Evaluation protocol to assess maternal and child health outcomes using administrative data: a community health worker home visiting programme

INTRODUCTION: Emerging evidence suggests community health workers (CHWs) delivering preventive maternal and child health (MCH) interventions through home visiting improve several important health outcomes, including initiation of prenatal care, healthy birth weight and uptake of childhood immunisations. METHODS AND ANALYSIS: The Arizona Health Start Program is a behavioral-based home visiting intervention, which uses CHWs to improve MCH outcomes through health education, referral support, and advocacy services for at-risk pregnant and postpartum women with children up to 2 years of age. We aim to test our central hypothesis that mothers and children exposed to this intervention will experience positive health outcomes in the areas of (1) newborn health; (2) maternal health and healthcare utilisation; and (3) child health and development. This paper outlines our protocol to retrospectively evaluate Health Start Program administrative data from 2006 to 2015, equaling 15 576 enrollees. We will use propensity score matching to generate a statistically similar control group. Our analytic sample size is sufficient to detect meaningful programme effects from low-frequency events, including preterm births, low and very low birth weights, maternal morbidity, and differences in immunisation and hospitalisation rates. ETHICS AND DISSEMINATION: This work is supported through an inter-agency contract from the Arizona Department of Health Services and is approved by the University of Arizona Research Institutional Review Board (Protocol 1701128802, approved 25 January 2017). Evaluation of the three proposed outcome areas will be completed by June 2020.