The Data Management & Methods Core (DM&M) will centrally unify the j-renewal program project and serve all three proposed research projects and the other cores by providing the data management backbone of the P01. Core C will build analysis files for each project in accordance with investigators' specifications, will work with project investigators to construct and test independent and dependent variables that are conceptually and theoretically appropriate for use across projects, will assemble data for dissemination on our program project website (www.LTCFocUS.org), and develop statistical models and methods specific to the complex administrative data utilized in the projects.
Specific aims are to: 1) assemble project data and develop methods for tracking and cleaning projects' longitudinal data; 2) create uniform, core measures for use in all project analyses; 3) provide analytic and statistical support; 4) develop novel methods of measuring functional status across post-acute settings; and, 5) assemble data for dissemination on the project website, managed by Core A. The DM&M Core is the heart of the common overlap across the projects since all rely upon the same core of longitudinal Medicare claims data and PAC assessment data all of which is matched to 100% of Medicare beneficiaries. Core C will integrate 2010 thru 2014 data to the existing decade of pane| data. We will apply Rasch measurement modeling within the framework of a missing data model in order to create synthetic measures of functioning that can be derived from MDS data on NH residents, IRF-PAI data on rehabilitation patients and OASIS data on home health patients. These integrated measures will be applied as independent and outcome variables in the 3 projects. Core C programmers will create analysis file programs for each project and continue to update the information on our web site and to disseminate it to dissemination partners ranging from the Commonwealth Fund to MedPAC and to trade associations in the nursing home industry
This research core provides methodological and data management support for all projects included in the Program Project and produces de-identified data for public dissemination on our project web site.
Showing the most recent 10 out of 160 publications
