Risk factors for nosocomial candidemia: A case-control study in adults without leukemia☆

Abstract

PURPOSE: The purpose of this study was to define risk factors for nosocomial candidemia in adult patients without leukemia at a tertiary care medical center. PATIENTS AND METHODS: All patients with nosocomial candidemia between August 1, 1981, and October 31, 1984, were included if they met strict selection criteria and did not have acute or chronic leukemia. For each case, one control was selected from among patients admitted during the same month/year and matched for hospital service and duration of hospitalization up to the first blood culture that grew Candida species. Logistic regression was used to obtain estimates of risk after simultaneously controlling for other variables. RESULTS: Candida albicans caused 24 of the 48 fungemias studied. The risk factors identified included the presence of a central line (odds ratio, 26.4; 95% confidence interval, 1.5 to 451.1); bladder catheter (13.0 1.3 to 131.4); two or more antibiotics (25.1, 2.1 to 318); azotemia (22.1, 2.2 to 223.2); transfer from another hospital (21.3, 1.7 to 274.5); diarrhea (10.2, 1.03 to 101.4); and candiduria (27.0, 1.7 to 423.5). A prior surgical procedure was associated with lowered risk (0.1, 0.01 to 0.9), suggesting perhaps that medical service patients are at higher risk than those on surgical services. Because total parenteral nutrition was always administered by means of a central line, it could not be shown to increase the risk over that conferred by a central line alone. CONCLUSIONS: This study has defined seven major risk factors for nosocomial candidemia. These findings should facilitate development of rational approaches to preventing infection and may assist clinicians in identifying those patients in whom this life-threatening complication is likely to occur.