TOPIC DEVELOPMENT

The VA has had a significant increase in requests for chiropractic care since these services became covered by the VHA. With an increased focus on interdisciplinary care within the VHA, findings from an evidence synthesis about the effectiveness of spinal manipulative therapy (SMT) or chiropractic care will help the VA identify approaches for treating acute neck and lower back pain and ensure the VA is providing Veterans with optimal healthcare services.

This report was developed based on a nomination from operational partners Lucille Beck, PhD, Deputy Chief Patient Care Services Officer, Rehabilitation and Prosthetic Services (10P4R), Anthony Lisi, DC, Director, VHA Chiropractic Service Rehabilitation and Prosthetic Services; Section Chief, Chiropractic Service, VA Connecticut Healthcare System; and David Chandler, PhD, Deputy Chief Consultant, Rehabilitation and Prosthetic Services (10P4R).

The proposed Evidence-based Synthesis Program (ESP) evidence synthesis will be used by the Office of Rehabilitation and Prosthetic Services (10P4R), Chiropractic Service, to inform VA clinical practice and national policy as the VA continues to implement chiropractic services across the country.

The Key Questions are:

Key Question 1.

What are the benefits and harms of spinal manipulation/chiropractic services for acute lower back pain (less than 6 weeks duration) compared to usual care or other forms of acute pain management?

Key Question 1A.

What is the relationship between the use of spinal manipulation/chiropractic services for lower back pain and the use of opiate medication?

Key Question 2.

What are the benefits and harms of spinal manipulation/chiropractic services for acute neck pain (less than 6 weeks duration) compared to usual care or other forms of acute pain management?

Key Question 2A.

What is the relationship between the use of spinal manipulation/chiropractic services for acute neck pain and the use of opiate medication?

The PROSPERO registration number is CRD42015017916.

SEARCH STRATEGY

Spinal manipulation is a topic that has been the subject of numerous prior systematic reviews, including 3 reviews by members of the ESP review team. Therefore, instead of searching for original evidence in databases such as PubMed, we instead began by reference mining existing systematic reviews, and then performing an update search to identify new studies published since the end date of the searches of the most recent reviews. Then we consulted our technical experts for any additional studies we might have overlooked. See Appendix A for full search strategy.

STUDY SELECTION

All reference titles and abstracts were screened independently by 2 reviewers. If either reviewer selected a title or abstract, it was included for further review. Full-text articles were then reviewed in duplicate, with all discrepancies discussed with the group. References were selected based on the following inclusion criteria:

Participants: Adults (ages 18 and older) with acute (defined as 6 weeks or less) neck or lower back pain. Patients with sciatica were included. Studies of patients with chronic back pain were excluded, as were studies where we could not determine the duration of pain. If studies included patients with longer durations of pain, we included them if they presented stratified results or if the majority of patients had pain for less than 6 weeks duration. Studies of children were excluded; this included pediatric populations or patients under the age of 18.

Intervention: Spinal manipulation by any provider type. Studies where spinal manipulation was given alone or as part of a “package” of therapies were included. “Chiropractic care” was considered as including SMT for the great majority of patients. The definitions of SMT types were refined during the data abstraction process, and a more detailed description of the intervention is given in the following Data Abstraction section.

Comparator (study design): Other forms of management for acute pain, such as analgesics, exercises, physical therapy, etcetera. Sham-controlled studies were included.

Outcome: Pain management, functional status, quality of life, opiate use, disability claims, return to work, health care utilization.

Timing: Studies had to report at least one outcome within 6 weeks to be eligible.

Setting: Ambulatory/outpatient settings. Studies in hospital settings were excluded.

Study design: Only randomized controlled trials (RCTs) were eligible for assessing benefits. Both RCTs plus observational studies were used for assessing harms.

DATA ABSTRACTION

Data were extracted by 2 reviewers, and discrepancies were reconciled after discussion. Articles had data abstracted on the anatomical location of the pain, authors’ description of the SMT provided, type of professional performing the treatment, co-interventions, whether that treatment was provided alone or as part of a package of other treatments, whether patients were selected as more likely to respond to SMT or unselected, data on any of the outcomes listed above (eg, pain, functional status, etc), as well as data needed to complete the Cochrane Back Group Risk of Bias assessment. For studies included in our own prior reviews of SMTs we used data abstracted from those studies at that time.

