Study setting
This paper is based on a secondary data analysis of the Performance Monitoring for Action Ethiopia (PMA) project [19]. PMA Ethiopia is implemented in collaboration between Addis Ababa University, Johns Hopkins University, and the Federal Ministry of Health to produce policy-relevant reproductive, maternal, neonatal, and child health results. Household and health facility data were collected from five regions and one city administration: Tigray, Afar, Amhara, Oromia, the Southern Nations, Nationalities and Peoples regions, and Addis Ababa. In Afar, only rural and in Addis Ababa, only urban strata were included, while the rest had both.
Study design and data
This is a secondary analysis of a longitudinal study collected from pregnant and postpartum women aged 15-49 years from five regions and one city administration, at initial screening, six weeks, six months, and one year after birth. We analyzed the six-week postpartum interview along with the PMA cross-sectional data assessing facility readiness.
Data were collected from 2714 women aged 15-49 years and 462 health facilities from five regions and one city administration from October 2019 to January 2020.
Data on antenatal care visits collected from women assessed at initial screening and six weeks postpartum were linked to the facility data assessing maternal health service delivery points for the included women. We used ecological linking method [20] to link facility readiness assessment data to women data using enumeration ID, as the same enumeration ID was given for the clusters from which households and health care facilities were selected.
Eligibility
For this analysis, we included consenting pregnant or postpartum women 15-49 years old who were regular household members. The analysis also included data on healthcare facilities in the enumeration areas where the eligible women resided.
Sampling and sample size
Household data were collected based on two-stage cluster sampling. Sample size was estimated to detect a 5% difference between groups of women defined by various reproductive, maternal and child health indicators. Considering an alpha level of 0.05 and a power of 0.8, a minimum sample size of 3100 women was needed [19].
Data collection
Trained field workers collected data. The training included a review of survey protocols, questionnaire content, and interview skills. In addition to classroom exercises, field staff training included three days of field exercises, during which data collectors practiced using the tools on the data collection device (phones) [19]. The following background characteristics and service details were included in the study: region, residence, household wealth quintile, family size, age of the woman, marital status, gestational age at first ANC visit, number of ANC visits, service provider at first ANC visit, components of ANC received, any complication during recent pregnancy, and partner's encouragement to utilize maternal health services.
Measurements
The contact coverage was defined as the proportion of women who received four or more ANC visits during their most recent pregnancy. The input-adjusted effective ANC coverage was defined as the proportion of women receiving four or more ANC visits from healthcare facilities with skilled professionals, equipment, drugs, and supplies necessary for ANC service delivery, and emergency transport, as defined in Box 1. The facility readiness score was computed by taking the mean score of tracer items that should be available to provide ANC services at the healthcare facilities serving the included women [21]. In computing the facility readiness score, we included all health facilities that provide antenatal care for the included women within each EA. Input-adjusted ANC coverage was computed by multiplying the proportion of women who attended four or more ANC by facility readiness score.
Intervention-adjusted coverage was defined as the proportion of women receiving four or more ANC visits with both tetanus vaccination and iron-folate supplementation in the facility where the necessary inputs were available. These two were chosen as they are compulsory interventions of great importance for the woman and her newborn. Intervention-adjusted coverage was computed by multiplying the proportion who attended four or more ANC visits by facility readiness score and the proportion receiving both tetanus vaccination and iron folate tablets.
The quality-adjusted coverage was defined as the proportion of women receiving four or more ANC visits and the two mentioned interventions per the recommended process quality at a facility where the necessary inputs were available (Box 2). The process quality score was calculated by taking the mean score of nine service components. Quality-adjusted coverage was computed by the proportion of women who received four or more ANC multiplied by the facility readiness score multiplied by intervention coverage multiplied by the process quality.
Effective ANC coverage: We took quality adjusted ANC coverage as a proxy measure of effective coverage, as we don’t have information on the other two components of the effective coverage cascade (client adherence and outcome adjusted coverage) as recommended by the “Effective coverage Think Tank Group” [11] in the PMA data.
Data analysis
Background characteristics of women were summarized using descriptive summary measures. We used a STATA command svy to account for the clustered PMA data taking enumeration area as a primary sampling unit and households as a secondary sampling unit. The analysis was weighted by enumeration area and household to ensure the sample’s representativeness [19]. ANC coverage was stratified by region (Tigray, Afar, Amhara, Oromia, the Southern Nations, Nationalities, and Peoples region, and Addis Ababa) and urban or rural residence.
Box 1: Services and tracer items used to measure facility readiness to provide antenatal care, PMA Ethiopia, 2019-2020
1.
|
Availability of hemoglobin test
|
2.
|
Availability of blood group test
|
3.
|
Availability of blood glucose test
|
4.
|
Availability of venereal disease research laboratory test
|
5.
|
Availability of urine dipstick
|
6.
|
Availability of fetoscope
|
7.
|
Availability of blood pressure apparatus
|
8.
|
Availability of weight scale
|
9.
|
Availability of iron-folate supplementation
|
10.
|
Availability of tetanus vaccination
|
11.
|
Availability of private room for antenatal care
|
12.
|
Access to emergency transport
|
Box 2. Components included in antenatal care processes quality
1.
|
Blood pressure measured
|
2.
|
Weight taken
|
3.
|
Blood sample taken
|
4.
|
Urine sample taken
|
5.
|
Provider discussed healthy diet
|
6.
|
Provider discussed where to go for delivery
|
7.
|
Provider discussed transport for delivery
|
8.
|
Provider discussed dangers of bleeding before delivery
|
9.
|
Provider discussed dangers of high blood pressure
|