General strengthening exercise for chronic low back pain

Objectives

This is a protocol for a Cochrane Review (intervention). The objectives are as follows:

The primary objective is to investigate the benefits and harms of general strengthening exercise therapy in adults with chronic non‐specific low back pain on pain intensity, functional limitations, health‐related quality of life, psychological functioning, and adverse events compared to placebo, sham, or attention control and no treatment (including waitlists, control groups with no treatment provided, usual/normal care not controlled by the trial available to all treatment groups, or when the exercise and comparison groups receive the same co‐interventions).

Background

Low back pain (LBP) is the largest musculoskeletal cause of years lived with disability (YLD) (Wu 2020), affecting around 577 million people globally at any given time (Wu 2020). LBP was responsible for approximately 65 million YLDs in 2017, an increase of 53% since 1990 (Wu 2020). LBP is a condition with widespread implications, affecting individuals' daily activities, physical and mental health, and work‐life (Dueñas 2016), as well as being a significant socioeconomic burden regarding loss of productivity and healthcare costs (Vos 2016; Wu 2020).

Description of the condition

This review will focus on chronic non‐specific LBP (CNSLBP). LBP is defined as pain, muscle tension, or stiffness below the costal margin and above the inferior gluteal folds, with or without leg pain (Chou 2010). When LBP persists for 12 weeks or more, it is defined as chronic (Chou 2010). Approximately 90% of those experiencing LBP are categorised as non‐specific LBP, defined as symptoms not attributable to a specific pathology (e.g. fracture, ankylosing spondylitis, infection, or metastasis) (Hartvigsen 2018; Maher 2017).

Description of the intervention

Exercise therapy is one of the most recommended interventions, as demonstrated in 2018 by Oliveira and colleagues, who reviewed 15 clinical practice guidelines for managing CNSLBP, of which 71% recommended exercise therapy (Oliveira 2018). The effectiveness of exercise therapy was assessed in a Cochrane review and a network meta‐analysis in 2021 (Hayden 2021; Hayden 2021a); based on 249 and 217 randomised controlled trials (RCTs), respectively, they concluded that exercise therapy is effective for CNSLBP compared to usual care, no care, or placebo (Hayden 2021; Hayden 2021a). Furthermore, Pilates, McKenzie therapy, and functional restoration were marginally more effective than other types of exercise therapy (Hayden 2021a).

We will focus on general strengthening exercise therapy in this review. We define general strengthening exercise therapy, inspired by the definition used in a Cochrane review by Hayden and colleagues (Hayden 2021), as load‐bearing or resistance exercises, including weights, weight machines, elastic cords, or body weight to increase an individual's ability to exert or resist force, to strengthen parts of or the whole body (i.e. at least two of the following body: areas a) spinal column, b) upper extremity, or c) lower extremity)).

How the intervention might work

The specific mechanism of effect for exercise therapy is yet to be fully understood, but a review synthesised the proposed mechanisms for exercise therapy in people with CNSLBP in 2021 (Wun 2021). The study included 110 RCTs and reported 33 unique mechanisms across the RCTs. These unique mechanisms were grouped into five themes (Wun 2021):

1. 44% of the RCTs proposed a neuromuscular cause of effect;

2. 36% of the RCTs proposed a psychosocial cause of effect;

3. 9% of the RCTs proposed a neurophysiological cause of effect;

4. 6% of the RCTs proposed a cardiometabolic cause of effect;

5. 5% of the RCTs proposed a tissue‐healing cause of effect.

However, many RCTs fail to report the mechanisms of effect (Wood 2020; Wun 2021).

Why it is important to do this review

Although exercise therapy is frequently recommended in clinical practice guidelines for managing CNSLBP (Oliveira 2018), there are still significant knowledge gaps, such as: Which exercise programme works best and for whom? Although Hayden and colleagues identified that certain types of exercise therapy might be more effective than others, albeit only marginally (Hayden 2021a), it is unknown how these exercise therapy types should be delivered according to frequency, intensity, dose, mode of delivery, and in combination with other treatments.

