The authors interviewed 32 individuals with expertise in learning health systems to explore how such systems can work towards diagnostic excellence. Data, management, and behavioral barriers are discussed, such as the need to standardize measurement, the need for measures that both define and track errors, and that clinicians lack tools to self-assess diagnostic skills. The authors discuss how machine learning and artificial intelligence can be leveraged to advance diagnostic excellence, but that any meaningful integration must be accomplished through mutually beneficial collaborations among researchers and care providers.

Identifying and addressing diagnostic error in the ambulatory setting remains an ongoing challenge. Incorrect or delayed diagnoses can subject patients to unnecessary testing and delays in care that lead to harm. Using AHRQ safety culture survey results from 925 medical offices across the United States, researchers sought to understand the association between safety culture, health information technology (IT) implementation, and the incidence of problems that could contribute to diagnostic error in outpatient care, such as missing or unavailable test results and records. The most frequently cited problem was missing test results, with about 15% of offices in the study citing that it occurred daily or weekly. Better safety culture scores were associated with fewer problems, and practices undergoing health IT implementation reported more problems. A past WebM&M commentary highlighted an incident involving a delay in cancer diagnosis.

Although advanced computing can assist in diagnosis, these systems are not routinely utilized. This commentary suggests a framework to develop diagnostic support technologies that capture physician knowledge to enhance diagnostic safety. The authors encourage drawing from crowdsourced data to guide improvements at a system level to address future practice and educational needs.

Improving patient safety requires development of a culture of safety and transformation into a learning organization—one that has the capacity to rapidly address problems through information sharing and learning from past experience. In this systematic review, the authors characterize the published literature on organizational safety programs, and summarize published data on error detection methods (such as incident reporting systems), error analysis, and systems to mitigate and reduce specific errors (such as diagnostic errors and medication errors). The review is limited by publication bias (the preferential publication of studies with positive results) and the descriptive nature of most studies, reducing the generalizability of these studies for other organizations. An AHRQ WebM&M perspective discusses organizational approaches to safety improvement in academic and community settings.