Computer-based stratified care in general practice for common musculoskeletal consultations: Results of the STarT MSK cluster randomised controlled trial (ISRCTN15366334)

Keywords: Stratified care; Randomised controlled trial; Clinical decision support tool Purpose: The Keele STarT Back approach has been shown to be effective for low back pain in UK primary care. This trial aimed to determine whether stratified care delivered by general practitioners (GPs) for patients with the most common musculoskeletal (MSK) pain presentations was more effective and cost-effective than usual non-stratified care. Methods: This was a pragmatic, two-arm parallel (stratified vs non-stratified care), cluster randomised controlled trial, with embedded health-economic and qualitative studies. 24 UK general practices were randomised in a 1:1 ratio (12-intervention: 12-control) with blinding of statisticians and outcome assessors. Units of observation were adult MSK consulters without indicators of serious pathologies, urgent medical needs, or vulnerabilities. GPs delivered stratified care using computer templates containing the Keele STarT MSK tool (classifying low-, medium- or high-risk of persistent pain) and recommended matched treatment options. The primary outcome using intention-to-treat analysis was time-averaged pain intensity (0-10 numerical rating scale) measured monthly over 6-months. Secondary outcomes included physical function, quality-of-life and satisfaction with care. An anonymised electronic medical record audit examined changes in clinician decision-making such as prescribing, referrals and imaging. The economic evaluation estimated healthcare cost differences and incremental cost-utility (cost per quality-adjusted-life-year (QALY) gained. Patient and clinician experiences were explored using qualitative interviews and focus groups. Results: Medical record data were available for 2494 patients (recruited May-2018 to April-2019). Of these, 1211 (49%) consented to patient-reported data collection via questionnaires, with follow-up data available for 88.5%. Mean age was 60 years (range 18-95), 58.9% were female. There were no signs of baseline selection bias. There were no statistically significant differences in time-averaged pain intensity over 6-months; mean (SD) values: 4.4 (2.3) intervention and 4.6 (2.4) usual care; adjusted mean difference of -0.16 (95%CI -0.65, 0.34) (p = 0.535). There were also no significant differences in any 6-month secondary outcomes; except stratified care significantly improved shoulder pain and function for patients with that pain presentation (p = 0.013) and led to greater satisfaction with care overall (p = 0.033). Exploratory subgroup analyses (for which the trial was not powered) showed benefits for patients at high risk. A number of significant changes in GP clinical decision-making were observed, including less imaging, and more over-the-counter analgesics, written information, and early referral to physiotherapy. The economic evaluation showed costs of care were similar in each trial arm (adjusted incremental cost of stratified care compared with usual non-stratified care was £6.85 (-107.82, 121.54)), with a small QALY gain of 0.0041 (-0.0013, 0.0094) and 73% probability of cost-effectiveness at a threshold of £20,000/QALY. Interviews revealed stratified care informed decision-making, e.g. greater attention on psychosocial issues and function; supported patient negotiations concerning imaging; and facilitated a holistic approach and interdisciplinary working/communication. Conclusion(s): GP delivered stratified care for common MSK presentations does not lead to superior clinical outcomes. It does significantly change GP treatment decision-making, increases patient satisfaction and provides positive clinician experiences. Impact: First Contact Physiotherapists in general practice may be considering implementing risk-stratified MSK consultations. This trial suggests the STarT MSK model of stratified care tested improves some aspects of clinical decision-making but not clinical outcomes. Funding acknowledgements: This paper presents independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (grant number: RP-PG-1211-20010) and the Center of Excellence funding from Versus Arthritis (grant reference: 20202). The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. Nadine Foster is a NIHR Senior Investigator and was supported through an NIHR Research Professorship (NIHR-RP-011-015).