A framework for defining diagnostically challenging conditions identifiable through electronic algorithms.

Diagnostic delays and errors are serious and costly, affecting approximately 5 % of US adults in the outpatient setting annually. Patients with difficult-to-diagnose conditions may spend months or years undergoing diagnostic evaluation in search of a correct diagnosis. Methods are needed to identify patients with diagnostically challenging conditions (DCCs) who are experiencing diagnostic odysseys and, as a result, potential missed opportunities in their diagnosis. Given the increasing availability of longitudinal EHR data to map a patient's journey, we propose a new framework to proactively identify patients with DCCs using electronic data. These patients are at risk for missed opportunities in diagnosis, and a timelier diagnosis can improve their outcomes. We propose criteria for identifying specific DCCs where the diagnostic process for that condition makes them amenable to detection using EHR-based algorithms. We discuss the application of the proposed framework to an exemplary case study of fibrotic interstitial lung disease and provide examples of algorithms that could be implemented in the future. This work can help identify patients earlier in their diagnostic journeys, resulting in adequate follow-up and fewer missed or delayed diagnoses. Our proposed framework can inform research and potential solutions that are more real-time to potentially mitigate and avoid delays in care and resulting harm.