This review will be conducted via The Collaborative Review Model (Hayden 2022). In short, The Collaborative Review Model aims to increase collaboration within the work on evidence synthesis by sharing resources, tasks, and tools to increase efficiency. The Collaborative Review Model will include conducting an overarching network review of all evidence evaluating the effectiveness of exercise therapy for CNSLBP (Hayden 2023), nine 'sub‐reviews' on specific exercise therapy types (e.g. general strengthening exercise therapy), and an evidence update for the World Health Organization (WHO) Guideline on the management of chronic primary low back pain in adults (Hayden 2022). Conducting this review as part of The Collaborative Review Model will provide substantial individual and complementary evidence using the same comprehensive methods (i.e. search strategy, data collection, risk of bias assessment, etc.) in the overarching review and the focused sub‐reviews. This will allow for a comprehensive comparison of the effectiveness of various exercise therapy types and build capacity for future research in this field (Hayden 2022).

The 2021 Cochrane review by Hayden and colleagues included RCTs published up to 2018 (Hayden 2021). Since then, approximately 200 eligible RCTs have been published (personal communication with Professor Jill A Hayden), meaning that approximately 450 RCTs are eligible for inclusion in the overarching network review. General strengthening exercise therapy was used in 49 of the 249 RCTs included in the Cochrane review by Hayden and colleagues (Hayden 2021). Since then, additional eligible RCTs have been published (personal communication with Professor Jill A Hayden).

Overall, the large and steadily increasing number of RCTs makes it difficult to provide detailed evidence about the benefits and harms of exercise therapy for CNSLBP. Furthermore, it remains unclear which exercise programme works best and for whom. This highlights the need for a separate Cochrane review focusing on general strengthening exercise therapy as part of The Collaborative Review Model (Hayden 2022). We therefore aim to investigate the benefits and harms of general strengthening exercise therapy in adults with CNSLBP. This review will include a subset of the RCTs included in the ongoing Cochrane review update 'Exercise therapy for chronic low back pain' (Hayden 2021).

Objectives

The primary objective is to investigate the benefits and harms of general strengthening exercise therapy in adults with chronic non‐specific low back pain on pain intensity, functional limitations, health‐related quality of life, psychological functioning, and adverse events compared to placebo, sham, or attention control and no treatment (including waitlists, control groups with no treatment provided, usual/normal care not controlled by the trial available to all treatment groups, or when the exercise and comparison groups receive the same co‐interventions).

Methods

Criteria for considering studies for this review

Types of studies

We will include published reports of completed RCTs with no restrictions on language or year of publication.

Types of participants

We will include RCTs involving adults (≥ 18 years) with CNSLBP (≥ 12 weeks). We will include RCTs with participants experiencing leg pain if CNSLBP is the primary complaint. Participants from any setting, including healthcare, occupational, general, and mixed populations, are eligible for inclusion.

We will exclude RCTs involving participants with LBP because of specific pathologies, including radiculopathy, spinal stenosis, piriformis syndrome, fracture, ankylosing spondylitis, spondyloarthritis, infection, neoplasm, or metastasis, pregnant or postsurgical people, and acute exacerbations of CNSLBP.

Types of interventions

We will include RCTs comparing general strengthening exercise therapy, if general strengthening exercise therapy is the primary part of the intervention, to (see Table 1 for comparison groups) (Hayden 2023):

1. Comparison types.

Category for analysis	Category extracted	Description or guidance
Placebo, sham, or attention control		The group is described as, or judged to be, a sham, placebo, or attention control by trial authors.
No trial treatment	Waitlist	The group is described as having received no treatment and to be assigned to waitlist.
	Control group	The group is described as no intervention or control group, or no other description is provided by the trial authors.
	Usual/normal care	The group received normal care, not controlled by the trial, which may have been offered to all treatment groups.
	Same co‐interventions	The exercise and comparison groups were offered, or received, the same co‐interventions, allowing the effect of general strengthening exercise therapy to be isolated.

1. placebo, sham, or attention control (which will be used for our primary analysis); and

2. no trial treatment (including waitlists, control groups with no treatment provided, usual/normal care not controlled by the trial available to all treatment groups, or when the exercise and comparison groups receive the same co‐interventions).

Types of outcome measures

We will extract all eligible outcomes measures as well as time periods and subsequently group them into:

1. short term, closest to three months after randomisation (6 to 12 weeks);

2. medium term, closest to six months after randomisation (13 to 47 weeks);

3. long term, closest to 12 months after randomisation (48 weeks or more).

We will use the time period closest to three months after randomisation (short term) for our primary analyses.

