Abstract
BACKGROUND: The effectiveness of nonsurgical treatment of patellar tendinopathy is questioned due to the conflicting results of placebo-controlled randomized controlled trials (RCTs) in which placebo arms often show impressive results. PURPOSE: To quantify the magnitude of placebo effect of the different nonsurgical treatments of patellar tendinopathy. We also evaluated the influence of patients and treatments characteristics on the response to the placebo. STUDY DESIGN: Systematic review; Level of evidence, 1. METHODS: We searched PubMed, Web of Science, Embase, Scopus, Cochrane Library, and gray literature databases on May 10, 2023, with no time limitation. RCTs on nonoperative treatment of patellar tendinopathy, including a placebo control arm reporting the evolution of symptoms after placebo administration, were included. A single-arm meta-analysis was performed with the Victorian Institute of Sport Assessment-Patella (VISA-P) at mid-term follow-up (3-6 months) as the primary outcome. The VISA-P score at short-term (1-3 months) and long-term (6-12 months) follow-ups, as well as visual analog scale (VAS) for pain at all 3 time points were also analyzed. A subanalysis based on the type of placebo and a meta-regression were conducted to look for potential determinants of the placebo effect. Risk of bias and level of evidence were also analyzed using the revised tool for risk of bias in randomized trials and Grading of Recommendations Assessment, Development and Evaluation. RESULTS: In total, 14 studies (251 patients) were included. VISA-P score at mid-term follow-up (3-6 months) showed statistically significant improvements of 13 of 100 points (P = .001). The change at short-term follow-up (1-3 months) was not statistically significant, whereas at long-term follow-up (6-12 months) it was 27 of 100 points (P < .001). Regarding VAS, results were statistically significant only at mid-term (MD = -1.5/10; P = .02) and long-term (MD = -3.2/10; P < .001) follow-ups. The meta-regression showed positive correlations between the response to placebo and the follow-up length (P < .001) and the effect size in the experimental group (P = .006). The level of evidence was moderate for mid- and long-term results and low for short-term results. CONCLUSION: The placebo effect for nonsurgical treatments of patellar tendinopathy is long-lasting (up to 12 months) and statistically and clinically significant. It has a perceived and true component and differs among treatments. The duration of follow-up and the effect size of experimental groups correlate with the magnitude of the placebo component, underlining the importance of RCTs to determine the effectiveness of new treatments of patellar tendinopathy.