Addressing maternal and child health equity through a community health worker home visiting intervention to reduce low birth weight: retrospective quasi-experimental study of the Arizona Health Start Programme

Objective: To test if participation in the Health Start Programme, an Arizona statewide Community Health Worker (CHW) maternal and child health (MCH) home visiting programme, reduced rates of low birth weight (LBW), very LBW (VLBW), extremely LBW (ELBW) and preterm birth (PTB).

Design: Quasi-experimental retrospective study using propensity score matching of Health Start Programme enrolment data to state birth certificate records for years 2006-2016.

Setting: Arizona is uniquely racially and ethnically diverse with comparatively higher proportions of Latino and American Indian residents and a smaller proportion of African Americans.

Participants: 7212 Health Start Programme mothers matched to non-participants based on demographic, socioeconomic and geographic characteristics, health conditions and previous birth experiences.

Intervention: A statewide CHW MCH home visiting programme.

Primary and secondary outcome measures: LBW, VLBW, ELBW and PTB.

Results: Using Health Start Programme's administrative data and birth certificate data from 2006 to 2016, we identified 7212 Health Start Programme participants and 53 948 matches. Programme participation is associated with decreases in adverse birth outcomes for most subgroups. Health Start participation is associated with statistically significant lower rates of LBW among American Indian women (38%; average treatment-on-the-treated effect (ATT): 2.30; 95% CI -4.07 to -0.53) and mothers with a pre-existing health risk (25%; ATT: -3.06; 95% CI -5.82 to -0.30). Among Latina mothers, Health Start Programme participation is associated with statistically significant lower rates of VLBW (36%; ATT: 0.35; 95% CI -0.69 to -0.01) and ELBW (62%; ATT: 0.31; 95% CI (-0.52 to -0.10)). Finally, Health Start Programme participation is associated with a statistically significant lower rate of PTB for teen mothers (30%; ATT: 2.81; 95% CI -4.71 to -0.91). Other results were not statistically significant.

Conclusion: A state health department-operated MCH home visiting intervention that employs CHWs as the primary interventionist may contribute to the reduction of LBW, VLBW, ELBW and PTB and could improve birth outcomes statewide, especially among women and children at increased risk for MCH inequity.