Mitigating Biases in a Cohort Study: A Critical Examination of the Silicone Arthroplasty in Rheumatoid Arthritis Study.

Abstract

Research bias, or the systematic errors of a study, can arise before, during, or after a trial ends. These biases hinder the internal validity of the study, which is the accuracy of a study's conclusions regarding the effects of an intervention on a given group of subjects. With the growing use of evidence-based medicine, there is a demand for high-quality evidence from the research community. Systematic reviews and meta-analyses of randomized controlled trials are considered the highest level of evidence, followed by individual randomized controlled trials. However, most surgical trials cannot be conducted as randomized controlled trials because of factors such as patient preferences and lack of equipoise among surgeons. Therefore, surgical trials may lack features that are held as important standards for high-quality evidence, such as randomization and blinding. To demonstrate the biases that surgical trials may encounter, the authors examined a prospective cohort study, the Silicone Arthroplasty in Rheumatoid Arthritis study. The authors focus on the challenges that arise during a surgical trial, including the design, implementation, and methods used to report the clinical evidence. By recognizing and addressing obstacles that exist in research, investigators will provide health care providers with high-quality evidence needed to make well-informed, evidence-based clinical decisions.