Focused microbiologic surveillance by specific hospital unit as a sensitive means of defining antimicrobial resistance problems

An annual summary of susceptibility patterns for the predominant clinical isolates from hospitalized patients can be of considerable assistance in selecting antimicrobial agents for sepsis of unclear etiology, as well as for guiding empiric therapy for other serious infections. Yearly summaries of the susceptibility patterns of the predominant clinical isolates from all patients hospitalized at Vanderbilt University Hospital (VUH) from July 1987 through June 1991 revealed only minor differences over time in susceptibility patterns. However, the clinical impression of physicians treating patients in various intensive care units (ICUs) was that there were serious resistance problems in some units. To better define the prevalence of clinical isolates and their susceptibility patterns for patients within ICUs at VUH, we utilized a "focused microbiologic surveillance" technique that addressed each unit separately. Both the predominant clinical isolates and their susceptibility patterns were determined and compared with those from the hospital as a whole. Because susceptibility patterns of clinical isolates by site of infection within these units were considered important, we also reviewed the summaries of susceptibility patterns for blood, sputum, and urine isolates from patients in ICUs and compared these with the summaries from each ICU and from the hospital. No major resistance problems were identified on a hospital-wide basis. In contrast, focused microbiologic surveillance by specific hospital ICU revealed important differences in the prevalence of pathogens among units and at different times. In 1987, Pseudomonas aeruginosa was the single most common Gram-negative organism isolated in the neonatal unit, while Acinetobacter spp. were rarely isolated. By 1991, this trend was completely reversed.