A health program's effect on neonatal mortality in eastern Kentucky.

We examined the impact of an innovative rural health program in central Appalachia on neonatal mortality by comparing improvements in the neonatal mortality rates of the serviced counties to those of the 13 control counties. Using data on all births and neonatal deaths in the study and control counties from 1970 to 1978 (tabulated by mother's county of residence), an expected neonatal mortality rate of each county was calculated by standardizing for the most important risk components: birth weight, sex, and birth multiplicity. An index of medical care effectiveness, the standardized mortality rate (SMR) for the study and control counties was computed by dividing the observed neonatal mortality rate by the expected rate. The SMRs of the study and control counties were then compared using a standard normal test statistic. The results indicated that the SMRs were significantly different in the control and study counties in the preintervention period 1970-1972. In subsequent years the SMRs converged, resulting in nearly equal rates for 1977-1978. The improvement in neonatal mortality in both areas may be due to their improved economic status rather than health care alone.