Major outcomes

Where available, we will extract the following outcomes based on the recommendations for the Core Outcome Set using the tools in the following hierarchy (Chiarotto 2018).

1. Pain intensity (visual analogue scale, numerical rating scale).

2. Functional limitations, i.e. back‐related disability (Roland‐Morris Disability Questionnaire, Oswestry Disability Index).

3. Health‐related quality of life (SF‐36 or SF‐12, PROMIS‐GH‐10 (10‐item Patient‐Reported Outcomes Measurement Information System Global Health short form), EQ‐5D. For SF‐36, SF‐12, or PROMIS‐GH‐10, the mental composite scores will be reported.

4. Psychological functioning, i.e. depressive symptoms (Beck Depression Inventory, Zung Depression Index, Patient Health Questionnaire‐9 (PHQ‐9), Montgomery‐Asberg Depression Rating Scale (MADRS), Hamilton Rating Scale for Depression (HRSD), Center for Epidemiologic Studies Depression Scale (CES‐D), Hospital Anxiety and Depression Scale (HADS), Hopkins Symptoms Checklist for anxiety and depression (HSCL)).

5. Adverse events.

Minor outcomes

We will describe whether the following minor outcomes are reported in the included RCTs.

1. Return to work/absenteeism.

2. Global improvement or perceived recovery.

3. Satisfaction.

4. General health.

5. Medication use.

6. Self‐efficacy.

7. Cost

However, we will not synthesise the results for the minor outcomes in this review.

Search methods for identification of studies

Electronic searches

The team behind the overarching network review will conduct electronic searches and search other sources to ensure adequate retrieval for the overarching network review and the focused sub‐reviews (Hayden 2023). We will search the following databases with no date or language restrictions (Hayden 2023):

1. Cochrane Central Register of Controlled Trials (CENTRAL) (Cochrane Library, Wiley) (latest issue);

2. MEDLINE (OvidSP) (1946 to current);

3. Embase (Embase.com) (1974 to current).

Citations will be managed using EndNote X8 software (Hayden 2023).

Searching other resources

We will evaluate any relevant systematic reviews identified by the electronic search for potentially eligible RCTs. We will search trial registrations and protocols in ClinicalTrials.gov (clinicaltrials.gov) and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (trialsearch.who.int). Finally, we will contact authors of all in‐progress RCTs via an emailed REDCap survey to determine if RCTs are completed and published. Before analyses, we will search for retractions and publication corrections within our set of eligible studies using the software Zotero, which is integrated with the Retraction Watch database (Hayden 2023).

Data collection and analysis

Selection of studies

Pairs of review authors from the pool of 35 team members of The Collaborative Review Model will independently conduct study selection based on title and abstract screening and subsequent full‐text review (Hayden 2023). We will resolve disagreements by consensus or by involving a third person. At the full‐text level, we will contact study authors when eligibility remains unclear. RCTs published in languages other than English will be assessed whenever feasible using English language abstracts, translation tools, and/or review by review team members and colleagues familiar with the language of publication.

Data extraction and management

For each RCT, a single review author from the pool of 35 team members of The Collaborative Review Model will extract data, and a second review author will conduct quality assurance by checking every extracted data point against the original study publication (see Table 2 for extracted data points) (Hayden 2023).

