Human- Versus Machine Learning–Based Triage Using Digitalized Patient Histories in Primary Care: Comparative Study

Abstract

Background Smartphones have made it possible for patients to digitally report symptoms before physical primary care visits. Using machine learning (ML), these data offer an opportunity to support decisions about the appropriate level of care (triage). Objective The purpose of this study was to explore the interrater reliability between human physicians and an automated ML-based triage method. Methods After testing several models, a naïve Bayes triage model was created using data from digital medical histories, capable of classifying digital medical history reports as either in need of urgent physical examination or not in need of urgent physical examination. The model was tested on 300 digital medical history reports and classification was compared with the majority vote of an expert panel of 5 primary care physicians (PCPs). Reliability between raters was measured using both Cohen κ (adjusted for chance agreement) and percentage agreement (not adjusted for chance agreement). Results Interrater reliability as measured by Cohen κ was 0.17 when comparing the majority vote of the reference group with the model. Agreement was 74% (138/186) for cases judged not in need of urgent physical examination and 42% (38/90) for cases judged to be in need of urgent physical examination. No specific features linked to the model’s triage decision could be identified. Between physicians within the panel, Cohen κ was 0.2. Intrarater reliability when 1 physician retriaged 50 reports resulted in Cohen κ of 0.55. Conclusions Low interrater and intrarater agreement in triage decisions among PCPs limits the possibility to use human decisions as a reference for ML to automate triage in primary care.

Artificial Intelligence-Based Application Provides Accurate Medical Triage Advice When Compared to Consensus Decisions of Healthcare Providers

“…Throughout the various phases of the study, there was high variability among the individual providers' decisions, similar to previous research [ 12 ]. In our study, ER providers generally agreed on clinical presentations that required emergency care but had more variability in their triage decisions for presentations that were more appropriate for urgent or outpatient care.…”

Health care utilization following “digi-physical” assessment compared to physical assessment for infectious symptoms in primary care

“…“Patient adherence” was 79.4% for recommended physical follow-up, and 82.4% for recommended self-care/no follow-up. Previous research on physician triage based on digital patient histories suggests high inter- and intra-rater variability in primary care triage thus making it difficult to optimize this process [ 26 ].…”

A Note on the Effects of Digital Primary Health Care on Utilization: Concepts, Evidence, and Descriptive Analysis of Non-Experimental Register Data from Sweden

“…To the extent that digital care is less costly in terms of direct and indirect resource use (including time) from the perspective of both providers and patients, these conditions should be managed by means of digital care. However, being able to identify cases that may need to be seen by a physician, is difficult a priori (Entezarjou, Bonamy et al 2020). To be able to perform such identifications, health systems need to be fully integrated, either organizationally or virtually through common medical records and communication technologies for sharing patient data.…”