risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	Unclear risk	Information currently not available
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Low risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Unclear risk	Information currently not available
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Fukusaki 1988.

Methods	Randomized Controlled Trial
Participants	53 subjects, 38 males and 15 female. Group 1 averaged 70 years of age and 79 days of symptoms on average, group 2 averaged 69 years of age and an average of 82 days of symptoms, group 3 averaged 72 years of age and 94 days of symptoms on average Setting: Anesthesia department in Japan
Interventions	1) Epidural injection with 8 mL of saline, repeated twice in the first week. (n=16) 2) Epidural injection with 8 mL of 1% mepivacaine, repeated twice in the first week. (n=18) 3) Epidural injection with a mixture of 8 mL of 1% mepivacaine and 40 mg of methylprednisone, repeated twice in the first week. (n=19)
Outcomes	1) Walking distance which was graded according to distance (Excellent, Good, or Poor)   Follow‐up: 1 week, 1 month, 3 months
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Unclear risk	Information currently not available
Was the care provider blinded to the intervention?	Unclear risk	Information currently not available
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	High risk
Was the timing of the outcome assessment similar in all groups?	High risk

Goren 2010.

Methods	Randomized Controlled Trial
Participants	45 subjects, 13 males, 32 females, average ages in groups of 57.4, 49.13, and 53.06. 7 subjects with pain duration of 3‐6 months, 7 with pain duration of 6‐12 months, and 31 with pain duration of greater than 12 months Setting: Rehabilitation center in Turkey
Interventions	1) Stretching and strengthening exercises for lumbar, abdominal, leg muscles as well as low‐intensity cycling exercises were given as therapeutic exercises. Ultrasound was applied with 1mHz, 1.5W/cm2 intensity, in continuous mode on the back muscle for 10 minutes (n=17) 2) Same as group 1 with ultrasound on off‐mode (n=17) 3) No exercise ‐ no treatment (n=16)
Outcomes	1) VAS (out of 10) 2) Treadmill test at 3 km/h for maximum of 15 minutes or 750m 3) ODI 4) Analgesic Consumption 5) Physiatrist Assessment   Follow‐up: End of 3‐week treatment period only
Notes	VAS = Visual Analogue Scale, ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	Low risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Koc 2009.

Methods	Randomized Controlled Trial
Participants	29 subjects, 21 male, 8 female, average ages of 62.6, 61.1, and 53.1 years in the three groups, average pain duration of 5.7 years, 5.0 years, and 5.7 years in the three groups Setting: Medical school department of physical medicine and rehabilitation in Turkey
Interventions	1) Conservative inpatient physical therapy program 5 days a week for 2 weeks. PT included applications of ultrasound 1.5 W/cm2 for 10min, hot pack for 20min, and TENS for 20min to the lumbar region (n=13) 2) Lumbar epidural steroid injections, 10 mL of solution containing 60mg of triamcinolon acetonide (1.5 mL), 15 mg of 0.5% bupivacain hydrochloride (3 mL), and 5.5 mL of physiologic saline (0.9%NaCl) was injected in 3.5 minutes (n=10) 3) Control group (n=10)   All patients included were trained to pursue a home‐based therapeutic exercise program performed twice daily for a period of 6 months, and oral diclofenac sodium 75mg was administered to all patients twice daily for 2 weeks
Outcomes	1) VAS 2) Treadmill Walk Test 3) Nottingham Health Profile 4) RMDI 5) Functional testing including finger to floor distance, sit‐to‐stand, and a weight carrying test   Follow‐up: 2 weeks, 1, 3, and 6 months
Notes	ODI = Oswestry Disability Index, RMDI = Roland Moris Disability Index, VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Low risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Malmivaara 2007.

