To estimate seroprevalence of antibodies to hepatitis C virus in healthcare workers at high risk for blood exposure.

A prospective anonymous seroprevalence survey of 243 healthcare workers.

A large referral hospital and 2 community hospitals in Connecticut.

Healthcare workers, including surgical personnel, dentists, hemodialysis workers, laboratory workers, and emergency room staff.

Antibody to hepatitis C virus was found in 1.6% (95% confidence interval [CI95] = 03.2%) of healthcare workers. None of the prevalent seropositives had a past history of clinical hepatitis or blood transfusion.

We conclude that the sero-prevalence of hepatitis C virus in healthcare workers with a high degree of blood exposure is low and is similar to seroprevalence rates reported for volunteer blood donors. However, first-generation hepatitis C serologic tests may underestimate the true prevalence of infection. Further studies, including prospective cohort studies, will be required to determine if the low seroprevalence is from low risk of acquisition of disease or from loss of measurable humoral antibody response to the virus.

To define the source of invasive coagulase-negative staphylococci (CNS) and the epidemiology of strain variation in immunocompromised patients.

Weekly microbial surveillance cultures were obtained from the nares, throat, skin, rectum, and urine. Plasmid pattern analysis was performed on all coagulase-negative staphylococci isolated from blood cultures and on selected strains from the surveillance sites.

A 902-bed, university-owned, tertiary-care referral hospital.

Forty-four patients on the bone marrow transplant or hematologic malignancy services.

Plasmid pattern analysis was performed on 340 surveillance isolates (median = 7 per patient) and 201 bloodstream isolates (median = 3 per patient). Patients were colonized with numerous unique strains (median= 5 per patient) of coagulase-negative staphylococci. The 44 patients had 108 episodes of positive blood cultures, 20 of which were preceded by colonization with the same strain. Isolation of the matching strain from surveillance cultures preceded the positive blood culture by 1 to 8 days in 9 episodes and 18 to 389 days in 11 episodes. The matching strain was isolated from the skin in only 6 (30%) of those episodes and from mucosal sites in 70%. Of the 108 episodes of positive blood cultures, 21 were identified as nosocomial bloodstream infections. Four of the 21 nosocomial bloodstream infections were preceded by colonization with the same strain. In all 4 episodes, the infecting strain was cultured from the nares before the blood cultures were obtained.

Our results suggest that mucous membranes might be sources for strains of CNS causing bacteremia.

Describe an outbreak of influenza A (H3N2); provide an analysis of vaccine efficacy; measure the sensitivity specificity, and positive predictive value of 3 clinical case definitions of influenza.

A nursing home in Washington County, Maryland. The outbreak involved 52 residents (attack rate = 47.7%) and at least 10 of 140 employees (minimum attack rate = 7.1%).

Twenty-five residents exhibited a 4-fold or greater increase in titer to influenza A/Sichuan/2/87. Vaccine efficacy was measured at -7.1%, suggesting that the influenza vaccine in 1988/1989 did not offer optimal protection against influenza A infection for the institutionalized elderly.

The outbreak was a clear indicator of the need for rapid diagnosis. With the use of rapid diagnostic tests, influenza A could have been detected in time to use amantadine.

To determine if the spread of Clostridium difficile-associated diarrhea is related to the use of electronic thermometers in an acute hospital and a chronic healthcare facility.

After finding that a significant percentage (20.8%) of electronic rectal thermometer handles were contaminated with C difficile, all electronic thermometers were replaced with disposables. A before/after trial was conducted to determine if the change to disposable thermometers would reduce the incidence of C difficile-associated diarrhea.

The study took place in a 343-bed acute hospital and a 538-bed skilled nursing facility.

All patients who underwent routine microbiological evaluation for nosocomially acquired diarrhea over a l-year period were included in the study. Nosocomial diarrhea was defined as 3 or more loose stools per day for 2 consecutive days and/or abdominal findings such as pain, distention, and ileus occurring 3 or more days after admission.

During the 6-month postintervention period, the incidence of C difficile-associated diarrhea was reduced from 2.71/ 1,000 patient days to 1.76/ 1,000 patient days in the acute hospital and from 0.41/1,000 patient days to 0.1 1/1,000 patient days in the skilled nursing facility The protective effect of the intervention was statistically significant for both facilities.

Replacement of electronic thermometers with single-use disposables significantly reduced the incidence of C difficile-associated diarrhea in both acute care and skilled nursing care facilities. Data suggest that the rectal route may be important in the transmission of C difficile in these settings.