Evaluation of Patient Satisfaction in Telemedicine Using Machine Learning Algorithms

Abstract

In telemedicine, patient satisfaction is a multi-dimensional notion that sheds light on numerous areas of healthcare quality. Although researchers have identified a variety of patient and provider-related factors that impact healthcare quality, the importance levels of these factors in telemedicine patient satisfaction remains unknown. Using survey data, we developed tree based machine learning (ML) algorithms, random forest (RF) and XGBoosting, to predict relationships between patient and provider-related characteristics and telemedicine satisfaction. An RF algorithm with high prediction capability was adopted for feature importance analysis to determine the relative importance of the factors influencing patient satisfaction. In all models, the traits of 'culturally acceptable' and 'comfortable communication' were found to be the most important determinants of patient satisfaction. The adoption of this prediction model and feature importance analysis can help healthcare practitioners and managers improve their healthcare settings and topics of interest.

Date of Award	May 2022
Original language	American English