Improving Maternal and Infant Health through CenteringPregnancy: Results of a 2-Year Retrospective Chart Review Using a Matched Comparison Design




Healthy People 2020 goals include reducing infant mortality and morbidity associated with pregnancy. CenteringPregnancy is a model of health care delivery that unifies health assessment, education, and support into a group setting that has been shown to improve prenatal care compliance, prenatal health knowledge, and prematurity.

Research Question

Is there a difference in select maternal health indicators and pregnancy outcomes between women participating in CenteringPregnancy and traditional care?


An analysis was conducted comparing CenteringPregnancy participants with a matched comparison group (traditional care). Data were collected through retrospective records review. Institutional review board approval was obtained from Western Michigan University and Borgess Medical Center prior to data collection. For the period of data collection (January 2010 through April 2012), 173 CenteringPregnancy participants were identified and a comparison group of 170 traditional certified nurse-midwife (CNM) care clients were selected matched on race, age, and insurance status. Additional maternal data collected included prior pregnancies, height, and prepregnancy weight. Prenatal variables included number and timing of prenatal care visits, weight gain, and smoking status. Intrapartum variables included mode of birth and gestational age. Postpartum data were collected regarding attendance at a 6-week postpartum visit and breastfeeding.


There was a significant between-groups difference in gestation at first prenatal visit (11.8 weeks for traditional care vs 10.3 for CenteringPregnancy, P = .031). There was not a significant difference in mean number of prenatal visits (14.2 for CenteringPregnancy vs 13.4 for traditional care, P = .266). There was not a significant difference in rates of smoking at pregnancy diagnosis (26% [n = 45] of CenteringPregnancy vs 30% [n = 49] of traditional care); however, 69% (n = 31) of the women in the CenteringPregnancy group quit with pregnancy diagnosis versus 18% (n = 9) of the traditional care group (P < .001). Additionally, 50% (n = 7) of those still smoking quit during CenteringPregnancy versus only 8% (n = 3) during traditional care (P < .001). There was no difference in pre-pregnancy weights or weight gain between groups, with only 28% of traditional care and 25% of CenteringPregnancy participants gaining the optimal amount of weight. There was no difference in mean gestational age at birth (39.3 weeks for CenteringPregnancy vs 39.5 weeks for traditional care) or rates of preterm births (5.8% in CenteringPregnancy vs 5.9% in traditional care). There were no differences in cesarean rates (17% for CenteringPregnancy vs 14% for traditional care, P = .443). Both groups had high rates of attendance at their 6-week postpartum visit (92% for CenteringPregnancy vs 88% for traditional care, P = .243). CenteringPregnancy participants were more likely to initiate breastfeeding (79% vs 60% for traditional care, P < .001) and more likely to still be breastfeeding at the 6-week postpartum visit (65% vs 45% for traditional care, P < .001).


In this study, CenteringPregnancy demonstrated promise in reducing smoking rates during pregnancy and improving breastfeeding rates. This study also demonstrates that optimal weight gain guidelines must be addressed with women during pregnancy regardless of type of prenatal care. Findings from this study must be viewed in light of the self-selection bias. Despite our attempts to control for variance by utilizing a matched comparison design, women who chose CenteringPregnancy initiated prenatal care earlier and also were more likely to quit smoking prior to onset of prenatal care than those who chose traditional care.