Clinical and statistical characterization of 30-day hospital readmission among patients with diabetes mellitus: a retrospective cohort study using Open Data

Abstract

Hospital readmission within 30 days is an important indicator of continuity of care and clinical vulnerability among people with diabetes mellitus. This study aimed to characterize sociodemographic, hospital, and therapeutic factors associated with early readmission using the open Diabetes 130-US Hospitals for Years 1999-2008 dataset. A retrospective cohort study was conducted on inpatient electronic health records from 130 United States hospitals. After excluding encounters ending in death or hospice referral, records with invalid sex data, and repeated encounters by retaining the first recorded admission per patient, 69,987 unique patients were analyzed. Early readmission (<30 days) was the dependent variable. Descriptive statistics and multivariable binary logistic regression were used to estimate odds ratios (ORs), 95% confidence intervals (95% CIs), and p values. Overall, 6,285 patients were readmitted within 30 days (8.98%). Older age, longer length of stay, a greater number of diagnoses, use of diabetes medication, and downward insulin adjustment were independently associated with higher odds of early readmission. Male sex and emergency admission were not statistically significant after adjustment. These findings identify clinically interpretable markers that may support structured discharge planning, medication reconciliation, diabetes self-management education, and timely post-discharge follow-up. Because the data are historical and observational, associations should not be interpreted as causal or directly generalized to current health systems.