CLN Daily 2024
Modern medicine cannot operate without diagnostic tests, which are intended to improve patients’ health and enhance healthcare efficiency without compromising outcomes. Yet demonstrating that tests are delivering on these goals is no easy task. In yesterday’s opening plenary lecture, Patrick Bossuyt, PhD, winner of the 2024 Wallace H. Coulter Lectureship Award, made a compelling case for why we need strong clinical trials of laboratory tests.
Bossuyt is a professor of clinical epidemiology at the University of Amsterdam whose research focuses on developing methods to evaluate diagnostic tests. He noted that clinical trials are already widely accepted as the standard for assessing the safety and efficacy of new pharmaceuticals and other interventions in clinical medicine, such as surgical strategies. However, laboratory medicine has lagged with respect to adopting a clinical-trial-style approach toward assessing diagnostic value.
Bossuyt explained that diagnostic-accuracy studies should aim to evaluate the degree to which medical tests correctly classify study participants as having a target condition. This type of investigation is difficult to do and requires precise scientific protocols. Researchers must carefully consider the challenges, which include figuring out how to recruit participants, define eligibility criteria, develop appropriate controls, and handle missing data.
Unfortunately, this work is not always done, and the ramifications can be significant. Bossuyt pointed out that study-design shortcomings may be partially to blame for the relative lack of successful new biomarkers and clinical tests.
To address this lack of rigor, Bossuyt led the effort to develop Standards for Reporting Diagnostic Accuracy Studies (STARD), which has identified 30 items that should be included when reporting on diagnostic tests to aid in evaluating study completeness and transparency. Importantly, he recommended considering these items as early as possible during a study’s design phase.
The first version of STARD, which was published in 2003, helped raise awareness about the importance of strong study designs, protocols, and analysis plans for diagnostic test evaluations. STARD has caught the eye of regulatory bodies, journal editorial boards, and clinical researchers around the world. Notably, since its update in 2015, the STARD guidelines have been cited over 1,000 times.
During his talk, Bossuyt shared some of his professional background and career path. He has authored more than 1,000 articles and continues to be incredibly prolific. He has also trained more than 100 doctoral students.
Bossuyt made it clear that rigorous clinical trials should be applied to medical tests. His talk was a call to action for laboratory medicine professionals to rethink how diagnostic tests are evaluated and ultimately improve the evidence base for tests used in medical decision-making.