Introduction: The Maternal Infant Health Program, MIHP, is Michigan's largest home visiting program for Medicaid-eligible pregnant women and infants. Prior quasi-experimental propensity score matched MIHP evaluations were limited by the possibility of selection bias due to the limited available baseline characteristics. The purpose of this study was to assess the effectiveness of MIHP including previously unobserved baseline characteristics to mitigate the possibility of selection bias.

Methods: We analyzed health care and outcomes of mothers and their infants using linked administrative and screening data. All Medicaid pregnant women with a singleton birth in Michigan between 2009 and 2012 who were prenatally screened into the statewide home visiting program were studied (n=69,408). Propensity score adjustment regressions were used to compare screened-only program participants to those screened who received additional program services. Data were assembled and analyzed between 2014-2016.

Results: Compared to screened-only pregnant women, those who received additional program services had higher odds of receiving adequate prenatal care (OR=1.12), an appropriate postnatal checkup (OR=1.27), reduced odds of low birthweight (OR=0.84), very low birthweight (OR=0.63), preterm (OR=0.83), or very preterm birth (OR=0.67), and infants had increased odds of receiving the appropriate number of well-child visits in the first year of life (OR=1.3, p

Conclusions: Favorable home visiting effects on maternal and infant care and health previously observed in quasi-experimental matched evaluations were supported by this study after accounting for an expanded set of characteristics and risk factors which reduced the possibility of selection bias.