Objectives: To determine if subgroups of patients with tennis elbow respond differently in treatment.
Methods: This study used individual patient data (n = 383) from two randomized controlled trials that investigated a wait-and-see policy, corticosteroid injections and physiotherapy. Common outcome measures were: pain severity, global improvement, severity assessed by a blinded assessor, elbow disability and pain free grip strength. Subgroup analyses for previous history of elbow pain, baseline pain severity, duration of the current episode and employment status were performed at 6 and 52 weeks.
Results: Patients' age, previous elbow symptoms and baseline pain severity were similar between trials, but other characteristics differed between trial populations. Based on individual patient data from both trials, we found that corticosteroid injections were statistically and clinically superior at 6 weeks, but significantly worse at 52 weeks compared with both wait-and-see and physiotherapy. Subgroup effects were scarce and small. Patients with higher baseline pain score showed less benefit on pain outcomes between physiotherapy and a wait-and-see policy at 6 weeks. It also appeared that non-manual workers who had an injection were the only work subgroup to follow the general trend that injections were significantly worse than a wait-and-see policy on global improvement at 52 weeks.
Conclusion: The treatment outcomes were largely similar between trials and not different between most subgroups studied. In tennis elbow, it would appear that patient characteristics play only a small role in predicting treatment outcomes, which supports the generalizability of individual trial results.