RESUMO
The purpose of this study was to characterize risk factors and groups at risk among people with diabetes and prediabetes for increased hospital utilization. Electronic health records for all people who visited the emergency department (ED) and had type II diabetes mellitus (PWD) or prediabetes (PWPD) were collected. ED use, hospital admissions, demographics, and clinical characteristics were compared between the groups. Multivariable logistic regression was used to compare the odds of ED high utilization (HU) (3+ visits per year) and hospital admissions between PWD and PWPD with interactions for socioeconomic status, race, marital status, and total comorbidities. PWD had higher mean ED visits per year compared with PWPD (1.5 vs. 1.2) and were more likely to be admitted (57.3% vs. 34.9%). PWD had higher odds of ED HU (2.1 [1.6, 2.7]) and hospital admissions (1.9 [1.6, 2.1]). Among PWD, Black, not married, and those with more than one comorbidity had the highest odds of ED HU. Among PWPD, those of low SES, Black, and divorced had the higher odds of ED HU. Hospital admissions were a risk for PWD and PWPD with increasing comorbidities. Early recognition and identification of prediabetes and clear criteria for diagnosis could reduce ED visits and hospital admissions.