Evidence for a time-dependent pain response to platelet-rich plasma versus corticosteroid injection in lateral epicondylitis: a cluster-stratified systematic review and meta-analysis of randomized controlled trials.
Abderraouf Elhadi, Anthony Kamel, Ernst Chavannes et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis of six randomised controlled trials examined how pain outcomes after platelet-rich plasma (PRP) versus corticosteroid injection for lateral elbow tendinopathy varied with follow-up time. The trials included 338 participants (180 PRP, 158 corticosteroid). Corticosteroid was favoured at four weeks or earlier (MD +1.88; 95% CI 1.05-2.71), while PRP was favoured beyond six months (MD -1.49; 95% CI -2.08 to -0.90). Meta-regression supported a time-dependent trend, although the authors describe the trajectory as hypothesis-generating.
Key findings
- At ≤ 4 weeks, corticosteroid was favoured (MD + 1.88; 95% CI 1.05-2.71; p = 0.006), exceeding the 1.5-point MCID.
- At 5-11 weeks, the estimate favoured corticosteroid but was non-significant (+ 1.62; - 1.69 to 4.93; p = 0.170).
- At 12-23 weeks, the estimate crossed zero (- 0.71; - 1.79 to 0.37; p = 0.152); beyond six months, PRP was favoured (- 1.49; - 2.08 to - 0.90; p = 0.001).
- Meta-regression found a time trend (slope - 0.087 per week; p = 0.003); certainty was moderate for early and late clusters and low for intermediate clusters.
Why it matters
The findings indicate that comparative pain outcomes may differ by follow-up period: early results favour corticosteroid, whereas later results favour PRP. This temporal pattern is relevant when interpreting trial results or discussing the evidence, since a single follow-up point may not represent the full course of comparative outcomes. The pooled estimates and certainty ratings should be considered alongside the authors’ caution about the analysis.
What to keep in mind
Repeated observations from the same trials across follow-up clusters violate independence assumptions, making the meta-regression descriptive and hypothesis-generating. The within-trial concordance analysis did not reach significance (p = 0.063).