2. Data for extraction.

Domain	Item
Source	Study title, Year of publication, Review first author, and Contact details (i.e. corresponding author’s last name and email address)
Registration(s)	Trial registration, and Protocol publication.
Methods	Study design, Country of origin, and Study setting (i.e. healthcare, occupational, general, and mixed populations).
Number of participants	Total number of participants, and Number of participants allocated to the intervention(s) and control group(s)
Participant characteristics	Mean age, Sex (% of females), Pain duration (months), Leg pain (%), and Baseline levels of: Pain intensity, Functional limitations (i.e. back‐related disability), Health‐related quality of life, and Psychological functioning (i.e. depressive symptoms). Number of participants before randomisation and number of participants randomised to each group.
Interventions	Treatment descriptions of general strengthening exercise therapy CERT item #1: Exercise equipment, CERT item #2: Care‐provider (i.e. type and qualifications), CERT item #3: Individual or group exercises, CERT item #4: Supervised or unsupervised exercises, CERT item #5: Adherence, CERT item #6: Motivation strategies, CERT items #7a and 7b: Decision rules for progressing exercises, CERT item #8: Description of each exercise to enable replication CERT item #9: Home programme component, CERT item #10: Non‐exercise components (e.g. education, massage, etc.), CERT item #11: Adverse events, CERT item #12: Setting, CERT item #13: Detailed description of the exercises (e.g. sets, repetitions, duration, intensity), CERT items #14a and 14b: Whether exercises are generic or tailored to the individual, CERT item #15: Decision rule determining the starting level of exercises, and CERT items #16a and 16b: Whether the exercise intervention is delivered and performed as planned. Treatment description of the control (see Table 1 for details)
Outcomes	Preferably, mean and standard deviations (see "methods" concerning other measures of central tendency and dispersion, if mean and standard deviations are not available) of: Pain intensity, Functional limitations, Health‐related quality of life, and Psychological functioning Measured at: Short‐term follow‐up, i.e. closest to 3 months postrandomisation (6 to 12 weeks) Medium‐term, i.e. closest to 6 months postrandomisation (13 to 47 weeks) Long‐term, i.e. closest to 12 months postrandomisation (48+ weeks) Further, we will describe whether the following minor outcomes are reported in the included RCTs: Return to work/absenteeism, Global improvement or perceived recovery, Satisfaction, General health, Medication use, Self‐efficacy, and Costs.
Miscellaneous	Conflict(s) of interest, Funding source, and Research ethics review

We will extract the following study characteristics.

1. Methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and location, study setting, withdrawals, and date of study.

2. Participants: N, mean age, age range, sex, disease duration, severity of condition, diagnostic criteria, important (condition‐specific) baseline data; inclusion criteria and exclusion criteria.

3. Interventions: we will describe the reporting quality of general strengthening exercise therapy according to the Consensus on Exercise Reporting Template (CERT) (Slade 2016).

4. Outcomes: major and minor outcomes specified and collected, and time points reported.

5. Notes: funding for trial, and notable declarations of interest of trial authors.

Two review authors will independently extract data from the included RCTs and subsequently rate the reporting quality of each CERT item within each RCT. Each CERT item will be rated as 'yes' (if the item is clearly reported); 'no' (if the item is not reported); or 'unsure' (if we are unsure whether the item is reported and/or the item is reported but not clearly). Subsequently, each RCT will be given a total score ranging from 0 to 19 (0 = no items clearly described to 19 = all CERT items clearly described) (see Table 2 for extracted data points).

We will follow structured procedures for citation management, study selection, and data extraction facilitated by evidence synthesis tools and online software. New search results will be pre‐screened using Cochrane Screen4Me resources (Noel‐Storr 2021), including a machine learning algorithm, Thomas 2021, and Cochrane Crowd RCT classification (Noel‐Storr 2021). Subsequent study screening, data extraction, and risk of bias assessment will be facilitated by DistillerSR (Version 2.42) (Hayden 2023). We will extract the number of events and number of participants per treatment group for dichotomous outcomes and means, standard deviations, and number of participants per treatment group for continuous outcomes. Further,

1. if both final values and change from baseline values are reported for the same outcome, we will extract final values;

2. if both unadjusted and adjusted values for the same outcome are reported, we will extract adjusted values; and

3. if data are analysed based on an intention‐to‐treat (ITT) sample and another sample (e.g. per‐protocol, as‐treated), we will extract per‐protocol for outcomes assessing benefits, and extract ITT for outcomes assessing harms.

We will note in the 'Characteristics of included studies' table if outcome data were not reported in a usable way and when data were transformed or estimated from a graph. We will resolve disagreements by consensus or by involving a third person.

Assessment of risk of bias in included studies

Pairs of review authors will independently assess risk of bias in each study. We will resolve any disagreements by discussion or by involving another author from the review team. We will conduct risk of bias assessment using a modified version of the RoB 1 recommended by Cochrane, Higgins 2011, and Cochrane Back and Neck, Furlan 2015, to assess the following domains:

1. selection bias (randomisation, treatment allocation concealment, and similarity at baseline);

2. performance bias (blinding of participants and care provider as well as ITT);

3. detection bias (blinding of outcome assessors and similar timing of outcome assessment);

4. attrition bias (missing outcome data);

5. reporting bias (selective outcome reporting); and

6. other biases to assess adherence (avoidance of co‐interventions and compliance).

