In an attempt to control Medicaid nursing facility utilization and expenditures, states have implemented preadmission screening (PAS) programs to assess applicant need for nursing facility services and, for some programs, their potential to remain in the community with the assistance of alternative long term care services. This study examined the structure and implementation of state PAS programs from 1978-1994 in all 50 states and the District of Columbia. Primary data on state PAS program characteristics were collected from state officials in three separate telephone surveys in 1989, 1992, and 1994.

Twenty-five states screened applicants by a paper or telephone review of information collected by a private provider in 1994. The other twenty-six states used state or contract agency staff to screen all or some applicants. The number of states implementing these types of PAS programs increased steadily until 1990. Most growth took place from 1981 (5 states) to 1984 (16 states). Programs became increasingly comprehensive until 1990, with a higher percentage using state or contract agency staff to conduct assessments, screening hospital and community-based applicants, and screening private pay applicants. A cross-sectional logistic regression of state socio-demographic, economic, and political characteristics found wealthier states with a large elderly population were positively associated with having a PAS program in 1941, while the percent of a state's elderly population with membership in the American Association of Retired Persons was negatively associated.

A stringency index was created by assigning states a score of 0 to 6 based upon the presence of selected program characteristics. Five states received the highest score of 6, while eleven received a score of 0. Sixteen states received a score of 4 or higher. Stringency scores were included in a cross-sectional two-stage regression analysis of state economic, socio-demographic, and health service data from 1992. No association was found between states with high stringency scores and Medicaid nursing facility utilization. A second analysis found PAS programs targeted towards Medicaid eligibles were significantly associated with lower Medicaid nursing facility utilization in 1991 and 1992. Future research should expand the utilization model, pool data from multiple years, and refine the stringency index.