Computer Detection of At-Risk Nursing Home Residents

This project reflects the next steps in developing an automated system for detecting changes in NHnutritional intake and body weight; changes assumed to be associated with health status. While NH remonitoring resident nutritional status and body weight, no reliable and efficient system for detectiexists. Pilot work indicates that automated resident-specific monitoring may result in more reliable"significant" changes than with commonly used methods. Consequently, Phase I is directed toward expawork, establishing operational definitions and criteria for significance of change from a broad groupractitioners, validating efficient data capture and analytical protocols, establishing the efficacyestablishing the relationship of detected changes to clinical implications, and establishing the endcharacteristics for the automated system. The prototype will be translated into a fully functioninggeneralization will be tested in diverse NHs, and we will further evaluate clinical diagnostic and iPhase III will focus on dissemination/marketing. The ideal result will be earlier detection of at-riintervention and health cost savings.