A Retrospective Analysis Examining the Impact of Coexisting Diabetes and Substance Use Disorder on Hospital Resource Utilization

Abstract

Background Substance use disorder (SUD) is characterized by the uncontrolled use of substances, leading to adverse health outcomes and increased strain on the healthcare system. Similarly, diabetes is a chronic and complex metabolic disease that significantly affects an individual's health and often necessitates additional healthcare resources for effective management. While existing literature highlights the individual impact of each condition on the healthcare system, evidence on their combined impact remains scarce. Objective The objective of this retrospective analysis was to compare hospital resource utilization among four groups: individuals with diabetes alone, SUD alone, coexisting diabetes and SUD, and neither condition. Methods The 2019 National Inpatient Sample (NIS), a large-scale inpatient discharge dataset, was used for this analysis. The dataset was filtered based on the objective of the present analysis into four groups: individuals with diabetes alone, SUD alone, coexisting diabetes and SUD, and neither condition. Diabetes types and substance related disorders were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes. For each group, hospital resource utilization was assessed across three metrics: length of hospital stay (LOS), hospital charges, and utilization of emergency department (ED) services. All statistical tests were conducted using the IBM SPSS Statistics for Windows software and were adjusted for age, gender, and race to control for potential confounding effects. Analysis of Covariance (ANCOVA) tests were performed to analyze differences in LOS and hospital charges, while logistic regression was conducted to examine the differences in the utilization of ED services among the four groups. Results Statistical analysis revealed significant differences in LOS, hospital charges, and utilization of ED services among the four groups (p<0.001). The diabetes alone group had the longest average LOS, followed by the coexisting diabetes and SUD group, the SUD alone group, and then the group with neither condition. Similarly, significant differences were observed in the hospitalization charges, with the diabetes alone group incurring the highest mean charges, followed by the group with coexisting diabetes and SUD, followed by those with neither condition, and lastly, the group with SUD alone. Finally, the group with coexisting diabetes and SUD had significantly greater odds of utilization of ED services compared to any other group. Conclusion Individuals with diabetes alone demonstrated the longest LOS and highest hospitalization charges, while those with coexisting diabetes and SUD had the highest rates of utilization of ED services. These findings suggest that diabetes primarily drives inpatient resource usage, whereas coexisting diabetes and SUD contribute more significantly to emergency care utilization. Implementing targeted strategies to address the specific needs of these populations could significantly improve patient outcomes and reduce the burden on the healthcare system.

Keywords: diabetes; emergency department utilization; hcup; healthcare cost and utilization project (hcup); healthcare costs; hospital burden; length of hospital stay (los); national inpatient sample (nis); substance use disorder (sud).