Among 89 episodes of Staphylococcus aureus bacteremia, factors identified as independent predictors of hospital mortality were Acute Physiology and Chronic Health Evaluation III score >60 (odds ratio [OR], 3.2; 95% confidence inter val [CI95], 1.7 to 5.9) and Lifestyle score >1 (OR, 2.1; CI95, 1.2 to 3.6). Future studies of S aureus bacteremia should take into consideration acute severity of illness (as well as treatment and source of infection) when evaluating outcome.

Two hundred seventy-five of 325 (85%) healthcare workers (HCWs) completed a questionnaire testing their knowledge of transmission of tuberculosis (TB) and infection control measures for TB. Of the 75 HCWs with no patient contact, 49 (65%) felt that masks should be worn, and 40 (53%) would use gowns. In contrast, 175 (88%) of 200 HCWs having contact with patients thought masks should be worn, and 70 (35%) would use gowns (P=.0001 and .0085, respectively). Only 87% of HCWs with patient contact felt that respiratory precautions should be instituted for TB patients. The results of our sur vey showed that HCWs should be reeducated about TB.

A cluster of four cases of surgical and burn wound aspergillosis occurred in a 900-bed, adult, tertiary-care hospital. The source was traced to the outside packages of dressing supplies, which had become contaminated during construction in the central Inventory Control area. This resulted in patients with large exposed surface areas being inoculated directly with Aspergillus spores.

A mail survey of surgeons at our medical center 5 years following adoption of Universal Precautions revealed that 29% estimated having ≥1 potentially serious blood or body fluid exposures (BBEs) per month. Failure to report BBEs (usually needlesticks) to the employee health department was common, and the majority of surgeons in practice for ≥10 years never reported such exposures. The most commonly cited reason for not reporting BBEs was perceived low risk of acquiring bloodborne infections.

To determine the prevalence of unrecognized human immunodeficiency virus (HIV)-1 infections in patients presenting to an inner-city hospital emergency department, medical records were reviewed from 1,945 patients diagnosed with diseases not related to HIV or acquired immunodeficiency syndrome. The overall seroprevalence was 2.1% (40):1.8% (11) in nontrauma versus 3.0% (29) in trauma patients. The highest prevalence was found in black, male, uninsured patients.

Four patients with atypical pneumonias admitted within a 1-week period to a community hospital were suspected and confirmed to have Legionnaires' disease. The patients had a common travel histor y. This cluster of cases uncovered a larger epidemic associated with cruise travel.

The in vitro activities of bacitracin and mupirocin were compared for seven different strains of methicillin-resistant Staphylococcus aureus. Six of seven strains showed bacitracin minimum inhibitory concentrations (MICs) of 0.5 to 1.0 units/mL, and all seven had mupirocin MICs of 0.5 to 2 μg/mL. Time-kill studies revealed 2.6- to 4.5-log reduction in 24 hours with strains susceptible to bacitracin (4 units/mL) and 0 to 2.2 reduction with mupirocin (16 μg/mL). Bacitracin should be considered further for in vivo studies because of enhanced bacteriocidal effect and lower cost.

We report an outbreak of Clostridium difficile-associated diarrhea at our medical center following adoption of Universal Precautions. Environmental cultures revealed unexpected contamination of blood pressure cuffs at a rate similar to that for bedside commodes (10% and 11.5%, respectively). An obser vational sur vey revealed that healthcare workers in the patient care areas not infrequently failed to remove their potentially stool-contaminated gloves prior to touching clean surfaces, which might have contributed to contamination of blood pressure cuffs.

Controversy has surrounded the use of liquid chemical germicides to reprocess medical instruments that are damaged by heat sterilization. A review of the literature was performed to assess and compare the efficacy of disinfection and sterilization processes. The results of this review demonstrate that high-level disinfection of thoroughly cleaned endoscopes is not associated with a higher infection rate than is “sterilization.” While there may be a theoretical distinction between the highest level of disinfection and sterilization, thorough cleaning eliminates clinical differences between the two. High-level disinfection is quick, effective, inexpensive, and recommended whenever heat sterilization is not feasible. A low-temperature sterilization process should be considered only if it is comparable in cost to disinfection or if it offers demonstrated advantages without damaging the instrument. Neither disinfection nor sterilization is likely to be effective if the instrument is not cleaned thoroughly after use.

The role of the hospital epidemiologist has changed substantially over the last 30 years as medical care has become more complex. The hospital epidemiologist needs training in methods for sur veillance, prevention, and control of nosocomial infections. The hospital epidemiologist also must know how to apply these methods to other areas, including the epidemiology of noninfectious adverse out-comes of medical care. Training in hospital epidemiology should be a defined part of every infectious disease fellowship training program. Ancillary and additional training is available from several sources.

The close relationship between colonization and the development of nosocomial infections has been demonstrated. Patient-related factors, such as underlying illness for all major sites of infection and advanced age, and pathogen-related factors, such as the ability of bacteria to adhere to epithelial cells, play the major roles in the pathogenesis of colonization. However, exact mechanisms of colonization have not been elucidated, and modulation of bacterial adherence as a method of infection prevention remains experimental. Current methods of infection prevention, therefore, focus either on preventing growth of colonizing microorganisms or on preventing patient-pathogen contact. Topical antibiotics have been used as a method of colonization prevention. However, their effectiveness may be limited by increases in antibiotic resistance; moreover, the effects on patient outcome are controversial. Maintenance of the physiologic mucosal environment using nonantimicrobial agents seems a promising approach, but only a few studies demonstrating efficacy have been published. Prevention of colonization still must rely heavily on basic infection control measures to prevent contact between patient and pathogen.