Article Text
PDF
Abstract
Introduction Subgroups for Targeted Treatment Back Tool (SBT) is a brief multiple-construct risk prediction tool for patients with low back pain (LBP). Thus far, the predictive ability of this tool has been inconsistent. Therefore, we aim to conduct a literature review on the predictive ability of the SBT to determine the outcomes of patients with LBP. The results of this review should improve the ability of the SBT to predict poor outcomes in patients with LBP.
Methods and analysis Databases including PubMed, EMBASE, Cochrane Central, Web of Science, Chinese National Knowledge Infrastructure Databases, Chinese Science and Technology Journal Database, and Wanfang will be searched for studies on SBT and LBP from their inception until 31 March 2023. Longitudinal studies investigating the association between SBT subgroups and LBP outcomes, including pain, disability and quality of life, will be included. The identified studies will be independently screened for eligibility by two reviewers. A standardised sheet will be used to extract data. The Newcastle-Ottawa Scale will be used to assess the methodological quality of the included studies. Heterogeneity will be evaluated by the χ2 test with Cochran’s Q statistic and quantified by the I2 statistic. The results will be synthesised qualitatively and presented as pooled risk ratios or beta coefficients quantitatively. The results will also be presented using their 95% confidence limits. Publication bias will be assessed using the method proposed by Egger and by visual inspection of funnel plots.
Ethics and dissemination This study is a secondary analysis of original studies that received ethics approval. Therefore, prior ethical approval is not required for this study. The findings will be submitted to relevant peer-reviewed journals for publication and presented at profession-specific conferences.
Trial registration number PROSPERO registration number
CRD42022309189.
- Protocols & guidelines
- Risk management
- Back pain
- PAIN MANAGEMENT
- REHABILITATION MEDICINE
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
Four major databases (PubMed, EMBASE, Cochrane Central and Web of Science) and three Chinese literature databases will be comprehensively searched.
Outcomes will include risk ratios and beta-coefficients.
Separate subgroup analyses will be conducted.
The Newcastle-Ottawa Scale will be used to ensure synthesis output confidence.
Heterogeneity across studies is still a concern owing to differences in follow-up, patient characteristics, measurement tools used and poor outcome definitions.
Introduction
Low back pain (LBP) is a leading cause of long-term disability and musculoskeletal disorders, and it limits effective participation in activities of daily living.1 2 Non-pharmacological interventions are recommended as first-line treatment for LBP.3 Even so, there is much variance in physical therapy outcomes, and nearly half of the patients fail to see significant improvements in their pain and disability.4 The poor outcomes of physical therapy may be owing to the complex, multifactorial, and biopsychosocial nature of LBP.5 It is proposed that treatment outcomes will improve by identifying prognostic factors for poor outcomes and implementing target interventions involving modifiable factors.6 Psychological factors, such as depressive symptoms, fear of movement and pain catastrophising, are associated with poor outcomes.7 Consequently, reducing these factors may improve patient outcomes.7 8 Administering multiple, full-length risk-factor questionnaires is burdensome and remains a barrier to assessors and patients.9
The Subgroups for Targeted Treatment Back Tool (SBT) is an easy-to-use questionnaire comprising nine items. It assesses modifiable factors that correlate with unfavourable prognoses in LBP patients. These modifiable factors include physical (comorbid pain, leg pain and disability) and psychosocial (catastrophising, bothersomeness, kinesiophobia, anxiety and depression) factors.10 The tool was initially developed by Hill et al to stratify patients into three risk subgroups based on persistent disability and to suggest matched levels of care.10 11 Patients with a score between 0 and 3 are classified as low risk, and their care consists of personalised consultation, including advice and education. Patients with a score above 3 and a psychosocial subscale score below 4 are classified as medium risk. The care of medium-risk patients includes exercise, manual therapy, and advice and education. Patients whose score increases to 4 or 5 points on the psychosocial subscale are classified as high risk. These patients receive psychological treatment in addition to the medium-risk treatment package. This brief multiple-construct risk prediction tool has a greater predictive strength and scope than several single-construct measures.12 Therefore, it is recommended by the National Institute for Health and Care Excellence guidelines as a screening instrument to identify patients at risk for unfavourable prognoses.13
The number of studies based on the predictive value of SBT has steadily increased beyond the clinical setting, outcome and participants in the original study. In addition to verifying Hill’s study on patients coming from primary care in different countries,14 15 patients visiting other clinical settings, such as physical therapy care,16–18 chiropractic practice settings,15 19 spinal care clinics20 21 and emergency department,22 23 were investigated. The results varied across care settings. Previous studies revealed that the prognostic ability of SBT also varied with the outcome, including pain intensity.19 24 The duration of symptoms in patients with LBP in most studies investigating SBT varied from acute to chronic.14 25 26 However, the predictive performance of the SBT was challenged when very acute patients were assessed.15 Therefore, a summary of all existing evidence is required to guide the clinical assessment and treatment of patients with LBP. To the best of our knowledge, a meta-analysis exploring the generalisability and utility of the SBT in different LBP patient groups, care settings and clinical outcomes has yet to be conducted.
