Abstract

Objectives: The triple aim of population health focuses on providing the right care to the right patient at the right time. Patient use of digital health tools may reduce the overuse of emergency services. We tested the safety and clinical applicability of a patient-facing, automated digital urgent care triage tool (UCTT) for patients with chest pain.

Methods: The automated digital health UCTT (IVisitMD, Roslyn, NY) uses evidence-based algorithms to curate on-demand patient access to physicians, health facilities, and emergency departments (EDs). A retrospective observational study was performed on patients who presented to the ED before the coronavirus disease 2019 pandemic with the complaint of chest pain. We evaluated 1372 patients who presented to the ED for chest pain; 383 patients met the criteria for the final diagnostic-related group (DRG) that reflected acute coronary syndrome. In total, 109 patients who had electronic records documenting all of the components of clinical history, medical decision making, and patient disposition were assigned to the study. Two physicians not involved in patient care independently reviewed records and determined whether the ED visit was warranted (ED+) or not (ED−), which was then compared with the UCTT recommendation.

Results: Most patients had coronary artery disease or cardiac risk factors. Cardiac DRGs were observed in 84.3% of participants; 86 patients had no high-risk DRG, with ED− 9.7% by UCTT, and 19.8% by a medical doctor (MD) (P < 0.05). The high-risk DRG patients had an acute infarction, stroke, or pulmonary embolism. Twenty-three patients with a high-risk DRG were 100% ED+ by the UCTT and MD. The estimated savings, assuming the average cost of an emergency evaluation for chest pain is $7000/patient, were $70,000 with UCTT per 100 patient visits.

Conclusions: No high-risk conditions were missed by the UCTT. Our UCTT was more conservative than MD decision making, providing a sizable safety margin and adequate patient triage.