Measurement of early initiation of breastfeeding: a mixed methods study exploring accuracy, challenges and implications for tracking newborn health targets

Background Early initiation of breastfeeding within one hour after birth is one of the recommended interventions for saving newborn lives. Despite WHO standards, there are still inconsistencies between surveys in how the indicator is measured. This thesis identified challenges and accuracy issues in the measurement of early initiation indicator and proposed improvements to household survey tools. Objectives Individual studies included in this thesis assessed the following: the evidence-base for the ‘one hour’ time limit in defining the indicator; mothers’ perception of the survey question; newborn care practices other than breastfeeding; and data collectors’ experiences of asking mothers the early initiation question. Methods Mixed methods were used and data was collected using a critical review of the literature, cognitive interviews, semi-structured interviews, focus group discussions, key informant interviews and a self-administered questionnaire. All primary data was collected from Ethiopia. Results Nine out of 11 studies included in the review provided evidence supporting the ‘one hour’ time limit in the early initiation indicator. Primary data collection in Ethiopia revealed that some mothers did not understand the question as intended. Considerable probing was required and there were inter-personal variations among data collectors in how they asked the question. In the context of home births, other competing newborn care practices were given priority over breastfeeding initiation. Conclusion This thesis provided evidence supporting the ‘one hour’ time limit for the indicator and accordingly, consistent use of standard WHO definition is recommended. To avoid ambiguity, the early initiation question should be revised as “how long after birth did you first put [name] to the breast even if your breastmilk did not arrive yet?” Standard probes or follow on questions are required to avoid misinterpretation of the indicator by data collectors. Probes should be designed using other newborn care practices as reference points to improve maternal recall.