Incidence and Predictors of Hospitalization for Bacterial Infection in Community-Based Patients with Type 2 Diabetes: The Fremantle Diabetes Study

BACKGROUND: The few studies that have examined the relationship between diabetes and bacterial infections have utilized administrative databases and/or have had limited/incomplete data including recognized infection risk factors. The aim of this study was to determine the incidence and associates of bacterial infection severe enough to require hospitalization in well-characterized community-based patients with type 2 diabetes. METHODS AND FINDINGS: We studied a cohort of 1,294 patients (mean±SD age 64.1±11.3 years) from the longitudinal observational Fremantle Diabetes Study Phase I (FDS1) and 5,156 age-, gender- and zip-code-matched non-diabetic controls. The main outcome measure was incident hospitalization for bacterial infection as principal diagnosis between 1993 and 2010. We also examined differences in statin use in 52 FDS1 pairs hospitalized with pneumonia (cases) or a contemporaneous non-infection-related cause (controls). During 12.0±5.4 years of follow-up, 251 (19.4%) patients were hospitalized on 368 occasions for infection (23.7/1,000 patient-years). This was more than double the rate in matched controls (incident rate ratio (IRR) (95% CI), 2.13 (1.88–2.42), P<0.001). IRRs for pneumonia, cellulitis, and septicemia/bacteremia were 1.86 (1.55–2.21), 2.45 (1.92–3.12), and 2.08 (1.41–3.04), respectively (P<0.001). Among the diabetic patients, older age, male sex, prior recent infection-related hospitalization, obesity, albuminuria, retinopathy and Aboriginal ethnicity were baseline variables independently associated with risk of first hospitalization with any infection (P≤0.005). After adjustment for these variables, baseline statin treatment was not significant (hazard ratio (95% CI), 0.70 (0.39–1.25), P = 0.22). Statin use at hospitalization for pneumonia among the case-control pairs was similar (23.1% vs. 13.5%, P = 0.27). CONCLUSIONS: The risk of severe infection is increased among type 2 diabetic patients and is not reduced by statin therapy. There are a number of other easily-accessible sociodemographic and clinical variables that could be used to optimize infection-related education, prevention and management in type 2 diabetes.