International Delphi on a medical screening tool for emergency department patients with psychiatric complaints

Abstract Background: Medical screening in the emergency department (ED) is performed to identify underlying medical conditions that may contribute to a patient’s psychiatric presentation and/or require treatment. This process frequently causes disagreement between emergency physicians and psychiatrists regarding objectives and scope of medical screening, leading to potentially unnecessary testing, prolonged ED length-of-stay, and increased costs. This study aims to develop an evidence-informed and consensus-based medical screening tool through a transdisciplinary, international Delphi process. Methods: This modified international Delphi study was conducted between May and November 2025 using a 4-round modified Delphi process with online surveys and online steering committee discussions. Participants included 25 society representatives from 7 emergency medicine and psychiatry societies across 4 continents and 6 patient representatives with prior ED medical screening experience as patients or patient relatives. Participants generated, rated, refined, and reached consensus on proposed screening elements. Consensus was predefined as ≥80% agreement in Round 3. In Round 4, the resulting tool was applied to 37 standardized clinical vignettes. Interrater agreement was assessed using Gwet agreement coefficient (AC). Results: Of 28 invited society representatives and steering committee members, 25 (89.3%) participated in Round 1, 23 (82.1%) in Round 2, 22 (78.6%) in Round 3, and 23 (82.1%) in Round 4. The 6 patient representatives completed all 4 rounds. In Round 1, participants generated 91 statements from society representatives and 27 from patient representatives. Following refinement in Round 2, consensus in Round 3 was achieved for 48 of 91 society representatives’ statements (52.8%) and 21 of 27 patient statements (77.8%), which were synthesized into the Delphi-Informed Medical Evaluation (DIME) tool. In Round 4, application of the draft DIME tool to 37 clinical vignettes demonstrated almost perfect agreement among society representatives (Gwet AC, 0.86; 95%-CI 0.78–0.93; p < .001) and substantial agreement among patient representatives (Gwet AC, 0.76; 95%-CI 0.66–0.86; p < .001). Conclusions This international modified Delphi study developed the DIME tool, a consensus-based medical screening tool to identify adult ED patients with psychiatric complaints who are at low risk for acute medical illness. Prospective multicenter validation studies are required before clinical implementation. Trial registration clinicaltrials.gov, NCT06936826