A prominent program that claims to reduce infant and maternal deaths in rural India by encouraging mothers to deliver in private hospitals has been unsuccessful, despite the investment of more than $25 million since 2005, a new Duke University study finds.

The Chiranjeevi Yojana program in Gujarat, a state in northwestern India, received the Wall Street Journal Asian Innovation Award in 2006 and has been hailed by some as a model for wide adoption throughout India.

The program was launched in 2006 to help address the shortage of obstetricians at public hospitals accessible to low-income women in rural areas. It aimed to provide free childbirth care at participating private-sector hospitals to women who are below the poverty line. The hospitals are paid 1600 Indian rupees per delivery, approximately $30 to $40. The hospitals may offer additional services to patients and charge separate fees for them. By 2012, approximately 800 private-sector hospitals were participating and the program had helped pay for more than 800,000 deliveries.

Manoj Mohanan, Duke assistant professor of public policy, global health and economics, led the research team. The team surveyed 5,597 households in Gujarat to collect data on births that had occurred between 2005 and 2010. They found no statistically significant change in the probability of delivery in health care institutions, the probability of obstetric complications or the probability that physicians or nurses were present during childbirth.

"We were surprised to find, as well, that even among those who delivered at health care facilities there were no significant reductions in households' out-of-pocket expenditures for deliveries."

The findings were published online by the peer-reviewed international public health journal, Bulletin of the World Health Organization, in an article titled, "Impact of Chiranjeevi Yojana on institutional deliveries and birth outcomes in Gujarat, India: a difference-in-differences analysis."

While the study did not determine why patients' delivery costs did not decline or why the program is ineffective, several explanations are possible, Mohanan said. Media reports in India suggest that despite the promise of free care, hospitals were billing women for extra, chargeable services. Providers also complained that the reimbursement amounts were not adequate to cover costs of providing the service.

In addition, mothers may perceive the quality of care at participating private hospitals to be poor, so even when the care is provided for free, demand does not rise. Transportation costs from rural villages also could be a factor, he said.

Mohanan said previous research, which had suggested the program was very successful, had methodological limitations. It did not address the role of self-selection of institutional delivery by pregnant women and did not account for unrelated increases in institutional deliveries that probably occurred as a result of rapid economic growth in the region.