Errors that occur during the diagnostic process can lead to missed or wrong diagnoses and can harm patients. Most patients will experience diagnostic errors in their lifetime.1 Many diagnostic errors result from clinicians inadequately synthesizing clinical information,2 such as weighing evidence and assigning proper probabilities to potential diagnoses.
Medical diagnosis is fundamentally based on probability.3 Thus, more accurate execution of probability-based diagnosis is needed to reduce diagnostic errors related to poor information synthesis. For example, a negative stress ECG test in a patient at high risk for cardiac ischemia may be misinterpreted by someone with poor understanding of probability as definitive for the absence of disease, whereas it more often represents a false-negative. The result is a missed diagnosis and missed opportunity for early intervention.
Understanding probability and managing related uncertainty are vital for making accurate, timely diagnoses. Although central to clinical practice, these topics are not emphasized as critical skills or typically included in medical education curricula.4,5 Despite longstanding public discussion around the need to improve diagnosis, probabilistic reasoning has not been emphasized as a core competency.6 Better development and dissemination of evidence-based methods for training on probabilistic thinking could improve diagnostic accuracy and reduce errors.
This issue brief presents a framework that outlines the diagnostic process and highlights the role of probabilistic understanding at each step. The framework lays out the information needed throughout the diagnostic process to help clinicians make accurate probability assessments. It also proposes innovative methods to train on probabilistic reasoning skills for improving diagnostic decision making.