The researchers analyzed data from the medical records of type 2 diabetes mellitus in-patients 18 years and older who received healthcare services at Ho Teaching Hospital from January 2017 to November 2022.
Sandra A. Darfour-Oduro, assistant professor of public health studies co-authored an article titled “Predictive models and determinants of mortality among T2DM patients in a tertiary hospital in Ghana, how do machine learning techniques perform?”
The prevalence of type 2 diabetes mellitus (T2DM) is increasing in lower and middle–income countries (LMICs) and this calls for preventive public health interventions. Studies have consistently reported high mortality rates from T2DM. Darfour-Oduro worked with collaborators at the University of Health and Allied Sciences in Ho, Ghana to determine the predictors of mortality among T2DM patients at the Ho Teaching hospital in Ghana.
The researchers analyzed data from the medical records of T2DM in-patients 18 years and older who received healthcare services at Ho Teaching Hospital from January 2017 to November 2022.
The data obtained included sociodemographic characteristics (age, sex, marital
status, family history, educational level, occupation and place of residence), lifestyle variables (smoking and alcohol intake), family history of diabetes, cardiovascular disease (CVD), asthma, diabetic complications and mortality outcome.
Descriptive and inferential statistics were generated to describe and build predictive models respectively. The performance of machine learning (ML) techniques such as support vector machine (SVM), decision tree, k nearest neighbor (kNN), eXtreme Gradient Boosting (XGBoost) and logistic regression were evaluated using the best-fitting predictive model for T2DM mortality.
Darfour-Oduro and her co-authors found that nephropathy was the significant predictor of T2DM mortality. Also, a 100% mortality was recorded among the T2DM patients with sepsis
The authors recommend that for better prediction of mortality outcome, a holistic assessment of sociodemographic characteristics, family history, lifestyle variables and complications of T2DM is required.