Interpreting an open-label collagen pilot before squad-wide adoption
A sports medicine team completed a 12-week pilot study of 15 g collagen plus 50 mg vitamin C daily for patellar tendinopathy in 28 jumping athletes. The intervention was open-label, and usual care was compared with collagen plus usual care. Seventeen athletes completed all visits; adherence in the collagen group averaged 62%. The primary outcome was VISA-P at 12 weeks, but the team also analyzed pain during single-leg decline squat, countermovement jump, and training availability at 4, 8, and 12 weeks. At 12 weeks, the collagen group improved by 9.8 VISA-P points more than usual care (95% CI −2.1 to 21.7), and the jump outcome favored collagen. The coach has seen the result and wants collagen for the whole squad. The team has limited funding, one research assistant, and access to a statistician. You are asked to interpret the pilot, identify the main threats to validity, and advise on next steps.
1. What can be concluded from this pilot, and what threats limit causal inference?
Read the worked answer
Interpretation. The pilot is hypothesis-generating. The between-group difference of 9.8 VISA-P points is promising but the 95% CI crosses zero (−2.1 to 21.7), so a clinically important effect is compatible with no effect. The open-label design allows response bias, and the unblinded therapist may influence usual care. Only 17 of 28 completed all visits, and collagen adherence was 62%, so both attrition and incomplete exposure threaten validity. Multiple outcomes and repeated time points increase chance findings. The jump result is supportive but not definitive. The coach should not mandate collagen for the squad on this evidence. It is reasonable to discuss a time-limited individual trial with shared decision-making, while correcting the study design and not replacing loading-based rehabilitation.
2. Design the next study to answer whether collagen improves tendon pain.
Read the worked answer
Next study. Use a randomized, placebo-controlled, assessor-blinded design with concealed allocation and identical sachets. Pre-specify one primary outcome (VISA-P at 12 or 24 weeks) and a sample size based on the minimal clinically important difference, allowing for 20% attrition. Use intention-to-treat and per-protocol analyses, record adherence objectively, and standardize usual care. Stratify by baseline severity and sport. Include adverse events and blinding checks. If funding is limited, a pragmatic trial at one site with blinded outcome assessment is better than another open-label pilot. Register the protocol and report confidence intervals, not just p-values.
3. How should you communicate the findings to coaches and athletes now?
Read the worked answer
Communicate now. Tell coaches and athletes that the pilot is not proof and that the confidence interval includes no benefit. Avoid banning collagen, but do not present it as standard of care or a squad-wide requirement. Explain the adherence and dropout problems and the need for placebo controls. Offer an individual decision aid: potential low risk, uncertain benefit, cost, and the importance of continued tendon loading. Share a one-page summary with limitations and the next study plan. Monitor for supplement contamination and batch issues. The sport scientist’s credibility depends on resisting pressure to overclaim from a small open-label study.