Evaluating Manual Therapy in Musculoskeletal Pain: Why Certain Trial Designs May Overestimate Effectiveness—A Scoping Review

Background and Objective Atraumatic musculoskeletal pain, regardless of the affected body region, is a highly prevalent condition impacting over 25% of the global population and contributing significantly to the burden of disease. A common study design compares physiotherapy or exercise therapy alone to the same intervention combined with MT (A vs. A + B). This study design is inherently flawed due to its inability to isolate the effect of treatment B, the potential for interaction effects, and the lack of control for non‐specific contextual factors. The goal of this study was to compile studies using that approach and to examine the short‐, medium‐, and long‐term effects of the addition of MT to a control treatment. Databases and Data Treatment This scoping review identified 95 randomised controlled trials (RCTs) with a systematic literature search in the electronic bibliographic databases MEDLINE (via PubMed), EBSCO, and PEDro. Results Long‐term effects were absent, and medium‐term effects were infrequent. Approximately half of the studies reported statistically significant effects in the immediate or short‐term follow‐up; however, these effects were of limited clinical relevance and susceptible to methodological issues. Furthermore, studies with lower methodological quality were more likely to report significant effects (85%), whereas medium‐ and high‐quality studies showed positive results in only 50% of cases. Conclusions This review highlights significant research gaps and provides methodological insights. The study design in question is therefore methodologically problematic, as it tends to generate positive short‐term results without providing clear answers or meaningful clinical implications for researchers and clinicians. Significance Statement This scoping review summarizes studies using an ‘A vs. A+B design’, where manual therapy is added to usual care, with or without a sham control. The review identifies a lack of medium‐ or long‐term effectiveness and highlights a methodological bias toward generating positive short‐ or immediate‐term results of questionable clinical relevance. Based on these findings, we provide several recommendations to improve future research and to support clinicians in interpreting the current evidence base.