Methods	Randomized Controlled Trial
Participants	94 subjects, 22% of surgical subjects were male, 45% of non‐operative subjects were male. Nonoperative group had average age of 62.9 years, surgical group had average age of 63.9 years. Surgical group averaged 14 years since onset of symptoms, non‐surgical group average 16 years since onset of symptoms.  Minimum of 6 months of symptoms for study inclusion Setting: Research Center in Finland
Interventions	1) Segmental decompressive surgery with facetectomy (n=50) 2) Non‐operative treatment – NSAIDS when indicated and seen one to three times by a physiotherapist, in addition to the standard visit at each follow‐up. The physiotherapist gave all patients educational brochure. The patients were encouraged to use their back in a normal way. Pain‐relieving body postures were taught as well as basic ergonomics related to lifting and carrying. Individually structured programs included trunk muscle endurance and stretching‐type exercises. Additional individual physiotherapy consisting of passive treatment methods (such as ultrasound and transcutaneous nerve stimulation) (n=44) The patients in the surgical group also received the brochure and the instructions described above
Outcomes	1) 11‐point numerical pain rating scale for back and leg pain 2) Walking ability (distance without a break) also via treadmill test 3) General health status on a 5‐point scale (Very Good, Quite Good, Average, Quite Poor, or Very Poor) 4) ODI 5) Ability to complete certain ADL without difficulty, some difficulty, marked difficulties or not at all 6) Radiographic examination Follow‐up: 6 months, 1, 2 and 6 years
Notes	ADLs = Activities of Daily Living, ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Matsudaira 2009.

Methods	Randomized Controlled Trial
Participants	79 subjects, 24 males and 24  females, with an average age of 69.6 years in the Limaprost group and 72.2 in the etodolac group Setting: Orthopedic surgery in a medical faculty in Japan
Interventions	1) Oral prostaglandin E1 derivative (15 g Limaprost) three times daily for eight weeks (n=39) 2) 400 mg of etodolac (NSAID) twice daily for eight weeks (n=40)
Outcomes	1) SF‐36 2) Verbal Pain Rating Scales 3) Walking Distance 4) LBP Severity 5) Leg Pain Severity 6) Leg Numbness Severity 7) Treatment Satisfaction   Follow‐up: 8 weeks
Notes	LBP = Low Back Pain
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Unclear risk	Information currently not available
Was the care provider blinded to the intervention?	Unclear risk	Information currently not available
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Unclear risk	Information currently not available
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Podichetty 2004.

Methods	Randomized Controlled Trial
Participants	55 subjects with an average age of 68.5 years and an average of 36.2 weeks of the condition in the intervention group and 29.8 weeks in the placebo group, 33 males and 22 females Setting: Spinal center in the United States
Interventions	1) 400 IU intranasal calcitonin daily for 6 weeks followed by open label 6‐week extension (n=36) 2) Placebo nasal spray daily for 6 weeks, followed by open label 6‐week extension, during which all patients received 400IU calcitonin (n=19)
Outcomes	1) VAS 2) Walking Capacity (time and distance to stopping) 3) ODI 4) Stenosis Specific Questionnaire 5) Satisfaction with pain levels, functional status, and treatment received 6) SF‐36 7) Symptom Diary Follow‐up: 12 weeks
Notes	VAS = Visual Analogue, ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	High risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	Low risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Low risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Porter 1983.

Methods	Randomized Controlled Trial
Participants	41 subjects with 10 in a double blind RCT crossover, 37 males and 4 females with ages between 41 and 67 years Setting: Infirmary in England
Interventions	1) 100 IU salmon calcitonin injection four times per week, sometimes with Maxalon for nausea (n=5) 2) Matching placebo (n=5)
Outcomes	1) Walking chart and ability to walk more than 1 mile 2) ODI Follow‐up: 10 weeks
Notes	Only responders randomized, ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Unclear risk	Information currently not available
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Information currently not available
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Unclear risk	Information currently not available
Were the groups similar at baseline regarding the most important prognostic indicators?	Low risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	High risk
Was the timing of the outcome assessment similar in all groups?	High risk

Porter 1988.

