Machine Learning Assisted Differentiation of Low Acuity Patients at Dispatch (MADLAD): A Randomized Controlled Trial

Resource Constrained Situations (RCS) at Emergency Medical Dispatch centers where there are more patients requiring an ambulance than there are available ambulances are common. Machine learning (ML) techniques offer a promising but largely untested approach to assessing relative risks among these patients. The study aims to establish whether the provision of ML-based risk scores predicting patient outcomes improves the ability of dispatchers to identify patients at high risk for deterioration in RCS and dispatch the first available ambulance to them. A parallel-grouped, randomized trial of adult patients assessed by a dispatch nurse in the Swedish regions of Uppsala or Västmanland as requiring a low-priority ambulance response in RCS. Patients were randomized 1:1 to be prioritized with the aid of a ML-based risk assessment tool, or per current clinical practice. Prioritization accuracy was assessed primarily in terms of whether the first available ambulance was sent to the patient with the highest National Early Warning Score (NEWS 2) based on subsequently collected vital signs. Trial registered at ClinicalTrials.gov (NCT04757194). A total of 1245 RCS were included in the study. In the intervention group, patients assigned the first available ambulance had the highest NEWS in 68.3% of cases vs 62.5% in the control group, corresponding to an odds ratio of 1.28 (95% CI 1.00 – 1.63, p = 0.047). Prespecified analyses also suggested that dispatchers complied with the tool in 80.9% (77.7 – 83.9) of cases, and that full compliance with the risk prediction instrument would have improved prioritization decisions further. This study suggests that clinical ML-based decision support tools have the ability to influence care provider decisions and improve their capacity to rapidly differentiate between high- and low-risk patients at dispatch.