Objectives
This protocol aims to provide a detailed plan for conducting a systematic review and meta-analysis that involves identifying, evaluating and summarising literature published on the relationships between SBT and the outcomes of patients with LBP. The systematic review and meta-analysis aims to determine whether patients with LBP classified as medium risk or high risk via the SBT have a higher risk of poor outcomes, such as pain, disability and decreased quality of life, compared with those in low-risk patients.
Methods
This systematic review and meta-analysis will adhere to the Meta-analysis Of Observational Studies in Epidemiology guidelines.27 The current protocol was drafted according to the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P)28 29 with PROSPERO registration number (CRD42022309189). The PRISMA-P checklist was attached as online supplemental file 1.
Supplemental material
Search strategy for identification of studies
The systematic screening of studies will be conducted on seven electronic English and Chinese databases, including PubMed, EMBASE, Cochrane Central, Web of Science, Chinese National Knowledge Infrastructure Databases, Chinese Science and Technology Journal Database and Wanfang, from their inception to 31 March 2023. The search strategy comprises two domains of interest: (1) LBP and (2) SBT. The search terms will be defined for each domain using MeSH terms, entry terms, related synonyms, abbreviations and spelling variations separated with the ‘OR’ operator. The Boolean operator ‘AND’ will combine search terms for the domains. This search strategy is first developed for PubMed. An adjusted search strategy specific to the database’s features will be applied to other databases. To ensure enough coverage of the search, references for the included publications and pertinent reviews will be screened to identify additional pertinent studies. Prior to the final publication, the databases will be searched once more to identify studies that may have been published while the systematic review is being performed. The search strategy was presented in online supplemental file 2 and flow chart was illustrated in figure 1.
Supplemental material
Flow diagram of literature selection.
Criteria for inclusion in this review
Population
The inclusion criteria for this systematic review will be patients aged 18 years or older with non-specific LBP with or without leg pain. Non-specific LBP will be defined as LBP not attributable to a specific pathophysiological mechanism,30 such as infection, rheumatoid arthritis, osteoporosis, fracture, herniated nucleus pulposus or tumour. LBP will be classified based on its duration as acute (<6 weeks), subacute (6–12 weeks) and chronic (>12 weeks).30
Exposure
This systematic review will include studies on SBT subgroups. Patients will be further classified based on their SBT results as medium risk or high risk.
Comparison
A comparative analysis will be performed on SBT subgroups (low, medium and high risk) in this systematic review.
Outcomes
The outcomes of this systematic review will include the risk ratio or beta coefficient for poor outcomes such as pain, disability and quality of life.
Study design
The study will include prospective and retrospective longitudinal studies focused on the relationship between SBT subgroups and LBP outcomes.
Setting
Restrictions will not be placed on the care setting or location of the study.
Time frame
This systematic review will include studies published in English or Chinese before 31 March 2023.
