Improved Targeting of Long Term Care Discharges during Hospitalization. Abstractand Executive Summary of Dissertation

摘要

Hospitals are now increasingly admitting large numbers of elderly patients, asubset of which are at risk of needing admission to long term care (LTC). Clinicians are not always able to identify correctly those patients at high risk of LTC discharge. Delays in discharge planning may result in increased time spent in the hospital. Since hospitals are not fully reimbursed for these days, financial losses are associated with them. Targeting patients at risk of needing LTC discharge has the potential of reducing hospital cost associated with discharge delays. The purpose of this study is to assess how the risk of three hospital discharge outcomes differ across subgroups of elders. This study relies on two administrative data files from a large tertiary care teaching hospital in the Northeast, which provide information on the daily nursing needs of patients. The study design includes two empirical models. Three survival analysis models measure the risk of the three discharge destinations, death, discharge to LTC and discharge home with services, conditional upon survival through the current hospital day. These models are compared to a multinomial logit model estimating the risk of discharge to the same three destinations compared to discharge home without services. The principal findings are that indicators of nursing need are strong predictors of all three destination outcomes under study. They are stronger predictors of the outcomes than are demographic or diagnostic covariates.