Prevention of Influenza in Health Care Settings: Models and Observations

Abstract

In this thesis we present five studies that focus on the assessment of prevention and control strategies for influenza in health care institutions. In chapters 2 and 3, we show the difficulties concerning recognition of influenza virus infection in hospital patients and the low willingness of health care workers (HCWs) to receive influenza vaccination. In chapters 4 to 6 we develop and use a mathematical model of influenza virus transmission in health care institutions. With this model we demonstrate how an increase in the proportion of HCWs that is vaccinated, decreases the number of influenza virus infections in long-term care nursing home patients. The relationship between the fraction of HCWs vaccinated and the fraction of patients with an influenza virus infection appears to be linear and a threshold for herd immunity cannot be detected. Our results confirm and extend several previous observational and experimental studies that attempted to quantify the effect of HCW vaccination. In chapter 5, we use the transmission model to extrapolate our findings from the long-term care nursing home environment to a regular hospital environment, for which no data on preventive measures against influenza for HCWs are available from experimental or observational studies. According to the predictions by the model, HCW vaccination in this setting will turn out to be at least as effective and efficient as observed for a nursing home setting. Finally, we modify the model to study the effects and the efficiency of two strategies of prophylaxis with the antiviral agent oseltamivir. Although oseltamivir has been shown to be effective in reducing susceptibility, infectiousness and the probability of developing disease in household studies, its efficacy for outbreak control in health care institutions remains to be determined. Our simulations suggest that post-exposure prophylaxis is not so effective, but far more efficient, than continuous prophylaxis in the protection of nursing home patients against influenza virus infection. Increasing resistance of influenza strains against oseltamivir will reduce the effectiveness and efficiency of both strategies, but post-exposure prophylaxis exerts a weaker selection pressure for resistance than continuous prophylaxis. From our studies we conclude that patients can be best protected by increasing HCW vaccination, which is more effective and more efficient than post-exposure prophylaxis with oseltamivir. Nevertheless, as long as HCW vaccine uptake is suboptimal or in case of a mismatch between the vaccine strains and the circulating viruses, post-exposure prophylaxis for patients is an effective secondary control measure. However, if we aim to protect patients by protecting HCWs, our efforts should be directed on increasing HCW vaccine uptake rather than HCW prophylaxis.