Assessing the diagnostic capabilities of the app DemDx in an acute care setting (Preprint)
UNSTRUCTURED DemDx is a differential diagnosis app for students and junior doctors. Starting with a patient’s presenting complaint the app goes through a step-by-process through history, examination and investigation findings to an increasingly refined differential diagnosis list until a single most likely diagnosis is reached. The aim of this project was to assess the accuracy of DemDx in an Emergency Department (ED) setting. Anonymised clinical records for 100 patients were retrospectively obtained from the ED in Beth Israel, Boston, USA. This contained the differential diagnoses from the clerking doctor, who performed the initial assessment (D1). The discharge diagnosis was used as the gold standard diagnosis (D2). D1 agreed with D2 in 74.44% of cases while DemDx agreed with D2 in 85.56% of cases (p=0.0003716). When the first, and thus most likely, differential was taken from D1 and DemDx, they agreed with D2 in 20 and 18.8% of cases, respectively (p=0.1428). This demonstration of the clinical accuracy of the app highlights how it can be a useful medical student education tool.