Exclusion criteria
The exclusion criteria are as follows: (1) patients who are pregnant, have failed back surgery syndrome, postpartum back pain and chronic diffuse pain; (2) studies such as conference abstracts, letters, notes, grey literature, comments, guidelines, protocols, editorials and case reports; (3) inability to independently extract data on LBP from a study that investigated more than one musculoskeletal pain disorder and (4) studies with less than 3 months of follow-up.
Study selection process
The literature search results will be imported into EndNote (V.X9.1). After removing duplicate records, two reviewers (YF and SL) will independently screen titles and abstracts to exclude irrelevant studies based on the exclusion criteria. The full text of all potentially relevant studies will be downloaded. Two reviewers (YF and SL) will independently assess the full texts for inclusion using a piloted standardised eligibility sheet and record the reasons for exclusion to construct a flow chart. The results of the reviewers will be compared, and any discrepancies will be resolved through discussion and consensus from a third researcher (JC). The study with the largest sample size will be included in cases of duplicate studies or studies published in multiple reports. Emails will be sent to the corresponding author(s) when the full text of a study is inaccessible or a figure instead of a table is used to show the results of a study. Emails that fail to receive a response within 15 days will be resent. Studies with inaccessible data without an email response 15 days after the second email requesting access was sent will be excluded.
Data extraction and management
Two reviewers (YF and SL) will independently extract data. A third investigator (JC) will evaluate their data. The following data will be extracted using Microsoft Excel: (1) the data source, including the name of the electronic database names or custom source; (2) the study information, including the title, authors, year of publication, journal, study design, care settings, funding sources, country of sampling, aims, inclusion and exclusion criteria, intervention (type, frequency, course), outcome measures (pain intensity, disability, quality of life), tools used to assess the outcomes (such as the numerical rating scale for pain, the Rolande-Morris Disability Questionnaire for disability, the SF-36 for quality of life), the definition for poor outcomes, follow-up time, time of SBT assessment, statistical analysis methods, adjusted confounders (age, sex, education, baseline pain, baseline disability) and effect sizes (β, RR, OR); (3) the total sample information, including the sample size, age, sex, education, employment, duration, baseline pain and baseline disability and (4) SBT subgroup information, including subgroup sample size, age, sex, education, duration, baseline pain, disability in each subgroup and the number of patients with poor outcomes.
When the OR or RR values could not be extracted directly, the number of exposure cases and the total number in each group will be used for the calculation.
Quality assessment
Two independent reviewers (YF and SL) will use the Newcastle-Ottawa Scale (cohort studies) (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp) to assess the methodological quality of the included studies. This tool contains eight items that assess the three following domains: (1) selection, (2) comparability and (3) outcomes. The maximum scores for the above domains are 4, 2 and 3 stars, respectively. The risk of bias in each study will be graded as good, fair or poor.31 Good quality means that at least 3, 1 and 2 stars are assigned to the selection, comparability and outcome domains, respectively. When the selection domain receives 2 stars, and the other domains receive stars similar to good quality, the risk of bias is determined to be fair quality. Poor quality refers to a selection domain below two stars, a comparability domain below one star, or an outcome domain below two stars. Final judgements will be made by a consensus between the reviewers. The assessment results for eligible studies will be graphically presented.