We will grade each risk of bias domain as low, high, or unclear, and provide a quote from the study report together with a justification for our judgement in the risk of bias table. We will summarise the risk of bias judgements across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the risk of bias table. When considering treatment effects, we will take into account the risk of bias in the studies that contribute to that outcome. We will present the figures generated by the risk of bias tool to provide summary assessments of risk of bias.

Measures of treatment effect

We will analyse dichotomous data (adverse events) as risk ratios (RR) and use 95% confidence intervals (CIs). We will estimate the treatment effects separately for pain intensity, functional limitations, health‐related quality of life, and psychological functioning as standardised mean differences (SMD) with 95% CIs (Higgins 2023). SMDs will be estimated and adjusted to Hedges g to compensate for the slight overestimation of small studies using Cohen's d, as multiple scales are expected to be reported across the RCTs (Higgins 2023). We will estimate SMDs as the difference between the final mean scores in the intervention and control groups divided by the pooled standard deviation (SD) (Higgins 2023). We will interpret SMDs as proposed by Cohen (Higgins 2023), considering:

1. 0.2 as a small effect;

2. 0.5 as a moderate effect; and

3. 0.8 as a large effect.

We will interpret clinically important difference as a difference of 10% for exercise treatment. This reflects a difference in pain intensity and functional limitations of 10 points out of 100 and is compatible with the smallest worthwhile effect based on a 20% reduction (Ferreira 2013). A valid clinically important difference does not seem to be available in the existing literature for health‐related quality of life and psychological functioning.

For dichotomous outcomes, we will calculate the absolute per cent change from the difference in risks between the intervention and control group using GRADEpro GDT software and expressed as a percentage (GRADEpro GDT). For continuous outcomes, we will calculate the absolute per cent change by dividing the mean difference by the scale of the measure and expressed as a percentage.

Unit of analysis issues

Where multiple trial arms are reported in a single trial, we will include only the relevant arms. If two comparisons are combined in the same meta‐analysis, we will halve the control group to avoid double‐counting (Higgins 2023a).

Dealing with missing data

We will request missing data from study authors via email with a link to a REDCap data capture form, where both extracted data and missing fields are displayed (Hayden 2023). If we are unable to contact the authors or they cannot supply missing data, we will impute missing variance scores using the mean‐variance from studies with similar populations of LBP. Where data in study papers are reported as a median and interquartile range (IQR), we will use the median to estimate the mean when sample sizes exceed 25 participants. In contrast, for studies with small sample sizes, we will use the formula proposed by Hozo and colleagues (Hozo 2005).

Where possible, we will calculate missing SDs from other statistics, such as standard errors, CIs or P values, according to the methods recommended in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2023).

Assessment of heterogeneity

We will statistically pool data if the RCTs are judged to be clinically homogeneous. We will assess statistical heterogeneity using I², which describes the percentage of the variability in effect estimates due to heterogeneity rather than sampling error (i.e. chance) (Deeks 2023). We will interpret I² as follows (Deeks 2023):

1. 0% to 40%: small heterogeneity;

2. 30% to 60%: moderate heterogeneity;

3. 50% to 90%: substantial heterogeneity; and

4. 75% to 100%: large heterogeneity.

Furthermore, we will estimate Tau² to express the absolute heterogeneity (i.e. the between‐study variance) (Deeks 2023).

Assessment of reporting biases

We will evaluate between‐study publication bias using Egger's test for continuous outcomes and visual inspection of funnel plots (Page 2023). Furthermore, we will assess within‐study publication bias by considering whether all expected outcomes were reported for each time period. If we suspect reporting bias, we will contact the study authors for additional information and attempt to locate the study protocol to determine if there are differences between the protocol and publication.

Data synthesis

We will conduct pairwise meta‐analyses for both comparisons (i.e. placebo, sham or attention control and no treatment) using a random‐effects model for all major outcomes (i.e. pain intensity, functional limitations, health‐related quality of life, psychological functioning, and adverse events) at the time point closest to three months. We will conduct pairwise meta‐analyses and meta‐regressions in RevMan, RevMan 2024, and Stata/MP, Stata 2023, respectively, using the DerSimonian and Laird method (Deeks 2023). We will conduct subgroup, meta‐regressions, and sensitivity analyses to explore the heterogeneity of the findings as suggested in the Cochrane Handbook (Deeks 2023). We will estimate prediction intervals to estimate the effects in individual study settings (Riley 2011). A forest plot for each primary outcome will supplement the results of the pairwise meta‐analysis.