Methods	Randomized Controlled Trial
Participants	42 subjects, 35 male, 7 female, average of 53.6 years in 20 subjects and 56.7 years in 22 subjects, median duration of back pain reported was 11 years for 19 subjects, and 14 years for 22 subjects. Median duration of claudication was 1.25 years for 20 subjects and 4.5 years for 22 subjects Setting: Infirmary in England
Interventions	1) 100 IU of salmon calcitonin injected subcutaneously four times per week for eight weeks (n=20) 2) 1 mL of saline injected four times per week for eight weeks (n=22)
Outcomes	1) VAS 2) Claudication threshold and tolerance for walking at constant speed with verbal description of walking pain on a 5‐point pain rating scale 3) 3 level mobility assessment 4) Analgesic requirements 5) 3 level sleep disturbance 6) Treatment success defined as 100% improvement in walking distance and able to walk 800m Follow‐up: 4 and 8 weeks
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	Unclear risk	Information currently not available
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Information currently not available
Was the drop‐out rate described and acceptable?	Low risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Unclear risk	Information currently not available
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Pua 2007.

Methods	Randomized Controlled Trial
Participants	68 subjects, 35 males, 33 females, average age of 58 years, 12 week median pain duration Setting: Hospital in Singapore
Interventions	1) Unweighted treadmill training ‐ Weeks 1 and 2, participants walked with a relatively pain‐free gait which translated to 30–40% of body weight. In Weeks 3 to 6, participants were encouraged to walk at a moderate intensity. The duration of each treadmill session was limited by participant tolerance or to a maximum of 30 minutes.  Two times per week for 6 weeks = 12 sessions (n=33) 2) Cycling on upright bicycle ‐ During Weeks 1 and 2, participants cycled at their comfortable pace at 50 to 60 rpm. Participants were instructed to assume a flexed posture. In Weeks 3 to 6, participants were encouraged to exercise at a moderate intensity and the duration of each cycling session was limited by participant tolerance or to a maximum of 30 minutes. Two times per week for 6 weeks for 12 sessions (n=35)
Outcomes	1) VAS for pain over past week 2) Patient perceived benefit on a 6‐point scale 3) ODI 4) RMDI 5) Walking Ability   Follow‐up: 3 and 6 weeks
Notes	ODI = Oswestry Disability Index, RMDI = Roland Moris Disability Index, VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	Low risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Low risk
Was the timing of the outcome assessment similar in all groups?	High risk

Sahin 2009.

Methods	Randomized Controlled Trial
Participants	45 subjects 31 males and 14 females, average ages of 57.65 years in the calcitonin group and 54.45 years in the paracematol group Setting: Physical and Rehabilitation Medicine Department in Turkey
Interventions	1) 200 IU intranasal calcitonin daily for 8 weeks (n=23) 2) Up to 1500mg of paracematol daily for 8 weeks (n=22)   Both groups took part in a physical therapy and exercise program five times per week for 15 sessions
Outcomes	1) VAS 2) Walking Capacity 3) RMDI 4) Ranges of Motion   Follow up: 8 weeks
Notes	RMDI = Roland Moris Disability Index, VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	Low risk
Were all randomized participants analyzed in the group to which they were allocated?	Unclear risk	Information currently not available
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Tafazal 2007.

Methods	Randomized Controlled Trial
Participants	40 subjects, 30 males, 10 females, average of 67 years in the intervention group and 70.2 years in the placebo group, average of 38.7 months with symptoms in the calcitonin group and 30.9 months in the placebo group Setting: University hospital in England
Interventions	1) Placebo nasal spray NaCl for four weeks (n=20) 2) 200 IU nasal salmon calcitonin for 4 weeks (n=20)
Outcomes	1) VAS 2) Shuttle Walking Test 3) 4 point subjective outcome of overall assessment (Excellent, Good, Fair, Poor) 4) ODI 5) Modified Somatic Perception Questionnaire 6) Modified Zung Depression Score   Follow‐up: Baseline, 4, 10, 16 weeks
Notes	ODI = Oswestry Disability Index, VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	High risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Low risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Waikakul 2000.

