Replication data for: Quantity and content of antenatal care visits increase the use of skilled birth attendance in rural Africa

Skilled birth attendance (SBA) at delivery is a critical public health intervention that plays an important role in preventing maternal and neonatal births around the world. The Safe Motherhood Interagency Group declared in 1997 that SBA was the “single most critical intervention” to reducing maternal mortality.1 The World Health Organization (WHO) advocates for “skilled care at every birth,” defining SBA as an accredited health professional who has the skills necessary to manage uncomplicated pregnancies, childbirth and the immediate postnatal period.2 Yet despite the evidence and global support for SBA, it remains an under-utilized intervention. From 2006 to 2013, only 46% of women in low-income countries had skilled care at delivery.2
Antenatal care (ANC) is one avenue through which SBA coverage can improve. Among its other benefits, such as identifying risk factors during pregnancy, ANC provides a venue in which to educate mothers about the benefits of SBA and encourage its use. WHO guidelines recommend that a woman should receive a minimum of four ANC visits in her pregnancy.3 These policy recommendations have shaped global health discussions and funding patterns. In 2010 alone, the international health community spent $6 billion on maternal, newborn, and child health, 37% of which went to ANC programs.4 Given the substantial amount of resources that are mobilized toward promoting antenatal care in place of alternative maternal and child health interventions, the effectiveness of ANC on increasing SBA is important to assess. This paper evaluates how both the quantity and the content of ANC visits affect the probability of SBA in four rural African settings.
While many studies have attempted to evaluate this effect, many fail to adequately account for potential endogeneity biases between ANC and SBA5. There may be many factors, such as distance to a health center, a woman’s perception of the health care system, and wealth that confound the relationship between ANC and SBA. Adjiwanou & LeGrand (2013) address this problem by using structural equation modeling (SEM) to eliminate the bias. Their study found a significant and positive effect of the quantity of ANC visits on SBA in Ghana, Kenya, Uganda, and Tanzania.5 This paper extends the work of Adjiwanou & LeGrand using more recent data, simplified methods of analysis, simulation of quantities of interest, and data visualizations. We focus on identifying the causal estimates of quantity and content of ANC visits on SBA and interpret how these estimates might be used by policy makers.