Nursing Home Policy via Hierarchical Duration Models. Abstract, ExecutiveSummary, and Final Report

摘要

The primary goal was to develop and apply statistical methods for estimation ofduration (survival) times in nursing home length of stays and in other relevant health policy applications. Inproved statistical analyses of duration data has increased our understanding of nursing home spend-down and of the long-term care demands on Medical resources. Additional, the methods developed expand the set of policy questions that can be answered in other data settings where problems of length-bias, multiple outcomes, heterogeneity, and censoring occur. Our contributions toward this objective resulted in six manuscripts on policy and methodological topics. (1) heirarchical models for duration analyses, (2) learning-by-doing in organ transplant centers, (3) multiple outcomes in survival models, (4) length bias correction in survival models, (5) parametric analysis of nursing home lengths of stay, and (6) spend-down in nursing homes. These manuscripts and the presentation of this work in medical schools, schools of public health, and in national and international conference have helped to improve understanding of long-term care and of analytical methods.