Based on the authors’ description of the SMT provided, we categorized the study as using thrust or non-thrust technique. Thrust was defined as high velocity low amplitude (HVLA), such as “a short-lever, high-velocity thrust directed specifically at a ‘manipulable lesion.’”³ Non-thrust was defined as low velocity high amplitude (LVHA), such as one study where “most participants had several low-velocity mobilization techniques.”⁴ Any studies where the research team had questions were brought to the Technical Expert Panel (TEP) for their input. In one case, we contacted the original author to clarify this (John Triano, personal communication, October 28, 2015).

Statistical data were extracted by the project statistician. We focused on data from follow-ups less than 6 weeks. For continuous outcomes, the sample size, mean, and standard deviation were extracted for each SMT group and comparator group within each trial. For count data, odd-ratios (OR) were extracted if means were not reported. Since pain and function scales often differed across studies, a standardized mean difference (SMD) was calculated between the SMT group and each comparator group. A negative SMD indicates that the SMT group is doing better at follow-up than the comparator group. The few studies that reported ORs were converted into SMDs and combined with the continuous trials.

QUALITY ASSESSMENT

We assessed the quality of studies using the Cochrane Back Group Risk (CBG) of Bias Tool (ROB) (see Appendix B for full tool). This tool has 11 items in the following domains: randomization, concealment, baseline differences, blinding – patient, blinding – care provider, blinding – outcome, co-interventions, compliance, dropouts, timing, and intent to treat. Prior research has shown the CBG ROB Tool to identify studies at an increased risk of bias using a threshold of 5 or 6 as a summary score.⁵

DATA SYNTHESIS

We constructed evidence tables showing the study characteristics and results for all included studies.

Studies were pooled within outcome measures and 95% confidence intervals were constructed. Studies using a 100 mm Visual Analog Scale or 11-point Numeric Rating Scale or other numeric pain scale were pooled together by converting all outcomes to a 0–100 measure (using the appropriate multiplier); studies reporting the Roland Morris Disability Questionnaire (RMDQ, scored on a 0–24 scale) and studies reporting the Oswestry Disability Index (ODI, scored as 0–100) were pooled as a functional outcome using an effect size approach. Studies reporting none of these were not pooled, but discussed narratively.

Random effects meta-analyses were conducted using the Hartung-Knapp Method.^6,7 Tests of heterogeneity were performed using the I² statistic.⁸ All meta-analyses were conducted with Stata statistical software, version 12.0⁹ and R3.2.2. The Begg’s rank correlation¹⁰ and Egger regression asymmetry test¹¹ were used to examine publication bias. To further explore possible sources of heterogeneity, such as timing, outcome, type of practitioner, and type of manipulation, bivariate meta-regressions were conducted.

The meta-analyses were organized based on 2 follow-up times and the 2 outcomes. Two studies^12,13 were in the gap between immediate and short-term outcomes; they were closest to immediate-term so they were classified in the immediate-term group. Within these 4 groupings the intervention was assessed in comparison to control interventions classified as either sham SMT or all other therapies. ¹⁴ Studies comparing SMT to sham-SMT were not pooled with studies comparing SMT to other therapies. Studies were included in each pooled analysis only once.

An a priori analysis considered 3 potential sources of heterogeneity: the comparison group, the outcome, and the timing of the outcome. In addition, 3 post-hoc hypotheses were developed to test possible explanations for observed heterogeneity: by type of manipulation, comparing thrust techniques to non-thrust techniques; by the types of patients enrolled (selected or not selected); and by study quality, comparing higher-quality trials to lower-quality trials.

The Minimum Clinically Important Difference (MCID) for either pain or function in acute low back pain is not well-established empirically. The MCID for the Roland Morris Disability Questionnaire has been proposed as low as 1.5 points and as high as 5 points. A recent systematic review of studies of the minimum clinically important difference for pain scales in acute pain concluded that no single value could be supported.¹⁵ Therefore, we have not chosen a MCID value for pain or function, but frame our results as a range or in comparison to other treatments for acute low back pain.

RATING THE BODY OF EVIDENCE

The evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria, which uses the domains of study design limitations, inconsistency, indirectness, and imprecision in results.¹⁶ The GRADE Working Group classified the quality of evidence across outcomes according to the following criteria:

High: We are very confident that the estimate of effect lies close to the true effect for this outcome.

Moderate: We are moderately confident that the estimate of effect lies close to the true effect for this outcome.

Low: We have limited confidence that the estimate of effect lies close to the true effect for this outcome.

Insufficient: We have no evidence, we are unable to estimate an effect, or we have no confidence in the estimate of effect for this outcome.

PEER REVIEW

A draft version of this report was reviewed by 3 technical experts and 2 members of VA operations. Reviewer comments were addressed and out responses were incorporated into the final report. The complete set of comments and responses can be found in Appendix C.