Effect of Stratified Care for Low Back Pain in Family Practice (IMPaCT Back): A Prospective Population-Based Sequential Comparison

The recent publication by Foster and colleagues (1) is an important addition to the low back pain literature and I congratulate them on completing this large, well conducted study. The study is an important follow up after the previous RCT, (2) and while not a RCT itself the results are largely consistent with the previous RCT and provide evidence for the ability to implement the STarT approach in general medical practices.

There has been a great deal of interest in the STarT back approach since the publication in Lancet in 2011 and this is justified based on the promising findings. However, as Carey et al (3) conclude in their accompanying editorial "These advances don't solve our problems with the large disability burden and high costs of low back pain, but they represent a promising start". The current study demonstrates small pre- post intervention improvements in patient outcomes and only in the high risk group, which is the smallest group. There were no pre-post intervention improvements in pain or disability scores for the low or medium risk group. While this is an expected (positive finding) for the low risk group, it is a disappointing finding for the medium risk group (largest group). In both low risk and high risk groups less than 50% of patients achieved the MCID which is particularly concerning for the low risk group and somewhat questions the validity of considering them low risk. The positive findings from the trial are mostly economic with reductions in work absenteeism being arguably the most impressive (50%). The trial reported small to moderate pre-post intervention improvements in pain and disability in the high risk patients only.

So where does this leave researchers and clinicians in the search to improve outcomes through stratification or subgrouping of low back pain? The STarT tool is in my opinion a good starting point to be used in primary care by General Practitioners and other clinicians (e.g. physiotherapists). There is much more to be done in improving patient outcomes particularly in the moderate risk group. Other stratification or subgrouping approaches have been developed to improve outcomes in patients receiving physical therapy, and while at a preliminary stage of development and validation, some of these approaches have been shown to have favourable outcomes in single RCTs. (4-5) An important area for future research would be to test if incorporating these subgrouping approaches into the physical therapy management of patients with moderate and high risk profile (i.e further stratifying these patients), improves the outcomes of STarT approach, particularly from the perspective of patient outcomes.

Finally I have one last thought regarding the STarT tool itself. In the development of the STarT tool the authors made a conscious choice to focus on modifiable risk factors, so the tool could be used to guide intervention rather than simply identify likely outcome. While this is entirely logical an important finding from the STarT RCT (2) and current study (1) is the equivalent findings in patients at low risk when provided with minimal intervention. This is responsible for much of the economic benefits associated with STarT approach. With this in mind, identification of those at low risk (good prognosis) using factors which are non- modifiable (e.g. duration of episode) may have little down side. If patients classified as low risk are provided with minimal care then it seems the focus should be on most accurately identifying these people, rather than identifying modifiable risk factors. For example in a previous study of patients with acute low back pain presenting to primary care, we found 78% of people with short duration of episode (< 5 days), lower baseline pain intensity (<7/10) and limited number of previous episodes (<1) were recovered by 2 weeks compared to only 25% in people without any of these characteristics. (6) Inclusion of some of these non- modifiable characteristics into the STarT tool or at least using this information in clinical decision making should significantly improve accuracy in identifying those people likely to do well with minimal intervention.

References
1. Foster NE, Mullis R, Hill JC, Lewis M, Whitehurst DG, Doyle C, et al. Effect of Stratified Care for Low Back Pain in Family Practice (IMPaCT Back): A Prospective Population-Based Sequential Comparison. Annals of Family Medicine. 2014;12:102-11.
2. Hill JC, Whitehurst DG, Lewis M, Bryan S, Dunn KM, Foster NE, et al. Comparison of stratified primary care management for low back pain with current best practice (STarT Back): a randomised controlled trial. The Lancet. 2011;378:1560-71.
3. Carey TS, Freburger J. Physical Therapy for Low Back Pain: What Is It, and When Do We Offer It to Patients? Annals of Family Medicine. 2014;12:99-101.
4. Vibe Fersum K, O'Sullivan P, Skouen J, Smith A, Kvale A. Efficacy of classification-based cognitive functional therapy in patients with non-specific chronic low back pain: A randomized controlled trial. European Journal of Pain. 2013;17:916-28.
5. Brennan GP, Fritz JM, Hunter SJ, Thackeray A, Delitto A, Erhard RE. Identifying subgroups of patients with acute/subacute "nonspecific" low back pain: results of a randomized clinical trial. Spine. 2006;31:623-31.
6. Hancock MJ, Maher CG, Latimer J, Herbert RD, McAuley JH, Hancock MJ, et al. Can rate of recovery be predicted in patients with acute low back pain? Development of a clinical prediction rule. Eur J Pain. 2009;13:51-5.

Competing interests: None declared