Methods	Randomized Controlled Trial
Participants	152 subjects, 68 males and 84 females with an average age of 66.8 years. 44 of the subjects had symptoms for less than one month, 98 had symptoms for more than one month Setting: Hospital in Thailand
Interventions	1) Conservative treatment consisting of education, activity modification, exercise and physical therapy. NSAIDs, muscle relaxants, and analgesics as necessary. Vitamin B1, B6, and B12 three times per day (n=82) 2) Conservative treatment plus methylcobalamin (Methlcobalin ESAI), 1.5mg per day in 3 divided doses after meals for 6 months (n=70)
Outcomes	1) Presence of pain on spinal motion 2) Claudication distance 3) Medication intake (NSAIDs, muscle relaxants, and steroids)   Follow‐up: every month for two years
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	Unclear risk	Information currently not available
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Weinstein 2007.

Methods	Randomized Controlled Trial
Participants	Subjects with image‐confirmed degenerative spondylolisthesis: 304 subjects in clinical trial, 303 in the observational cohort, 31% male in the surgical group, 33% male in the surgical group. Average age of 64.7 years in the surgical group and 68.2 years in the non‐surgical group. Subjects had symptoms for at least 12 weeks Setting: Multicentred orthopedic departments in the United States
Interventions	1) Assigned to surgery (standard laminectomy with or without fusion) (n=159) 2) Assigned to non‐surgical treatment – usual non‐operative care (n=145) 3) Chose surgery (n=173) 4) Chose non‐surgical treatment (n=130)
Outcomes	1) SF‐36 Bodily Pain 2) SF‐36 Bodily Function 3) Low Back Pain Bothersomeness Scale 4) Leg Pain Bothersomeness Scale 5) ODI 6) Subjective self‐reported improvement, satisfaction with current symptoms and care 7) Stenosis Bothersomeness Index Follow‐up: 6 w, 3 and 6 mo, 1, 2 and 4 yrs
Notes	ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Unclear risk	Information currently not available
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Low risk
Was the timing of the outcome assessment similar in all groups?	High risk

Weinstein 2008.

Methods	Randomized Clinical Trial
Participants	289 in the RCT, 365 in the observational cohort. 62% male in the surgical groups, 59% male in the non‐surgical groups. Average age of 63.8 in the surgical group, 66.1 in the non‐surgical group. 60% in the surgical group and 55% in the non‐surgical group had symptoms for over 6 months Setting: Multicentered ‐ orthopedic departments in the United States
Interventions	1) Assigned to surgery – standard laminectomy with or without fusion (n=138) 2) Assigned to non‐surgical treatment ‐ usual non‐operative care ‐ recommended to include at least active physical therapy, education or counseling with home exercise instruction, and the administration of NSAIDs, if tolerated (n=151) 3) Chose surgery (n=219) 4) Chose non‐surgical treatment (n=146)
Outcomes	1) SF‐36 Bodily Pain 2) SF‐36 Bodily Function 3) Low Back Pain Bothersomeness Scale 4) Leg Pain Bothersomeness Scale 5) ODI 6) Subjective self‐reported improvement, satisfaction with current symptoms and care 7) Stenosis Bothersomeness Index   Follow‐up: 6 weeks, 3 and 6 months, 1, 2 and 4 years
Notes	ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	Low risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Unclear risk	Information currently not available
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Low risk
Was the timing of the outcome assessment similar in all groups?	High risk

Whitman 2006.

