Poor quality patient reported outcome measures bias effect estimates in orthopaedic randomized studies

Area of Science:

• Orthopedics
• Clinical Trial Methodology
• Health Outcomes Research

Background:

• Patient-reported outcome measures (PROMs) are vital in assessing treatment efficacy for rotator cuff disease (RCD).
• The psychometric quality of PROMs can vary significantly, potentially introducing bias into clinical trial results.
• Assessing the impact of PROM quality on treatment effect estimates is crucial for reliable research interpretation.

Purpose of the Study:

• To evaluate whether the psychometric quality of PROMs influences treatment effect estimates in randomized clinical trials (RCTs) for RCD.
• To determine the extent of bias introduced by lower-quality PROMs in RCD research.

Main Methods:

• Systematic search of orthopedic journals for RCTs on RCD published between 2011-2016.
• Data extraction included study design, risk of bias, PROM psychometric quality, and effect estimates.
• Multilevel linear regression analysis was used to associate PROM quality with effect size, controlling for other factors.

Main Results:

• Seventy-two RCTs comprising 174 outcomes were analyzed.
• Higher-quality PROMs were associated with smaller treatment effect estimates (β = -0.32, P = 0.001).
• Longer follow-up periods were linked to slightly increased effect estimates (β = 0.08, P = 0.007).

Conclusions:

• PROMs with poor or unknown psychometric properties can overestimate treatment effects in RCD research by approximately 68.4%.
• This study provides the first empirical evidence linking PROM quality variations to biased treatment effect estimates.
• Researchers and clinicians must critically assess PROM quality to prevent biased decision-making in RCD management.