A comparative study was conducted in two neighboring states, Tennessee and North Carolina, to determine whether Medicaid managed care (implemented in Tennessee as TennCare) affected prenatal care, care patterns at labor-delivery, and birth outcomes. A pre- and post-design coupled with a difference-in-difference approach—using North Carolina as a control—was used to assess TennCare's effects for all births and for three categories of high-risk mothers (under age eighteen, unwed, or living in high poverty areas). Data from 328,296 singleton births in birth files and matched birth-death files for 1993 and 1995 in both states were used to analyze a number of variables related to maternal behavior during pregnancy,utilization of care before and after labor-delivery, patterns of obstetrical care at delivery, and birth outcomes.
Under TennCare, Tennessee mothers were relatively more likely to obtain no prenatal care or to wait and initiate third trimester care as compared to those in North Carolina. Relative utilization of specific prenatal procedures declined, Apgar scores fell very slightly, and birth abnormalities increased in the poverty subsample. TennCare had no significant effect on infant mortality. Utilization reductions in obstetrical services were achieved with apparent spillovers to non-TennCare births, but without adverse effects overall. TennCare was neither a panacea nor an unmitigated disaster. It is a model worth examining, but not uncritically.