To describe the epidemiology of vancomycin-resistant enterococci (VRE) in a university hospital in Taipei, Taiwan.

Retrospective review over a 27-month period, from March 1996 to May 1998.

A tertiary-care teaching hospital in Taiwan.

Patients with VRE isolated from any body site.

Patients were identified through hospital microbiology and infection control records. Patient charts were reviewed for clinical and epidemiology data, including age, gender, previous hospital admissions, underlying diseases, types of infection, and recent antibiotic use. VRE isolates were characterized by their typical biochemical reactions, cellular fatty acid profiles, and the presence of van genes. Antibiotypes using the E-test and randomly amplified polymorphic DNA (RAPD) patterns of these isolates were used to determine the clonality.

Twenty-five isolates of VRE recovered from 12 patients were identified. One patient with a perianal abscess had 12 isolates of VRE (4 Enterococcus faecalis, 7 Enterococcus faecium, and 1 Enterococcus casseliflavus) recovered from perianal lesions. Among 3 patients who were hospitalized in the same room, 1 had a community-acquired cellulitis over the left leg caused by E faecalis, and the other 2 patients both had anal colonization with 2 isolates of E faecalis. The other 8 patients had 1 E faecalis isolate each from various clinical specimens. All isolates possessed vanA resistance phenotype and vanA genes. Different antibiotypes and RAPD patterns of the isolates from different patients excluded the possibility of nosocomial spread at the hospital.

Multiple species of VRE (E faecalis, E faecium, and E casseliflavus) and multiple clones of E faecium could colonize or infect hospitalized patients. In addition, clones of VRE can persist long-term in patients' lower gastrointestinal tracts. These results extend our knowledge of the coexistence and the persistence of multiple species and multiple clones of VRE in hospitalized patients.

To describe the incidence and patterns of nosocomial fungal infection in a large teaching hospital in Taiwan.

Prospective, hospitalwide nosocomial surveillance data from 1981 through 1993 were analyzed to show the secular trend in nosocomial fungal infection rates and to identify the most common pathogens and sites of infection (other than skin) in this hospital.

The National Taiwan University Hospital is a medical school-affiliated hospital in the city of Taipei, Taiwan, with a 1,200-bed capacity before 1991 and 1,500 beds since 1992. It provides both primary and tertiary medical care.

The overall nosocomial fungal infection rate rose from 0.9 infections per 1,000 discharges in 1981 to 6.6 per 1,000 discharges in 1993, with the highest rate at the medical intensive-care unit (26.5/1,000 discharges in 1993). This increase in infection rate was found at four major anatomic sites of infection, particularly including the bloodstream (0.08-2.19/1,000 discharges) and the urinary tract (0.36-2.95/1,000 discharges). Of 256 pathogens causing nosocomial fungemia from 1981 through 1993, Candida albicans was the most commonly isolated (50.8%), followed by Candida tropicalis (17.6%), Candida parapsilosis (11.7%), and Candida glabrata (8.2%). As compared to isolates from 1981 through 1988, the proportion of C parapsilosis and C glabrata isolated between 1989 and 1993 increased more than sixfold and fourfold, respectively. The increasing importance of fungal infections was confirmed further by the increased use of amphotericin B and azoles in this hospital.

Candida species and other yeasts have become a prominent cause of nosocomial infections in this hospital. These fungal pathogens accounted for a higher proportion of nosocomial bloodstream and urinary infections than any single bacterial species. Therefore, it is important to conduct a prospective epidemiological study and to establish in vitro antifungal susceptibility testing to enhance efforts to control nosocomial fungal infections and to minimize the risk of emergence of antifungal resistance.