DESCRIPTION (provided by applicant): Long-stay nursing home residents, those who stay in the facility six months or longer, often remain institutionalized until their death. Intervening illnesses are common, so residents' care goals can evolve dramatically over time. Activities of Daily Living (ADL)-basic personal care activities such as dressing, eating, and mobility-are closely related to health, mortality, and quality of life for nursing home residents. Thus, while no single piece of information can determine care for a resident, change in a resident's ADL status can indicate new care needs as diverse as starting rehabilitation or switching to comfort care. Despite the importance of ADL in the care planning process, we have only a rudimentary understanding of the timing and pattern of ADL change among nursing home residents and the important predictors of these patterns. Moreover, nursing home residents are frequently hospitalized, an event that is often associated with a decline in ADL function. Few studies, however, have examined the relationship between acute hospitalizations and the patterns of ADL change in nursing home residents, particularly the expected ADL course after return to the nursing home. This application proposes to advance the field by examining the patterns of ADL change through time (ADL trajectories). Two large national data sets, Medicare claims data and Minimum Data Set (MDS) assessments, will be used to investigate ADL trajectories for a representative cohort of long-stay nursing home residents. Merging these resident-level data with national data on facilities and health care resources, a multifaceted study of ADL trajectories will be conducted, focusing particularly on the role of acute events involving hospitalization. Aim 1 will begin with hierarchical modeling, with repeated measures of residents' ADL function as the dependent variable. Using the output from these analyses, cluster analysis will be employed to identify ADL trajectory categories (1) preceding and following hospital stays and (2) removed from acute hospitalizations. In Aim 2, the resulting ADL trajectory categories will constitute a dependent variable for multinomial logistic regressions that model post-hospital ADL trajectory categories and mortality conditional on pre-hospital trajectory categories and resident characteristics. Aim 3 will build on Aim 2 by testing the association of facility characteristics (e.g., staffing and rehabilitation environment) with ADL trajectory. Aim 3 will use the same multinomial dependent variable as Aim 2, and also include facility case-mix and resident characteristics as control variables. The results of this innovative study will provide an important first step towards providing residents, nursing home staff, physicians and families the opportunity to include likely ADL trajectories when weighing care options. Such information can help define the appropriate care pathway, whether it is prolonging life, preserving physical and cognitive function, or maximizing comfort. PUBLIC HEALTH RELEVANCE: Although a nursing home resident's ability to perform basic activities such as dressing and eating is important to his or her health and quality of life, we know little about how these abilities change with time or what causes the changes. We propose to study how illnesses that lead to hospitalization affect a resident's ability to perform these important tasks. This research will give residents, physicians, nursing homes, and families important information to help define the resident's main care goals, such as a focus on prolonging life, preserving physical and mental abilities, or maximizing comfort.