Measuring the effectiveness of a telephonic maternal health program

Background: Age, preexisting conditions, and pregnancy complications can place the mother, the newborn, or both at risk for adverse outcomes and increased cost. Telephonic care management programs provide guidance and ongoing monitoring to expectant mothers and can improve pregnancy outcomes and reduce costs. Objectives: The purpose of this study is to demonstrate methods to measure the impact of participation in a telephonic maternity support program. Methods: Using medical claims and program operational data, we analyzed mothers in a commercial insurance program who had live births between 2010 and 2011 (n= 18,500). We compared costs between mothers who were actively engaged in the program at least 120 days before delivery (n=5,721) and mothers who were referred to the program but did not participate (n=10,636). A propensity score model was developed to account for differences in demographics and risk factors between the engaged and non-engaged mothers. Using inverse propensity score weights, a series of generalized linear models were built to measure the impact of program engagement on pregnancy, delivery, and resulting NICU newborn costs. Results: This evaluation showed statistically significant (p-value <0.05) reductions in NICU costs, pregnancy costs, and delivery costs ($33,623; $226; $245 in savings, respectively) among the mothers engaged in the program as compared to the non-engaged mothers. Conclusions: By adjusting for differences in demographics and risk factors and creating separate models for each cost component, this study has provided a reliable method for demonstrating the impact of a maternal support program.