Prediction of in-hospital mortality rate in COVID-19 patients with diabetes mellitus using machine learning methods

Pooneh Khodabakhsh; Ali Asadnia; Alieyeh Sarabandi Moghaddam; Maryam Khademi; Majid Shakiba; Ali Maher; Elham Salehian

doi:10.1007/s40200-023-01228-y

Prediction of in-hospital mortality rate in COVID-19 patients with diabetes mellitus using machine learning methods

J Diabetes Metab Disord. 2023 May 13;22(2):1-14. doi: 10.1007/s40200-023-01228-y. Online ahead of print.

Authors

Pooneh Khodabakhsh¹, Ali Asadnia², Alieyeh Sarabandi Moghaddam³, Maryam Khademi⁴, Majid Shakiba⁵, Ali Maher⁶, Elham Salehian⁷

Affiliations

¹ Department of IT and Computer Engineering, Azad Islamic University South Tehran Branch, Tehran, Iran.
² Institute of Social Sciences, Department of Business Analytics, Marmara University, Istanbul, Turkey.
³ School of Computer Engineering, Iran University of Science and Technology (IUST), Tehran, Iran.
⁴ Department of Applied Mathematics, Azad Islamic University South Tehran Branch, Tehran, Iran.
⁵ Advanced Diagnostic and Interventional Radiology Research Center (ADIR), Tehran University of Medical Sciences, Tehran, Iran.
⁶ Department of Health Policy, Economics and Management, School of Management and Medical Education, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
⁷ Department of Information Technology, Medical Science of Shahid, Beheshti University, Tehran, Iran.

Abstract

Background: Since its emergence in December 2019, until June 2022, coronavirus 2019 (COVID-19) has impacted populations all around the globe with it having been contracted by ~ 535 M people and leaving ~ 6.31 M dead. This makes identifying and predicating COVID-19 an important healthcare priority.

Method and material: The dataset used in this study was obtained from Shahid Beheshti University of Medical Sciences in Tehran, and includes the information of 29,817 COVID-19 patients who were hospitalized between October 8, 2019 and March 8, 2021. As diabetes has been shown to be a significant factor for poor outcome, we have focused on COVID-19 patients with diabetes, leaving us with 2824 records.

Results: The data has been analyzed using a decision tree algorithm and several association rules were mined. Said decision tree was also used in order to predict the release status of patients. We have used accuracy (87.07%), sensitivity (88%), and specificity (80%) as assessment metrics for our model.

Conclusion: Initially, this study provided information about the percentages of admitted Covid-19 patients with various underlying disease. It was observed that diabetic patients were the largest population at risk. As such, based on the rules derived from our dataset, we found that age category (51-80), CPR and ICU residency play a pivotal role in the discharge status of diabetic inpatients.

Keywords: Coronavirus; Decision Tree Classifier; Diabetic patients; Epidemiology Research.

© The Author(s), under exclusive licence to Tehran University of Medical Sciences 2023, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.