Selection Bias in Prenatal Care Utilization: An Interdisciplinary Framework and Review of the Literature

Abstract

Since there are no randomized trials of standard prenatal care, it is difficult to assess its impact on birth outcomes without controlling for selection processes that can inhibit or promote its use. In this article, we develop a typology of selection processes in prenatal care utilization, identifying four distinct types of selection and their possible biasing effects on estimates of prenatal care efficacy. Second, using an interdisciplinary frame-work, we review the published studies on birth outcomes that consider selectivity in prenatal care, all of which have been published in the economics literature. The results from these studies suggest that selectivity in the use of care does exist and that the predominant process is one of adverse selection. This implies that analyses failing to control for selection will underestimate the effects of prenatal care. Third, we discuss the public health policy implications of these findings and offer an agenda for future research.