BackgroundA peculiar phenomenon of engaging insurance intermediaries for government funded health insurance schemes for the poor, not usually found globally, is gaining ground in India. Rajiv Aarogyasri Scheme launched in the Indian state of Andhra Pradesh, is first largest tax funded community health insurance scheme in the country covering more than 20 million poor families. Aarogyasri Health Care Trust (trust), the scheme administrator, transfers funds to hospitals through two routes one, directly and the other through an insurance intermediary. The objective of this paper is to find out if engaging an insurance intermediary has any effect on cost efficiency of the insurance scheme.MethodsWe used payment data of RAS for the period 2007–12, to find out the influence of insurance intermediary on the two variables, benefit cost ratio defined as benefit payment divided by premium payment, and claim denial ratio defined as benefit payment divided by treatment cost. Relationship between scheme expenditure and number of beds empanelled under the scheme is examined. OLS regression is used to perform all analyses.ResultsWe found that adding an additional layer of insurance intermediary between the trust and hospitals reduced the benefit cost ratio under the scheme by 12.2 % (p-value = 0.06). Every addition of 100 beds under the scheme increases the scheme payments by US$ 0.75 million (p-value < 0.001). The gap in claim denial ratio between insurance and trust modes narrowed down from 2.84 % in government hospitals to 0.41 % in private hospitals (p-value < 0.001).ConclusionsThe scheme is a classic case of Roemer's principle in operation. Introduction of insurance intermediary has the twin effects of reduction in benefit payments to beneficiaries, and chocking fund flow to government hospitals. The idea of engaging insurance intermediary should be abandoned.