To determine the accuracy with which circulating nurses (CNs) classify surgical procedures by risk of contamination in the operating room.

Classification of surgical procedures by CNs was compared with the classification of surgical procedures by a physician observer.

University-affiliated, tertiary care hospital.

Circulating nurses used the traditional wound classification system of clean, clean-contaminated, contaminated, and dirty-infected to classify surgical wounds in the operating room. A physician remained in the operating room throughout each of 100 surgical procedures and simultaneously classified surgical wounds without the knowledge of the CNs.

Classification of surgical wounds by CNs was compared with classification by the physician observer for 50 cases in general surgery and 50 cases in trauma surgery. Compared with the physician observer, the overall accuracy of classification by CNs was 88% (95% confidence interval [CI] of 81.6% to 94.4%; Kappa statistic, 0.83). Classification of surgical wounds was more difficult in trauma surgery (accuracy of 82%) than in general surgery (accuracy of 94%). Accuracy increased for both services when surgical wounds were classified into just two categories (clean or clean-contaminated versus contaminated or dirty-infected).

Surgical wounds can be classified in the operating room with a high degree of accuracy by CNs. Classification was more difficult in trauma than in general surgery, but classification in trauma surgery improved with feedback to and additional education of CNs. The accuracy of classification by CNs was even higher when classifications were divided into just two categories.

To investigate the possibility of interhospital spread of multiresistant Staphylococcus aureus in Sao Paulo, Brazil.

We evaluated 13 nosocomial S aureus strains selected because of resistance to oxacillin and ciprofloxacin.

The strains were collected between March 1991 and September 1991 from four hospitals in Sao Paulo. Two were teaching hospitals, and two were private hospitals.

Each strain was isolated from a different patient. All patients were hospitalized when the strains were isolated.

The strains were typed by restriction endonuclease analyses of plasmid DNA (REAP) using EcoRI, HindIII, RsaI, and Alu1 and by extended antibiogram profile (34 drugs).

All strains had identical plasmid and antibiogram profile. They demonstrated the same plasmid pattern as previously described in one of the hospitals studied.

Our results suggest the dissemination of a unique oxacillin- and quinolone-resistant strain of S aureus in several hospitals of Sao Paulo, Brazil.

To report experience with brucellosis among expatriate hospital employees in an area of the world where brucellosis is endemic.

Review of cases from the viewpoints of direct supervisor and treating physician, including epidemiologic interviews and clinical follow-up.

Tertiary care hospital center in a developing country that increased from 125 beds to 450 beds during the period studied.

Employees with clinical findings suggesting brucellosis and in whom a positive blood culture or significantly elevated serologic test confirmed the diagnosis.

Nine hospital employees from nonendemic areas developed brucellosis. Five were from England and four from North America. Seven were bacteriology technologists, one was a nurse, and one was an obstetrician. Each had an acute febrile illness with malaise, arthralgia, or headache. Brucella titers were ≥ 1:1,280. Five were bacteremic with positive Brucella cultures. All responded to anti-Brucella therapy. Three patients had relapses, but there were no complications.

Among Saudi patients, brucellosis is generally attributable to drinking infected milk or contacting infected animals. On the other hand, among expatriate hospital employees, the infection is likely due to processing Brucella cultures or dealing with infected body fluids. In 1988, the laboratory began stricter infection control measures. Since then, there has been only one case of brucellosis among the hospital employees.