Methods	Randomized Clinical Trial
Participants	58 subjects, 31 males, 27 female, 29 (group 1) with an average age of 70 years, 29 (group 2) with an average age of 68.9, median low back pain duration of 108 months in Group 1 (n=29) and 60 months in Group 2 (n=29), lower extremity median pain duration of 48 months in Group 1 (n=29) and 24 months in Group 2 (n=29) Setting: University in the United States
Interventions	1) Flexion Exercise and Walking Group – 45‐60 minutes twice per week for 6 weeks. ‐ Lumbar flexion exercises along with performance of a progressive level self pace treadmill walking program, and subtherapeutic ultrasound. The duration of each treadmill session was based on patient’s tolerance on that specific day and could extend up to 45 minutes (n=29) 2) Manual Therapy, Exercise and Walking Group ‐ 45‐60 minutes twice per week for 6 weeks ‐ Manual physical therapy (thrust and non‐thrust) to the thoracic and lumbar spine, pelvis, and lower extremities and specific exercises at discretion based on the underlying impairments. Patients received specific exercises to address impairments in mobility, strength, and/or coordination. Exercises were performed in the clinic and as part of a home exercise program. Patients also underwent a body‐weight supported treadmill ambulation program using a cable and trunk harness system to unload a specific amount of weight from the patient while the patient walks as comfortably as possible on a treadmill (n=29)
Outcomes	1) Global Rating of Change (15 point scale) 2) Numerical Pain Rating Scale for lower limb 3) Walking Tolerance test 4) ODI 5) Medication consumption 6) Satisfaction subscale of the Spinal Stenosis Scale 7) Additional use of healthcare resources   Follow‐up: 6 weeks, 1 year, long‐term mail survey (averaging 29 months)
Notes	NPRS = Numerical Pain Rating Scale, ODI = Oswestry Disability Index
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Yaski 2007.

Methods	Randomized Clinical Trial
Participants	55 subjects, 22 males, 33 females, average age of 50.8 years Setting: Hospital department of physical medicine and rehabilitation in Turkey
Interventions	1) 900 mg of gabapentin per day increased weekly by 300 mg to a maximum of 2400 mg (n=28) 2) Placebo (n=27)   Both groups received physical therapy exercises, a lumbosacral corset with steel bracing and NSAID treatments
Outcomes	1) VAS – low back and leg pain during movement 2) Walking Distance 3) Presence or absence of motor and/or sensory deficits   Follow‐up: 15 days, 1, 2, 3, 4 months
Notes	VAS = Visual Analogue Scale
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	Low risk
Was the drop‐out rate described and acceptable?	Unclear risk	Information currently not available
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	Unclear risk	Information currently not available
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	High risk

Zahaar 1991.

Methods	Randomized Clinical Trial
Participants	30 subjects, 37 male and 26 female. Steroid group averaged 46.5 years of age and 36.6 months of symptoms, control group averaged 49 years of age and 29.4 months of symptoms Setting: Medical facility in Egypt
Interventions	1) Steroid injection ‐ 5mL of hydrocortisone acetate suspension, 2x2mL carbocaine, 4% volume completed with sterile saline to 30mL (n=18) 2) Control ‐ 2x2mL of carbocaine, 4% injected into epidural space. Volume completed with sterile saline to 30mL (n=12)
Outcomes	1) Subjective percentage of improvement where 75% or more was deemed successful and surgery after injection was considered a failure   Follow‐up: 24 hours, then every three months up to 36 months averaging 20.2 months in the steroid group and 21.5 months control group
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	Unclear risk	Information currently not available
Was the patient blinded to the intervention?	High risk
Was the care provider blinded to the intervention?	Unclear risk	Information currently not available
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Low risk
Was the compliance acceptable in all groups?	Unclear risk	Information currently not available
Was the timing of the outcome assessment similar in all groups?	Low risk

Zucherman 2004.

