Effects and use of a computerized diagnostic decision support system during the diagnostic process in emergency...

This study's aim was to bridge the gap between the theoretical effectiveness of clinical diagnostic decision support systems (CDDSs) and their practical implementation in emergency medicine. By focusing on one CDDS, we sought to understand how residents use these systems, and how that affects the diagnostic process. Specifically, we investigated the influence of CDDS use on residents' advice-seeking behavior.

This mixed-methods study combined exploratory analyses of data obtained during the Differential Diagnosis Broadening (DDx-BRO) cluster-randomized trial and its accompanying preplanned mixed-methods substudy. The DDx-BRO trial employed a multicenter, outcome assessor, patient-blinded, cluster-randomized, multiperiod crossover design across four Swiss emergency departments. The use of the CDDS in clinical practice was investigated by assessing self-reported use and tracked use, as well as observed workflow indicators such as contact time and advice-seeking behavior.

A total of 1,204 patients were included. Observations revealed a significant decrease in consultations with senior physicians in intervention periods, during which 91 % of observed physicians reported using the CDDS because the use was mandatory. When asked, users rated the CDDS as supportive for 23 % the cases. When this was the case, physicians cited the CDDS broadening their list of differential diagnoses and helping confirm their diagnosis as reasons.

Our study offers insights into the multifaceted impact of a CDDS on residents' decision-making processes, highlighting the role of such systems in confirmation-oriented diagnostic approaches and changes in consultation patterns.