Objectives: To generate regional, residential, and nationally representative estimates of knowledge of abortion legality and availability among women of reproductive age in Ethiopia, and examine how knowledge varies across regions and by urban/rural residence. Study Design: Our study draws on data from a nationally representative, cross-sectional survey implemented in 2019 in each of Ethiopia’s regional and administrative states, yielding a sample of 8,837 women aged 15 to 49. We compare weighted estimates and regional distributions of 3 outcomes: (1) general awareness and (2) correct knowledge of the abortion law, and (3) knowledge of facility-based abortion service availability. Results: Significant regional and urban/rural disparities in knowledge of abortion legality and availability exist. Nationally, 27% of women are aware of the abortion law and just 5% of women have comprehensive knowledge of the law, while 30% know where to access facility-based abortion services. Regionally, estimates range significantly, from 2% in Somali to 45% in Addis Ababa for general awareness of the law, 0% in Afar to 27% in Harare for comprehensive knowledge of the law, and 8% in Afar to 57% in Tigray for knowledge of abortion availability. Knowledge of all measures is higher in urban than in rural areas. Conclusions: Regional disparities in abortion knowledge may contribute to geographic inequities in access to and use of safe abortion care. Efforts to expand knowledge of abortion legality and availability are needed and should be tailored to regional contexts. Implications: Knowledge of abortion legality and availability is imperative to protecting and expanding access to safe abortion care, especially in contexts like Ethiopia where abortion is available for multiple indications. Efforts to improve knowledge of abortion legality and availability are needed, and should be locally tailored to address regional inequities.
Our study draws on data collected by Performance Monitoring for Action (PMA) Ethiopia. PMA Ethiopia is a 5-year project implemented collaboratively by Addis Ababa University (AAU) and Johns Hopkins University (JHU). PMA Ethiopia conducts nationally representative surveys measuring a range of key reproductive, maternal, and newborn health indicators. The study design and survey administration have been described in detail elsewhere [19]. Data come from a cross-sectional survey of women aged 15 to 49, collected from October through December 2019. A 2-stage cluster sampling design was used, and data were collected in each of 9 regional states and 2 administrative states. A total of 265 geographic enumeration areas (EAs) were selected by the Central Statistical Agency, and a cross-section of 35 households were selected randomly within each EA. Women were eligible to complete the survey if they were aged 15 to 49 years, lived within the selected EA boundaries, and slept in the selected household the night prior to the survey. Participants provided oral informed consent. Altogether, 8890 women from 9202 households participated in the female survey, with a response rate of 98.4%; our final analytic sample included 8837 women from whom we had complete data on the outcomes of interest (knowledge of abortion legality and availability), as well as current contraceptive use. The study received ethics approval from the JHU and AAU Institutional Review Boards. Our 3 dependent variables comprised (1) general awareness of abortion legality, (2) comprehensive knowledge of the national abortion law, and (3) knowledge of facility-based abortion service availability. We assessed awareness and knowledge of the abortion law using responses to 2 survey questions. First, we asked participants if they knew about a law regulating abortion in Ethiopia. Per the definition used by Assifi et al. in their systematic review on women’s knowledge of abortion laws, we consider this question to measure women’s general awareness of abortion legality [2]. Women who replied affirmatively were then asked to indicate under which circumstances it is legal to have an abortion in Ethiopia. The response options were read aloud to participants and included: (1) rape, when the (2) pregnancy poses a risk to the life of the woman and/or fetus (risk to the woman), (3) fetus has been diagnosed with an incurable disease or serious deformity (fetal impairment), (4) pregnant woman is incapacitated, or physically/mentally unfit to be a mother (maternal disability), and (5) under no circumstances. We used women’s responses to these items to measure comprehensive knowledge of the abortion law. We examined binary measures of knowledge of each specific legal ground, knowledge of any legal ground, and whether the respondent had comprehensive knowledge of the law (i.e., correctly identified that abortion is legal under the 4 circumstances indicated).1 Women who reported that abortion is not legal under any circumstances are not included as having comprehensive knowledge. Finally, a binary measure of knowledge of facility-based abortion service availability was assessed by asking women, “Do you know where a woman can access facility-based abortion services in the community where you live?” (Yes/No). These questions were asked to all women who completed the survey, regardless of age, marital status, or reproductive history. Our primary independent variables were region, comprising Ethiopia’s 11 regional and administrative states: Addis Ababa, Afar, Amhara, Benishangul-Gumuz (BG), Dire Dawa, Gambella, Harare, Oromiya, Southern Nations, Nationalities, and People’s Region (SNNP),2 Somali, and Tigray, and urban and/or rural residence. We first explored sample characteristics and calculated prevalence estimates by region and urban and/or rural residence, assessing differences in women’s awareness and knowledge of abortion legality and facility-based abortion availability. We then examined bivariate distributions of regional and urban/rural residence by women’s general awareness of the abortion law, knowledge of specific legal grounds, comprehensive knowledge of the abortion law and knowledge of facility-based service availability; statistical differences between regions and residence were assessed via one-way ANOVA tests and design-based F statistics, respectively. Finally, we examined the relationship between knowledge of legality and availability by assessing knowledge of facility-based abortion service availability among women who correctly reported at least 1 legal ground for abortion, and vice versa. We used design-based analyses to account for the complex sampling design, including accounting for clustering of women within EAs and applying survey weights to account for differential probability of selection, in order to generate nationally and regionally representative estimates. We did not further account for clustering of responses within households as all women in the household share the same probability of selection. As there was slightly fewer than 1 woman (0.98) per household among all sampled households, there was not significant evidence of clustering within households. We analyzed data using Stata 16.1 (Statcorp LP, College Station, TX).
N/A