Methods	Randomized Clinical Trial
Participants	191 subjects, 57% male and 43% female in the X STOP group. 52% male and 48% female in the non‐operative group. Average age of 70 years in the X STOP group and 69.1 years in the non‐operative group. Average of 3.5 years symptom duration in the X STOP group and 4.7 years in the non‐operative group. Setting: Spine center in the United States
Interventions	1)  X STOP Interspinous Process Decompression System (n=100) 2) Non‐operative treatment – subjects received an epidural steroid injection on enrolment and were eligible for additional injections as needed, as well as NSAIDS, analgesic agents, and physical therapy. Physical therapy consisted of education on back care and modalities such as ice packs, heat packs, massage, stabilization exercises, and pool therapy. Braces such as abdominal binders and corsets were permitted, but body jackets and chair back braces were not (n=91)
Outcomes	1) SF‐36 2) ZCQ 3) Worker’s Compensation Claims 4) ODI 5) Radiographic Changes   Follow‐up: Surgery: 7 (2 years) Control: 19 (2 years)
Notes	ODI, ZCQ = Zurich Claudication Questionnaire
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Information currently not available
Allocation concealment (selection bias)	High risk
Was the patient blinded to the intervention?	Low risk
Was the care provider blinded to the intervention?	Low risk
Blinding of outcome assessment (detection bias) All outcomes	High risk
Was the drop‐out rate described and acceptable?	High risk
Were all randomized participants analyzed in the group to which they were allocated?	High risk
Selective reporting (reporting bias)	High risk
Were the groups similar at baseline regarding the most important prognostic indicators?	High risk
Were co‐interventions avoided or similar?	Unclear risk	Information currently not available
Was the compliance acceptable in all groups?	High risk
Was the timing of the outcome assessment similar in all groups?	High risk

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Akyol 2009	Not published in English
Canovas 2009	Not published in English
Comer 2010	No imaging confirmation requirement for inclusion
Ghosh 1981	Not neurogenic claudication
Huda 2010	No imaging confirmation requirement for inclusion
Khoromi 2007a	Not neurogenic claudication but radiculopathy as inclusion criterion
Khoromi 2007b	Not neurogenic claudication but radiculopathy as inclusion criterion
Lee 2003	Patients with leg pain were excluded
Levendogluo 2009	Not an RCT
Makki 2010	Not neurogenic claudication but radiculopathy as inclusion criterion
Manchikanti 2008	Neurogenic claudication not explicitiy stated as inclusion criterion
Mariconda 2002	Not neurogenic claudication but radiculopathy as inclusion criterion
Mathews 1975	Not neurogenic claudication but sciatica or radiculopathy as inclusion criterion
Ng 2005	Mixed population not analysed separately
Oguz 2007	Not an RCT
Owlia 2007	No data on number of patients with lumbar spinal stenosis, patients did not explicitly have neurogenic claudication, subacute population
Paker 2005	Not an RCT
Price 2000	Neurogenic claudication was not an inclusion criterion
Price 2005	Not neurogenic claudication but radiculopathy as inclusion criterion
Riew 2000	Mixed population, neurogenic claudication was not an inclusion criterion
Riew 2008	Mixed population, neurogenic claudication was not an inclusion criterion
Sell 2006	Duplicate of accepted study (see Tafazal 2007)
Simopoulos 2003	Not neurogenic claudication but radiculopathy as inclusion criterion
Thackeray 2009	Not neurogenic claudication but radiculopathy as inclusion criterion
Wiilson‐MacDonald 2005	Neurogenic claudication was not an explicit requirement, patients had nerve root pain, patients were subacute
Wu 2003	Not published in English

Contributions of authors

Conception and design: C Ammendolia, KJ Stuber
 Analysis and interpretation of the data: C Ammendolia
 Drafting of the review: all members
 Critical revision of the article for important intellectual content: all members
 Final approval of the article: all members
 Statistical Expertise: A Furlan
 Administrative, technical, or logistical support:
 Collection and assembly of data: KJ Stuber, ER Rok

Sources of support

Internal sources

• No sources of support supplied

External sources

• Canadian Chiropractic Research Foundation, Canada.

Declarations of interest

The principal author received funding from the Canadian Chiropractic Research Foundation.

New