Subgroup analysis and investigation of heterogeneity

To explore heterogeneity, we plan to conduct subgroup and meta‐regressions, as suggested in the Cochrane Handbook (Deeks 2023).

We plan the following meta‐regressions, adjusting for the following continuous variables:

1. sex distribution (% of females), as a 2023 review on gender‐related issues in the management of LBP concluded that epidemiology and clinical outcomes might depend on both sex‐ and gender‐related factors (Bizzoca 2023);

2. total number of exercise therapy sessions (number per week multiplied by programme duration in weeks);

3. baseline pain (on a standardised 0‐to‐100 scale, where 0 equals no baseline pain and 100 equals maximum baseline pain), as a 2021 review on prognostic factors for pain chronicity in LBP indicates that higher pain intensity might be related (Nieminen 2021);

4. baseline functional limitations (on a standardised 0‐to‐100 scale, where 0 equals no baseline functional limitations and 100 equals maximum baseline functional limitations), as a 2021 review on prognostic factors for pain chronicity in LBP indicates that higher levels of functional limitations might be related (Nieminen 2021).

We will conduct subgroup analyses on the placebo, sham, or attention control comparison and no treatment comparison for pain and function.

We will use the formal test for subgroup interactions in RevMan and use caution in the interpretation of subgroup analyses as advised in Chapter 10 of the Cochrane Handbook (Deeks 2023). We will compare the magnitude of the effects between the subgroups by means of assessing the overlap of the CIs of the summary estimate. Non‐overlap of the CIs indicates statistical significance.

Sensitivity analysis

We will conduct the following sensitivity analyses using standard meta‐analysis for each of the continous outcomes (i.e. pain intensity, functional limitations, health‐related quality of life, and psychological functioning), excluding RCTs where data have been imputed or where median values were presented.

Summary of findings and assessment of the certainty of the evidence

We will follow the guidelines in Chapters 14 and 15 of the Cochrane Handbook for interpreting results and will be aware of distinguishing a lack of evidence of effect from a lack of effect (Schünemann 2023; Schunemann 2023a). We will base our conclusions only on findings from the quantitative analysis of the studies included in this review. We will avoid making recommendations for practice, and our implications for research will suggest priorities for future research and outline the remaining uncertainties in the area.

We will create a summary of findings table using the following outcomes: pain intensity, functional limitations, health‐related quality of life, psychological functioning, and adverse events. The comparison in the first summary of findings table will be: placebo, sham, or attention control. The comparison in the second summary of findings table will be: no trial treatment.

Two review authorswill independently assess the certainty of the evidence, with any disagreements resolved by discussion or by involving a third review author. We will use the five GRADE considerations (study limitations (overall risk of bias), consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of a body of evidence as it relates to the studies which contribute data to the analyses for the prespecified outcomes, and report the certainty of evidence as high, moderate, low, or very low. We will justify, document, and incorporate our judgements into the reporting of results for each outcome.

We will use GRADEpro GDT software to prepare the summary of findings tables (GRADEpro GDT). We will use version 3 of the GRADEPro view to display our summary of findings tables which has the 'What happens' column. We will justify all decisions to downgrade the certainty of the evidence for each outcome using footnotes and make comments to aid the reader's understanding of the review where necessary.

Contributions of authors

Kasper Bülow contributed to the writing of the protocol draft.

Alessio Bricca contributed to editing and commenting on the protocol draft.

Søren Thorgaard Skou contributed to editing and commenting on the protocol draft.

Jan Hartvigsen contributed to editing and commenting on the protocol draft.

Alice Kongsted contributed to editing and commenting on the protocol draft.

Carsten Bogh Juhl contributed to editing and commenting on the protocol draft.

Sources of support

Internal sources

• No sources of support provided

External sources

• No sources of support provided

Declarations of interest

None to declare, as Kasper Bülow, Alessio Bricca, Søren Thorgaard Skou, Jan Hartvigsen, Alice Kongsted, and Carsten Bogh Juhl have no conflicts of interest.

New