Table 1

Methods to construct service readiness indices

Item selection
SARA tracer items for obstetric care*15 items for health centres plus seven additional items for hospitals based on the WHO SARA basic and comprehensive obstetric care tracer items.23 These correspond to three readiness domains: (A) staff and training; (B) equipment and (C) medicines and commodities.
DHS analytical study’s obstetric and newborn care readiness indicators*30 items for health centres plus three additional items for hospitals based on obstetric and newborn readiness indicators described in the DHS analytical studies No. 65.24 25 This includes items across five readiness domains: (A) performance of signal functions for emergency obstetric care; (B) performance of newborn care functions; (C) general requirements; (D) equipment and (E) medicines and commodities. The DHS programme proposes an additional domain for ‘guidelines, staff training and supervision’; however, this domain is excluded from the domain-weighted addition given limited availability of these items in the PMA-ET survey.
WHO standards for improving quality of maternal and newborn care readiness items*44 items for health centres plus eight additional items for hospitals available in the PMA-ET survey instrument mapped to the WHO Standards for improving quality of maternal and newborn care in health facilities.39 These include three ‘provision of care’ standards and two ‘cross-cutting’ standards: (1) evidence-based practices for routine care and management of complications; (2) actionable information systems; (3) functional referral systems; (4) competent, motivated human resources and (5) essential physical resources. These items are also grouped in four readiness domains: (A) equipment, supplies and amenities; (B) medicines and health commodities; (C) staffing and systems for quality and safety; and (D) performance of signal functions.
Aggregation method
Simple addition of itemsThe number of items that is available on the day of the assessment is added together. The number of available items is divided by the total number of possible items to compute a score ranging from 0 to 1. Each item is given equal weight.
Weighted addition of items by readiness domainWithin each readiness domain, the number of items that is available on the day of the assessment is added together. The number of available items per domain is divided by the number of possible items per domain to compute a domain score. The sum of the domain scores is divided by the number of domains to compute a score ranging from 0 to 1. Each domain is given equal weight.
Principal components analysis (PCA)PCA is a data reduction technique that converts a set of correlated items into orthogonal components. Each component explains some proportion of the variation across the items, with the first component explaining the largest proportion. The first component is extracted and rescaled to a score ranging between 0 and 1.
Composite indices
Combination of item selections with aggregation methodsEach of the item selections (1=SARA, 2=DHS, 3=WHO standards) are aggregated using three different methods (1=simple addition, 2=weighted addition, 3=PCA) to generate nine childbirth service readiness indices.
  • *Please refer to online supplemental tables S1 and S6 for the complete list of items selected for each of the readiness indices and for information on any items excluded due to lack of available data.

  • DHS, Demographic and Health Survey; PMA-ET, Performance Monitoring for Action Ethiopia; SARA, Service Availability and Readiness Assessment.