Endemic acinetobacter in intensive care units: epidemiology and clinical impact

To assess whether Acinetobacter isolates obtained over 20 months in a tertiary care hospital were epidemiologically related; to establish the clinical importance of the organisms; and to identify the isolates according to the recent taxonomy.

Fifty eight Acinetobacter isolates from 49 patients collected during 1984 and 1985 were investigated. Most isolates were from respiratory tract specimens from intensive care patients. The organisms were typed by cell envelope protein electrophoresis and by a quantitative carbon source growth assay; patients' charts were reviewed to differentiate between colonisation and infection; representative isolates were identified to species level by DNA-DNA hybridisation.

Twelve protein profiles were distinguished in the isolates. Forty two isolates were of the same protein profile (profile I); other profiles were observed in a few or single isolates. Cluster analysis of carbon source growth divided profile I isolates into two groups--one of isolates from 1984 and one from 1985. They were identified as A baumannii and associated with infections in eight patients. Four other infections were caused by acinetobacters with other protein profiles (three of A baumannii; one of the unnamed DNA group 3).

Apart from sporadic strains, two strains of the same protein profile, but distinguishable by carbon source growth, were successively endemic. Cluster analysis was a valuable tool in the interpretation of typing and epidemiological data. The 12 (28%) infections of Acinetobacter in 43 patients in intensive care suggest that the presence of these organisms in wards of severely ill patients should be a cause of concern.