The Impact of Flu and Flu Vaccine on NH Resident Morbitidity

点击翻译按钮获取中文摘要

DESCRIPTION (provided by the applicant): Influenza is one of the most important causes of excess morbidity and mortality which disproportionately affects the elderly, leading to death and hospitalization. Nursing home (NH) residents are most likely to bear the full impact of influenza-related morbidity and mortality. Their immune and physiologic senescence, multi-morbidity, and greater exposure risk through close living quarters and shared caregivers contribute to their increased susceptibility. Influenza is one of the leading causes of infectious disease outbreaks in NH and efforts to increase the rates of preventative vaccination of this population have had varying levels of success. To date, few studies have examined the effect of influenza and preventative vaccination on NH residents' decline in physical functioning, something that is most germane to this population for both residents and their caregivers. This proposal addresses the AHRQ research agenda related to prevention and care management as well as patient safety, all framed within the broader agenda of long term care quality and quality measurement. Using ten years of comprehensive, longitudinal resident assessment data (the MDS) on all NH residents in all US NHs linked to Medicare enrollment and hospital claims data as well as to weekly CDC influenza prevalence and severity data in 122 surveillance cities, we propose to build a model to estimate the impact of localized influenza severity and annual vaccination coverage in each resident's nursing home on the rate of functional decline, infection, weight loss and both influenza related and all cause hospitalizations. Our modeling approach to estimation and testing the significance of the vaccination effect on hospitalization rates and our MDS-based morbidity outcomes is based on an extension of the differences-in-differences (DID) model applied to longitudinal data that incorporates the strong seasonal aspects of influenza exposure. In addition to generating unbiased estimates of the effect of influenza and influenza vaccination on NH residents, our model will make various other important predictions relevant to clinical and policy concerns and document the need for "seasonally adjusting" CMS' publicly reported nursing home quality measures.