Reliability and validity of a new preadmission acuity tool for long-term care.

ABSTRACT

The purpose of this project was to develop a new tool designed to assess the acuity of applicants to long-term facilities. Using a known groups method, residents of long-term care facilities were categorized as independent (n = 68), assist (n = 71), and dependent (n = 65) and assessed using the Preadmission Acuity Inquiry Tool (PAI) to generate the data base for examining its psychometric properties. The progression of the mean scores from independent to assist to dependent demonstrates the ability of the tool to differentiate across varying levels of functional independence. Using discriminant analysis, the degree of accuracy in predicting group membership for the independent group was 88%, for the dependent group 75%, and for the assist group 17% with the overall agreement between PAI scores and a priori classifications at 60%. Kappa coefficients for interrater reliability ranged from .54 to 1.0 with a mean of .90. The Cronbach's alpha coefficient was .81. The PAI has demonstrated sufficient reliability and validity to warrant further refinement and testing. This new instrument may enable health care professionals to provide the appropriate level of care in a cost-effective manner.