Management Systems for Home Health Care Clients with AIDS. Final Report, Abstractand Executive Summary

摘要

The primary purpose of this project is to refine an existing home health patientclassification system to predict more accurately the length of home health visits to persons with AIDS. The system will also apply to other complex or lengthy visits and to the general home health caseload. The system will assist home health agencies to determine and monitor trends in acuity and staff utilization. In addition, the project will assist developers of home health prospective payment systems by generating information on home health utilization by persons with AIDS and others with lengthy visits, and by assessing the importance of including visit length in prospective payment systems. The study builds on previous work by the researchers on factors predicting home health visit length and by others on home health patient classification systems. A previously developed instrument. The Patient Classification System for Home Nursing (PCSHN), will be modified for this project based on recommendations from nurse focus groups and an Advisory Group to include the factors influencing AIDS' clients home health utilization. Data on 3,219 home health visits will be collected prospectively by nurses from 14 participating Massachusetts home health agencies over a 12-month period, using the modified PCSHN. Models to predict nurse time in minutes will be developed using regression analysis, with characteristics of the client, nurses, agency, and visit as the independent variables. Descriptive statistics and measures of variability will be compared across categories of patients. Actual visit length will be compared to length predicted by the modified PCSHN. Each client included in the study will also be classified using two case-mix systems currently under development for proposed federal home health prospective payment systems. The performance of those systems and the modified PCSHN in predicting resource needs of persons with AIDS will be compared.