An outbreak of vancomycin-resistant enterococci (VRE) began at the University of Massachusetts Medical Center in May 1993. As of September 1995, we had a total of 253 patients infected or colonized with VRE, with consequent increasing demand for private rooms. We analyzed results of surveillance cultures for VRE of 49 patients known to be colonized or infected with VRE. Of these, 34 (70%) were classified as persistent carriers, defined as patients with at least three consecutively positive cultures from any site taken over at least a 2-week period. The length of carriage varied from 19 to 303 days (median, 41 days); 11 patients were converters, defined as patients with three consecutive negative cultures from all previously colonized sites taken over a 3-week period. These patients were free of VRE for 39 to 421 days (median, 142 days). Four were recolonizers after they were documented to be clear of VRE for 33 to 106 days. VRE carriage tends to be prolonged, and hospitalization of patients with VRE will require continued isolation and contact precautions for control of transmission.

A review of our infection control records revealed 3,159 new isolations of methicillin-resistant Staphylococcus aureus (MRSA) from 1988 to 1994. Prior to this period, our approach to MRSA had changed from eradication to containment measures. We found a decline in MRSA rates from 11.4 to 5.2 first isolations per 1,000 deaths and discharges over the study period.

A one-month prospective survey of all inpatients given vancomycin was performed in a community teaching hospital with a low prevalence of vancomycin-resistant enterococci. Only 20 of the 97 vancomycin orders written from August 1 to September 1, 1996, were consistent with Hospital Infection Control Practices Advisory Committee (HICPAC) guidelines. Surgical prophylaxis accounted for 37 of the 77 orders inconsistent with HICPAC guidelines.

Between 1987 and 1989, two sequential outbreaks of nosocomial infection caused by Enterobacter cloacae occurred in the pediatric intensive-care unit of a tertiary-care teaching hospital. Seventeen strains retrieved from the outbreaks and two control strains identified in other wards were typed by ribotyping and random amplification of polymorphic DNA (RAPD). The results indicated that the genomic pattern of strains identified between the first and second outbreaks was different. We conclude that both ribotyping and RAPD are highly discriminatory and reproducible methods for typing E cloacae. RAPD seems to be more efficacious and cost-effective.

Clonal spread of vancomycin-resistant Enterococcus faecium among seven patients on one ward of a community teaching hospital was identified by contour-clamped homo-geneous electric-field gel electrophoresis. Environmental cultures isolated the same strain from the handle of a shared electronic ear-probe thermometer. Cross-contamination of the clonal strain between two geographically separate units on this ward, sharing equipment but not personnel, suggests the possibility of an environmental source.

A cluster of serious Escherichia coli infections was identified among patients in a neonatal intensive-care unit. Infection control staff identified the outbreak because they realized that E coli rarely caused infections in this unit. Pulsed-field gel electrophoresis confirmed that one strain of E coli was transmitted among patients.

The cumulative yield from cultures of separate sites was determined during the investigation of a methicillin-resistant Staphylococcus aureus (MRSA) outbreak. Surveillance cultures were submitted from clinical sites, nose, groin, and axilla of 421 patients on two different occasions. MRSA was recovered most often from various clinical sites, including lower respiratory tract, surgical wounds, urinary tract, and decubitus ulcers (total, 13 patients). Four additional patients were identified as positive from the first nasal swab, one patient from the second nasal swab, and two others from swabs of the groin. The submission of axillary swabs or a second groin swab did not identify additional MRSA-colonized patients and resulted in additional costs of $4,525.

The rate of nosocomial respiratory syncytial virus infection was measured in a large pediatric hospital using an incidence density method. The at-risk days for nosocomial RSV were summed during a defined winter period in which there were 54 admissions with community-acquired RSV infection giving a rate of 2.9 cases per 1,000 at-risk days (95% confidence interval, 0.3-5.4 per 1,000).

This study was performed to determine whether van-comycin use at our pediatric hospital was consistent with modified Centers for Disease Control and Prevention guidelines. Vancomycin use was inappropriate in 54% of patients. Inappropriate use briefly decreased by 14% after educational efforts. Further education regarding van-comycin use was deemed necessary and is continuing.

Nosocomial pneumonia is defined as an infection of lung parenchyma that was neither present nor incubating at the time of the patient's admission to the hospital. In the United States, hospital-acquired pneumonia is the second most common nosocomial infection and accounts for the most deaths from nosocomial infection. We describe how infection control personnel can use targeted surveillance to identify clusters of cases and to prevent pneumonia. We also discuss common pathogens that cause nosocomial pneumonia; ventilator-associated pneumonia; and strategies for prevention of hospital-acquired pneumonia.