Citation

Landy DC. Am. J. Sports Med. 2023; 51(8): 1969-1970.

Copyright

(Copyright © 2023, American Orthopaedic Society for Sports Medicine, Publisher SAGE Publishing)

DOI

10.1177/03635465231178801

PMID

37392076

Abstract

The principles of evidenced-based medicine have been increasingly used since the 1990s to help health care providers and surgeons understand and compare treatment options.3,10 This process promotes research quality through more objective and standardized evaluations, which has highlighted the randomized controlled trial (RCT).2 Through randomization, RCTs attempt to create 2 or more similar groups from a shared study population. In theory, randomization will balance known and unknown confounders across groups. Having equivalent groups at baseline reduces the potential that between-group differences are responsible for treatment effects. This is a large part of why the evidence from RCTs is often referred to as level 1.

Even though RCTs are highly valued, they still require critical evaluation to ensure their results live up to the hype. We are lucky that a culture of evidence exists within sports medicine where the importance of identifying and promoting quality is well appreciated.5 As a part of this, it is important that authors, reviewers, editors, and readers continue to evolve with respect to our understanding of research quality and its assessment. While new methodologies such as mediation analysis and machine learning are creating an obvious need for new insights, it may be even more important that we continue to improve with respect to the quality and consistency of reviewing more common study designs such as the RCT.

While randomization attempts to balance patient characteristics across study groups, this may not be achieved.1 For instance, researchers may take a study population of patients with a chronic rotator cuff tear and randomize them to 2 different treatments. Through chance alone, the 2 study groups may differ with respect to an important factor. It is possible that one group may have a greater proportion of patients with full-thickness tears or a greater proportion of women. This lack of balance may confound estimates of treatment effect and produce biased results. It is possible that the improved outcomes seen in one group are not because of the treatment but rather baseline differences...

Language: en

Keywords

epidemiology; randomized controlled trials; biostatistics; treatment outcome