scalability of health services: an empirical test

Abstract

The implementation of five programs considered to contribute to an array of comprehensive health services is investigated by Guttman scale analysis in a sample of United States hospitals and health departments to determine whether the programs are undertaken in a unidimensional or a random sequence. Evidence is presented that the mode of implementation of these programs is unidimensional in hospitals, but program implementation in health departments is found to follow a different pattern and to show no scalability. The conclusions are not changed when the data are controlled for size of facility and regional location. Implications of these findings for evaluating comprehensiveness of health services and for predicting future innovations are discussed.