Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
2639848 | American Journal of Infection Control | 2006 | 7 Pages |
BackgroundRisk factors for bacteremia in patients with hospital-acquired bacteriuria are largely unknown. Given the morbidity and costs associated with nosocomial bacteremia, determining risk factors could enhance the safety of hospitalized patients.MethodsWe conducted a case-control study within the Veterans Affairs Puget Sound Health Care System. A patient hospitalized between 1984 and 1999 from whom a urine culture and a blood culture grew the same organism ≥48 hours after admission was considered a case. Control patients were those with significant bacteriuria detected ≥48 hours after admission who did not have a positive blood culture. We used logistic regression to determine independent risk factors for bacteremia.ResultsThere were 95 cases and 142 controls. Independent, statistically significant predictors of bacteremia included immunosuppressant therapy within 14 days of bacteriuria (odds ratio [OR], 8.13); history of malignancy (OR, 1.94); male sex (OR, 1.88); cigarette use in the past 5 years (OR, 1.26); number of hospital days before bacteriuria (OR, 1.03); and antibiotic use within 3 days of bacteriuria (OR, 0.76). Corticosteroid use within 7 days of bacteriuria predicted bacteremia in patients <70 years old (OR, 14.24). Similarly, patients <70 years old were more likely to develop bacteremia if they had diabetes mellitus (OR, 6.19).ConclusionDelineating risk factors for nosocomial urinary tract-related bacteremia can help target appropriate preventive practices at the highest risk patients.