Accepted for/Published in: Journal of Medical Internet Research
Date Submitted: Feb 17, 2023
Date Accepted: Jul 1, 2024
Predictors of medical and dental clinic closure by machine-learning methods: a cross-sectional study using empirical data
ABSTRACT
Background:
Many small clinics play an important role providing healthcare in local communities. Predicting their closure would be helpful in managing healthcare resource allocation.
Objective:
There have been few studies on the prediction of clinic closure using artificial intelligence methods. The objective of this study was to test the feasibility of predicting their closure using machine-running techniques.
Methods:
The units of analysis were medical and dental clinics. This study used health insurance claims data. The study subjects were running (RN) and closed (CL) clinics between January 1, 2020 and December 31, 2021. Using all CL clinics, CL and RN clinics were selected at a ratio 1:2 based on the locality of study subjects using the propensity matching score of logistic regression. Four machine-learning techniques were used: Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and Extreme Gradient Boost (XGB). The study evaluated the accuracy of the modeling using the Area under Curve (AUC) method and presented important factors critically affecting closures. The study used SAS version 9.4 and Python version 3.7.9.
Results:
The best-fit model for closure of medical clinics with cross validation was SVM (AUC 0.762, 95% CI [CI] 0.746–0.777, p<0.001), followed by RF (AUC 0.736, CI 0.720–0.752, p<0.001), XGB (AUC 0.720, CI 0.704–0.736, p<0.001), and LR (AUC 0.533, CI 0.515–0.550, p=0.004). The best-fit model for dental clinics was XGB (AU 0.700, CI 0.675–0.725, p<0.001), followed by RF (AUC 0.687, CI 0.661–0.712, p<0.001), LR (AUC 0.652, CI 0.626–0.678, p<0.001), and SVM (AUC 0.593, CI 0.566–0.620, p<0.001). The most significant factor associated with the closure of medical clinics was years of operation, followed by population growth, population, and percentage of medical specialties. In contrast, the main factor affecting the closure of dental clinics was the number of patients, followed by annual variation in the number of patients, year of operation, and percentage of dental specialists.
Conclusions:
This study showed that machine-running methods are useful tools for predicting the closure of small medical facilities. Important factors affecting medical facility closure also differed between medical and dental clinics. Developing good models would bring the prevision of unnecessary medical facility closure at the national level. Clinical Trial: none
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.