To determine risk factors for outbreaks of influenza virus infections in chronic-care facilities for the elderly, the authors compared the characteristics of two groups of nursing homes in Genesee County, Michigan, in 1982-1983, following a community-wide epidemic caused by A/Bangkok/1/79-like (H3N2) viruses: seven homes in which an outbreak occurred (case homes) and six homes with sporadic illnesses only (control homes). The two groups were similar in many respects, including the physical characteristics of the facilities, visitation and staffing patterns, infection control practices, and demographic and clinical profiles of residents. Case homes, however, had larger resident populations than control homes (median 160 vs. 92, respectively; p = 0.028) and a lower proportion of residents who had been vaccinated with A/Bangkok antigen the previous autumn (median 51% vs. 81%, respectively; p = 0.047). Based on a retrospectively determined estimate of the efficacy of influenza vaccine in preventing illness (43%), the number of residents presumed to be susceptible to the epidemic strain was higher in case homes than in control homes (median 102 vs. 61, respectively; p = 0.0047) and was the most important factor selected by a stepwise, multivariate logistic regression model (p less than 0.0005). These findings are consistent with mathematical models of vaccine-induced herd immunity in closed populations and suggest that high rates of vaccination can reduce the risk of influenza outbreaks in this setting.

From December 10, 1982, to March 4, 1983, when influenza A (H3N2) viruses circulated in Michigan, outbreaks of influenza-like illness were identified in seven nursing homes in Genesee County; 272 (27%) of 1,018 residents were affected. Unvaccinated residents were more likely than vaccinated residents to become ill (risk ratio [RR], 2.6; 95% confidence interval [Cl], 1.8-3.6) and were subsequently more likely to be hospitalized (RR, 2.4; 95% Cl, 1.2-4.8), develop roentgenographically proven pneumonia (RR, 2.9; 95% Cl, 1.6-5.3), or die (RR, 5.6; 95% Cl, 1.2-9.1). Similar observations were made during investigations in six of the eight remaining nursing homes in Genesee County, in which 57 (12%) of 458 residents became ill sporadically. These findings suggest that influenza vaccine can reduce the incidence and severity of influenza virus infections among the elderly and chronically ill and underscore the importance of vaccination programs for those in nursing homes and in the general community.

No abstract

The organization and outcome of influenza immunization programs were studied in 67 randomly or systematically selected nursing homes (8354 residents) in six states during the autumn of 1982 and/or 1983. In each home, influenza vaccine was usually offered to all residents on a voluntary basis, independent of their age, level of required nursing care, or underlying medical conditions. However, the proportion of residents who were vaccinated ranged from 8 to 98% (mean, 62% overall), with significantly lower rates in homes that also required consent from relatives (usually by return mail) than in homes that did not (P less than .00001; median, 57 versus 90%, respectively). These observations suggest that distribution of educational materials about the risks and benefits of influenza vaccine and systematic follow-up of relatives who fail to return the consent form may be useful strategies to further increase the number of nursing home residents who are immunized.

No abstract