Does health insurance matter? An analysis of how the type of health insurance affects medical procedure use during pregnancy and birth

The hypothesis, that more and better health insurance will lead to better health outcomes is intuitively easy to support. Like others, health providers are expected to respond to economic incentives. However, previous studies of this important issue by health economists do not conclusively answer the question of whether the type of health insurance affects inpatient hospital treatment, and if so, by how much.; Due to shortcomings in the data, recent studies use “probable” Medicaid eligibility, a limited number of hospital procedures and/or restricted demographic characteristics. This study alleviates these deficiencies by using the Statewide Planning and Research Cooperative System (SPARCS) data for New York State (NYS) merged with the Vital Statistics Linked Birth/Infant-Death data to examine the effects of insurance on procedure utilization during pregnancy and at birth. The data contain detailed information for each hospital stay including procedure utilization, risk factors, demographic data and physician identifiers.; To test the validity of previous methodologies, probable Medicaid eligibility is estimated for each woman giving birth in NYS, excluding New York City, for 1990–96. Theory dictates, that increasing Medicaid eligibility will increase procedure utilization for teens/dropouts, due to their going from being uninsured to having Medicaid and decrease utilization for non-teen/higher-educated women, due to their being crowded off private insurance onto Medicaid. The effects of probable Medicaid eligibility on procedure utilization supports the theory and previous studies for three of four procedures for teens and two of four procedures for non-teens. However, the theory is no longer supported when probable eligibility in each cell is replaced with the percentage of individuals who are actually covered by Medicaid.; The results from using individual observations in analyzing the effects of 10 types of HMO's and fee-for-service, insurance are small, but significant. Births involving medical risk factors and complications have greater probability of receiving 21 of 22 procedures. Controlling for individual hospital and physician heterogeneity further reduces the effects of insurance on procedure utilization. The presence of medical risk factors and/or complications of labor and/or delivery have a much larger effect on procedure utilization at birth than the type of health insurance.