Statistical analysis
Crude and adjusted estimates will be pooled independently. The pooled estimates will be presented as the risk ratio (binary outcomes) or beta coefficient (continuous outcome) and their 95% confidence limits. Meta-analysis will be performed using the Stata software (release V.16, Stata) with a random-effects model. We will evaluate the magnitude of heterogeneity using the χ2 test on Cochran’s Q-statistic and I2 statistic. A p<0.05 shows significant statistical heterogeneity, and an I2 statistic greater than 50% will mean substantial heterogeneity. When sufficient data are available, subgroup analysis will be performed on studies with outcomes adjusted for variables such as age, sex, education and baseline scores for pain or disability. Subgroup analyses will be conducted according to the care setting that the patients visited (primary care, physical therapy care, chiropractic practice settings, spinal care clinic and emergency department), the duration of the disease (acute, subacute and chronic) or outcome measures (pain, disability and quality of life). Meta-regression and sensitivity analyses will be conducted to explore sources of heterogeneity when the adjusted outcomes are insufficient or unavailable. Meta-regression will also be used to assess the role of age, sex, education and baseline pain or disability scores in determining the association between SBT subgroups and poor outcomes. Publication bias will be assessed by applying the method proposed by Egger and by visual inspection of funnel plots. If quantitative synthesis is inappropriate, such as in cases of substantial clinical heterogeneity or insufficient data, a systematic narrative synthesis will be accomplished with a summary table that presents the characteristics and findings of the included studies. Inter-rater agreements between the investigators will be assessed with Cohen’s κ coefficient.
Patient and public involvement
No patient involved.
Discussion
The findings of the current systematic review will reveal the ability of the SBT to predict poor outcomes in patients with LBP. The relative risk for poor outcomes in the medium-risk and high-risk subgroups, with reference to the low-risk subgroup, might differ based on clinical settings, duration of LBP and outcome measures.
Previous studies suggested that differences in the predictive ability of the SBT in primary and secondary care settings were attributable to differences in casemix.25 These studies showed that LBP patients who visited secondary care settings tended to have more intense pain, longer histories and higher incidences of referred leg pain than those who visited primary care settings.32 Another study failed to show an increased risk for poor outcomes between SBT subgroups of LBP patients seeking chiropractic therapy.19 The findings showed that a small number of LBP patients who visited chiropractors had adverse psychological factors linked to poor prognosis.33 A further research suggested that the duration of incidences of LBP varied between patients who visited different care settings; these differences affected the ability of the SBT to predict the outcome.15 The relationships between outcome measures of LBP and the SBT are not parallel.34 The fact that the predictive ability of SBT for disability, pain and global perceived change were discrepant from each other is an example of this.24
The SBT was developed to stratify care based on the results of SBT risk grouping.10 Caring for patients with LBP based on the result of the SBT optimises the use of medical resources. It is particularly important to determine whether additional psychotherapy is necessary for high-risk patients since psychological interventions tend to be expensive. In some conditions, psychological distress can be resolved to some degree by a treatment that does not involve psychological intervention.35 36 Studies have shown that LBP patients in the high-risk subgroup receive more improvements than those in the medium-risk or low-risk subgroup after a non-psychotherapeutic intervention,18 19 and can convert to the medium-risk or low-risk subgroup.23 Therefore, monitoring the patient serially with the SBT during the treatment was suggested.16 23 Unfortunately, there is no consensus on when or under what circumstances the SBT should be re-evaluated.16 23
The findings of the relative risks of unfavourable outcomes in the three subgroups assessed by the SBT under differing conditions will improve the credibility of the evidence and assist in elucidating inconsistencies identified in previous findings.37–39 Moreover, the results of this systematic review might provide a fresh perspective for investigating no difference in the efficacy of using the same intervention for medium-risk and high-risk groups in a specific condition. A meta-analysis and systematic review of studies that test the predictive performance of SBT in patients with LBP is warranted as they will facilitate the development of treatment plans based on predictions made by the SBT.
Ethics and dissemination
This study is a secondary analysis of original studies that received ethics approval. Therefore, prior ethical approval is not required for this study. The findings will be submitted to relevant peer-reviewed journals for publication and presented at profession-specific conferences.
Review status
Piloting of the full-text screening process.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors YF and L-HH contributed to the idea. YF, SL and YC designed and conceived the protocol. YF and JC developed the database search strategy. YF drafted the manuscript. YF, SL and YC revised the manuscript. YF and H-LH are the guarantors of the review. All authors approved the submission of the final version of manuscript.
Funding This work was supported by Natural Science Foundation of Fujian Province, China (Grant No. 2021J05071) and Startup Fund for Scientific Research, Fujian Medical University (Grant No. 2020QH1144). No funding body involved in any aspect of this project.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.