Lack of impact of rapid identification of rotavirus-infected patients on nosocomial rotavirus infections.

The efficacy of rapid identification of rotavirus-infected patients in the control of nosocomial rotavirus infections on an infant and young toddler ward by use of a rotavirus antigen detection test on stool from patients with diarrhea was evaluated by comparing the rate of nosocomial rotavirus infection in children during two separate 5-week periods in the winters of 1984 and 1986. In contrast to 1984 rapid rotavirus antigen testing by latex agglutination of stool from patients with diarrhea was instituted in 1986, in addition to testing for rotavirus by enzyme immunoassay, to determine whether use of rapid antigen testing resulted in an increased incidence of appropriate isolation and a decrease in nosocomial infections. In 1986 rapid identification of rotavirus resulted in an increase in hospitalization of rotavirus-infected patients in single bed rooms from 68% to 100% (P = 0.02, chi square test) but no significant increase in the use of enteric precautions for these patients. The total number of cases of nosocomial rotavirus infection in the two periods did not differ. In both periods 11 cases occurred; the nosocomial infection rate in 1984 was 18.9 cases/1000 days of exposure whereas in 1986 it was 20.2 cases/1000 days. These findings indicate that the use of rapid rotavirus antigen testing of patients with diarrhea is not of appreciable benefit in preventing the nosocomial spread of rotavirus to infants on the ward.