Patients’ Preferences for Artificial Intelligence Applications Versus Clinicians in Disease Diagnosis During the SARS-CoV-2 Pandemic in China: Discrete Choice Experiment

Background Misdiagnosis, arbitrary charges, annoying queues, and clinic waiting times among others are long-standing phenomena in the medical industry across world. These factors can contribute to patient anxiety about misdiagnosis by clinicians. However, with increasing growth use of big data biomedical health care communities, performance artificial intelligence (Al) techniques diagnosis is improving help avoid practice errors, including under current circumstance COVID-19. Objective This study aims visualize measure patients’ heterogeneous preferences from various angles AI versus clinicians context COVID-19 epidemic China. We also aim illustrate different decision-making latent class a discrete choice experiment (DCE) prospects for application judgment management during pandemic SARS-CoV-2 future. Methods A DCE approach was main analysis method applied this paper. Attributes dimensions were hypothesized: diagnostic method, outpatient time, accuracy, follow-up after diagnosis, expense. After that, questionnaire formed. With collected questionnaire, we apply Sawtooth software construct generalized multinomial logit (GMNL) model, mixed model sets. Moreover, calculate variables’ coefficients, standard error, P value, odds ratio (OR) form utility report present importance weighted percentage attributes. Results total 55.8% respondents (428 out 767) opted regardless description In GMNL found that people prefer 100% accuracy level most (OR 4.548, 95% CI 4.048-5.110, P<.001). For acceptable consists 3 classes respondents. The attributes substantial effects highest weights (39.29% general) expense (21.69% general), especially “accuracy” attribute, which constant classes. 1 3, + (class 1: OR 1.247, 1.036-1.463, P<.001; 3: 1.958, 1.769-2.167, 2, 1.546, 0.883-2.707, P=.37). levels increases increase all Conclusions Latent prominent useful quantifying choice. People’s “diagnostic expenses” palpable. will have potential market. expenses need be taken into consideration.