Risk Stratification for Musculoskeletal Pain Did Not Improve Outcomes

Stratifying patients with common musculoskeletal presentations by risk in the primary care setting was not found to improve pain or functional outcomes, according to results of a study published in The Lancet Rheumatology.

The STarT MSK trial recruited 24 general practitioners (GPs) in the United Kingdom for this parallel-group, pragmatic, cluster-randomized controlled trial. Physicians who consulted for musculoskeletal presentations were randomly assigned in a 1:1 ratio to the intervention — which involved the use of a digital template to assist in the delivery of risk-based stratified care — or usual care. Patients who received risk-based care were stratified as being at low, medium, or high risk for poor disability, and 1 of 15 risk-matched treatment options were suggested. The primary outcome was pain intensity over 6 months as measured using the numeric rating scale (NRS).

The mean ages of participants in the intervention (n=534) and usual care (n=677) groups were 57.8 (standard deviation [SD], 15.3) and 61.8 (SD, 15.0) years (P =.004), respectively; 59% of participants in each group were women, nearly all study participants were White, and overall pain NRS scores were 6.8 (SD, 1.9) and 6.7 (SD, 2.0), respectively.

The location of pain most frequently reported was the back (36%-40%) and knee (29%-33%), followed by the neck (10%-11%) and shoulder (9%-13%). Multisite pain was reported in 6% to 13% of study participants.

Time-averaged pain intensity over the 6-month period was rated as 4.4 (SD, 2.3) among patients receiving the intervention and 4.6 (SD, 2.5) among patients in the usual care group (P =.535). Patients tended to report more pain in the final 3 months of the study compared with the first 3 months (P =.211).

No significant difference in minimal clinically important change in pain intensity, defined as a greater than 1 point change, was observed overall (odds ratio [OR], 1.66; 95% CI, 0.98-2.82; P =.061). Comparing month 4 to 6 indicated a significant clinically important change (OR, 2.22; 95% CI, 1.26-3.89; P =.0063).

Stratified by gender, a significant effect of the intervention on pain intensity was noted among men (mean difference [MD], -0.67; 95% CI, -1.27 to -0.07) but not among women (MD, 0.14; 95% CI, -0.42 to 0.69), indicating a significant gender interaction (P =.019).

Stratified by pain location, the intervention had a significant effect among patients with shoulder pain (MD, -11.10; 95% CI, -19.80 to -2.30; P =.013).

A review of medical records indicated that participation in the intervention group required some significant clinical decision-making changes compared with participation in the control group. More patients in the intervention cohort received written information, referral for physiotherapy, and over-the-counter analgesics. There was also noted to be an increase in short-term strong opioid prescriptions, which was not an intended outcome.

The cost of care was similar between the intervention and usual care groups (adjusted incremental cost, £6.85; 95% CI, -107.82 to 121.54).

A major limitation of this study was not evaluating whether patients pursued referral consultations or filled prescribed medications.

Study authors concluded that “risk-based stratified care in general practice for patients with common musculoskeletal pain presentations does not lead to significant improvements in patient outcomes, despite some benefits to GP decision making and positive GP and patient experiences of care.”