Factors associated with unintended pregnancy in Ethiopia; further analysis of the 2016 Ethiopian demographic health survey data

We used the 2016 EDHS data for this analysis. The 2016 EDHS is a community based cross-sectional data collected from January 18 to June 25, 2016. The survey was designed to provide key indicators at national and regional level. The survey used a two-stage stratified random sampling technique. First, each region was stratified into rural and urban areas. Then, enumeration areas (EA) were selected with probability proportional to enumeration area size. After this, household listing in the selected EAs was done. In the second stage, a fixed number of 28 households from each cluster were selected and included using systematic random sampling technique. In both surveys, women aged 15–49 who were either permanent residents of the selected households or visitors who stayed in the household the night before the survey were interviewed. An interviewer administered questionnaire was used to collect data. The details of the methodology; sampling technique, data collection and data quality assurance are available from EDHS reports [24]. A total of 15,683 reproductive age women were included in the survey. From these, 1135 women were pregnant at the time of the survey and included in the analysis.

The outcome variable for this analysis was unintended pregnancy, which had two categories (yes or no). The 2016 EDHS questionnaire asked all reproductive age women involved in the survey whether they were pregnant or not at the time of the survey. If the woman responded she was pregnant, then she was asked if the pregnancy was wanted at the time, wanted later or not wanted at all. When the woman reported the pregnancy was wanted at the time of the survey, the outcome variable was considered “no”, meaning the pregnancy was intended. On the other hand, if the woman reported the pregnancy was wanted later or was not wanted at all, the outcome variable was considered unintended and coded “yes.”

The independent variables included in this analysis were socio-demographic (age, educational status, marital status, household wealth index, residence, religion, working status of the woman and region), reproductive health (parity, history of abortion, history of previous contraceptive use and knowledge of contraceptives), access to health service (distance to health facility to seek medical care) and woman’s autonomy related variables (can refuse sex, refusing sex is justified if she suspects (STI).

We used STATA software for this analysis. The data were weighted to adjust for over sampling or under-sampling and non-response. Descriptive statistics were calculated for all variables. Correlation between independent variables was checked before fitting the final regression model. Multivariable logistic regression analysis was done to identify factors associated with unintended pregnancy. When two independent variables were found to be correlated, one was dropped. In addition, complex survey analysis techniques were employed when computing odds ratios since DHS used a two-stage stratified sampling technique.

Five hundred fifty five (48.93%) respondents were in the age range of 25 to 34 years. Six hundred three (53.20%) respondents reported that they didn’t attend formal education. Two hundred fifty seven (22.67%) respondents were from households with the poorest wealth index. Majority of the respondents (96.59%) were married or living in union. Nine hundred seventy four (85.83%) respondents were rural residents. About three-fourth of these pregnant women were not working at the time of the survey (Table 1).

Four hundred fifty six (40.23%) respondents were multiparous. One thousand twenty five (90.33%) respondents had no history of abortion. Six hundred seventeen (54.37%) respondents had never used contraceptives. Six hundred thirty five (56.01%) respondents reported distance was a big problem to get medical care. Six hundred fifty two (57.51%) respondents reported that they would not refuse sex when the husband request. One thousand one hundred six (97.49%) respondents were knowledgeable about contraceptives (Table 2).

The level of unintended pregnancy among Ethiopian women was 336 (29.66%) with (95% CI, 25.33–34.39). Unintended pregnancy was high in Oromia region and low in Harari region.

Among women who have unintended pregnancy, 169 (50.36%) respondents were in the age groups of 25 to 34 years. One hundred and ninety (56.69%) respondents who had an unintended pregnancy had not attained formal education. Among women who have unintended pregnancy, three hundred eight (91.63%) respondents were living in rural area (Table 3).

On the multivariable analysis, marital status, region, perceived distance to seek medical care, parity and history of contraceptive use were found significantly associated with unintended pregnancy.

Women who were married had 66% lower odds of unintended pregnancy compared to women who were single, divorced and widowed AOR; 0.034; 95% CI: (0.01–0.14). Women who were living in developing regions were 84% less likely to have unintended pregnancy compared to women who were living in major regions AOR; 0.14, 95% CI: (0.07–0.27). Women who reported distance to health facility was not a big problem to get medical care were 41% AOR 0.59; 95% CI: (0.36–0.99) less likely to have unintended pregnancy compared to those who reported distance was a big problem to seek medical care (Table 4).

The odds of having unintended pregnancy among multiparous and grand multiparous women (women with five or more births) was AOR; 3.77, 95% CI: (1.71–8.33) and AOR; 6.72, 95% CI: (2.74–16.49) respectively times higher compared to nulliparous women. Similarly, women who ever had used contraceptives had AOR; 1.86; 95% CI: (1.06–3.26) times higher odds of having unintended pregnancy compared to those who never used (Table 4).