Obstetrical ultrasound in remote communities: an approach to health program evaluation

This paper illustrates one approach to evaluating the impact of an obstetrical ultrasound outreach program in remote Indian communities. The pregnancy characteristics and obstetrical outcomes of 240 deliveries in 2 groups of communities with and without the program are compared. After controlling for maternal age, parity, obstetrical risk score and presence of diabetes and hypertension during pregnancy, Program Area women are found to differ significantly from Control Area women in several medical care measures. There are more health care contacts, fewer medical referrals out of the community and a lower proportion of non-hospital births. There are however more inductions of labour and longer hospital stays. The small sample size was not sufficient to detect differences in rare events such as perinatal deaths. Despite its shortcomings, the quasi-experimental design provides health planners with important information regarding medical technologies and interventions, particularly when randomized controlled trials are deemed not to be feasible and the only alternative is unsubstantiated opinions.