Safety of self-triage by patients using a symptom-checker: a prospective surveillance study (Preprint)

BACKGROUND Symptom-checker self-triage apps assist patients to determine the urgency of medical care. To be safe and effective, these tools must be validated, particularly to avoid potential hazardous undertriage. OBJECTIVE To investigate the safety of patient’s self-triage using a symptom-checker. METHODS A single centre, prospective clinical trial comparing the individual outcomes of patients’ self-triage with the assessment of the clinical urgency made by three successive interdisciplinary panels of physicians. Data collected between 25 November 2019 and 1 May 2020. Panel assessments and data analysis completed on 29 August 2022. Setting: Walk-in-Clinic and Interdisciplinary Emergency Department (WIC/ED) of the cantonal hospital of Baden, Switzerland. Partcipants: All patients ≥ 18 years attending the WIC/ED between 8:00 a.m. and 5:00 p.m. Exclusion criteria included ESI 1 and presence of symptoms not encompassed by the symptom-checker. 22 676 (12 655 ER, 10 021 WIC) used the facility during the enrollment period. 7550 patients attended WIC/ED within the recruitment window. 2550 gave informed consent, 7 patients withdrew. Intervention: Participants assessed their symptoms using SMASSPathfinder, a web-based symptom-checker based on a computerized transparent neural network. Main Outcome and Measure: The assessment by the panels encompassed the appropriate time-to-treat and the adequate point-of-care. If a case was adjudicated as undertriaged by the first two panels, the third panel assessed the patient’s risk to health or life, making a decision on whether a potentially hazardous undertriage had been present. Using a Clopper-Pearson confidence interval, we assumed that in order to confirm the symptom-checkers safety, the upper confidence bound should lie below 1%. RESULTS 2543 patients were included in the study, 1227 (48.25%) female, 1316 (51.75%) male, 1397 (54.94) 18-49 y, 668 (26.27%) 50.65y, 360 (14.16%) 66-80y, 118 (4.64%) >80y. Of the 2543 cases none reached the pre-specified criterion for a potentially hazardous undertriage. This resulted in an upper 95% confidence bound for the probability of a potentially hazardous undertriage of 0.1184%. CONCLUSIONS The symptom checker proved to be a safe triage tool, avoiding undertriage in a real-life clinical setting of emergency consultations at a WIC/ED. Our data suggest the symptom checker may be safely used in clinical routine. CLINICALTRIAL ClinicalTrials.gov (NCT04055298)