A patient with chronic myelogenous leukemia became colonized with a Staphylococcus haemolyticus strain and experienced a septic episode caused by this strain during a cytostatic course. The strain was multiply resistant to antibiotics; the MIC and MBC of vancomycin were 2 and 4 mg/liter, and the MIC and MBC of teicoplanin were 4 and 16 mg/liter, respectively. We performed a surveillance study on the carriage of S. haemolyticus in medical and nursing staff of the hospital ward where the patient was treated. S. haemolyticus was isolated from 18 sites on 12 of the 39 people tested. A number of typing methods were performed in order to investigate the possible relationships among the isolates. Methods used were immunoblotting of staphylococcal peptides, plasmid analysis, restriction fragment length polymorphism of chromosomal DNA, and pulsed-field gel electrophoresis of total DNA. Compared with the immunoblotting technique, the molecular methods were more discriminative. The strain colonizing the patient showed a consistent pattern by all typing methods during isolation. When the immunoblot technique was used, similar patterns were found with isolates from hospital staff and isolates from unrelated sources. With the molecular techniques, no evidence of a local spread of the patient's strain was found. However, plasmid profiles and restriction fragment length polymorphism and pulsed-field gel electrophoresis patterns showed that S. haemolyticus isolates collected from hospital ward personnel were related, which was not the case with isolates collected from unrelated sources. Restriction fragment length polymorphism analysis was more discriminative when IS431 was used as a DNA probe instead of a probe based on the 16S rRNA gene. S. haemolyticus, as in this case, may develop resistance to vancomycin and teicoplanin. These antibiotics are considered the last-resort drugs for the therapy of nosocomial gram-positive infections. Thus, local spread of staphylococci resistant to these drugs is an important problem, which should be prevented by strict hygienic measures and antibiotic policy.

OBJECTIVE: To apply PCR-based DNA fingerprinting in a clinical microbiology laboratory to investigate nosocomial infections with Staphylococcus haemolyticus. METHODS: DNA fingerprints were generated by PCR on 99 S. haemolyticus isolates using different primer combinations based on ERIC, REP or arbitrarily chosen simple repeat sequences. RESULTS: Primer combinations REP1+(GTC)6 and ERIC1+ERIC2 had sufficient discrimatory power and were chosen to analyze the clinical isolates. DNA fingerprint patterns from strains isolated from the patients nursed in the same hospital ward in the period 1991--94 were approximately 90% similar to each other. One staff member, sampled in 1991, carried a strain with a similar fingerprint. CONCLUSIONS: PCR based DNA fingerprinting is a suitable method to perform in a clinical laboratory. An S. haemolyticus strain appeared to be endemic in the hospital ward and had most probably been transmitted from patient to patient. S. haemolyticus may carry glycopeptide resistance and needs attention as a causative agent of nosocomial infections.

Comparison of culture methods for monitoring Legionella species in hospital potable water systems and recommendations for standardization of such methods.