Safe abortion service utilization and associated factors among insecurely housed women who experienced abortion in southwest Ethiopia, 2021: A communitybased cross-sectional study

Background Insecurely housed women are more vulnerable to physical and mental health issues than the general population, making access to a safe abortion more difficult. Though Ethiopia has a penal code regarding safe abortion care, there has been a dearth of studies investigating the safe abortion care practice among those insecurely housed women. Thus, this study aimed at assessing the magnitude of safe abortion service uptake and its determinants among insecurely housed women who experienced abortion in southwest Ethiopia. Methods A community-based cross-sectional study was conducted in three towns in southwest Ethiopia from May 20-July 20, 2021. A total of 124 street-involved women were included in the study. They were selected by snowball sampling technique and data was collected through a face-to-face interview. The data were entered into Epi-data Version 3.1 and exported to SPSS 21 for analysis. A bivariable and multivariable logistic regression analyses were performed to determine the association of independent variables with the outcome variable. The level of significance was determined at a p-value <0.05. To determine whether the model is powerful enough in identifying any significant effects that do exist on the dependent variables, a power analysis was performed via a Post-hoc Statistical Power Calculator for Multiple Regressions. Results The magnitude of safe abortion service utilization among insecurely housed women was found to be 27.9% [95% CI: 20.1, 34.2]. Average daily income [AOR:3.83, 95% CI: 1.38, 10.60], knowledge of safe abortion services [AOR:3.94; 95% CI: 1.27,9.24], and affordability of the service [AOR: 3.27; 95% CI:1.87, 8.41] were identified as significant predictors of safe abortion service among insecurely housed women. Conclusion and recommendation The magnitude of safe abortion service utilization among insecurely housed women in the study area was low. The respective town health offices and health care providers at the facility level should strive to improve awareness about safe abortion service's legal framework, and its availability. In addition, a concerted effort is needed from local administrators, NGOs, and healthcare managers to engage those insecurely housed women in income-generating activities that allow them to access safe abortion and other reproductive and maternal health services. From May 20 to July 20, 2021, a community-based crossectional study was conducted in three purposively selected towns in southwest Ethiopia: Jimma, Bonga, and Mizan-Aman. Jimma town is the capital of the Jimma zone in the Oromia region of Ethiopia, which is located 345 km from Addis Ababa, the capital city of Ethiopia. The town has 17 kebeles (13 urban & 4 rural) and based on the 2007 national census, the estimated population of Jimma town in 2020/21 is 220,609. Bonga town, the capital of the Kaffa zone, is 105 kilometers and 465 kilometers away from Jimma town and Addis Ababa, respectively. The town has three administrative kebeles and a total population of 27,634 people, with 13624 men (49.3%) and 14010 women (50.7%). Mizan-Aman is the capital of the Bench-Sheko zone and is located 651 kilometers southwest of Addis Ababa. The total population of the town is estimated to be 48,934, with 24,956 women (51%) living in all five kebeles. All insecurely housed women with abortion experience living in three selected towns of southwest Ethiopia (Jimma, Bonga, and Mizan Aman) were the source population. The study populations were those selected insecurely housed women with abortion experience residing in the towns. All insecurely housed women with a previous history of abortion were eligible to take part in this study. Participants who were seriously ill during the data collection period were excluded from the study. The sample size for the study was determined using the rule of thumb of 10 participants per measurement variable [30–32]. The number of explanatory variables after an in-depth review of the literature was 11, and thus the minimum sample size required for this study was (10 x 11 = 110). After accounting for a 10% non-response rate, the final sample size was 124. A snowball technique was used to choose study participants. The data was collected from the first group of eligible insecurely housed women who experienced abortion and agreed to participate in the study. These women were then used as a reference for recruiting another respondent who could be included in the study. The data collection proceeded until a large number of women had been identified and the required sample size had been met. Data were collected using a pre-tested structured questionnaire that was prepared after reviewing relevant literature in trying to attain the study’s objectives [1, 2, 20, 24]. The tool had six sections: Background characteristics, obstetric characteristics (pregnancy and abortion experiences), individual factors (knowledge of safe abortion services), health system-related characteristics, and utilization of safe abortion services. Four BSc midwives collected data through face-to-face interviews under the supervision of three public health officers. Before data collection, the respondents were informed of the study’s purpose and asked if they were willing to take part in it. The questionnaire was translated from English to Amharic, and then English language instructors will translate it back into English. Cross-cultural and conceptual translations were prioritized during the translation process over terminological literality or linguistic equivalence. Furthermore, the questionnaire was pre-tested on 5% of the total sample size in Serbo town, outside the study area. The internal consistency between the items in the knowledge and practice assessment questions was evaluated, and the corresponding Cronbach’s alpha values were 0.83 and 0.87, respectively. The overall internal consistency composite score was 0.85, which indicates that 85% of the variance in the scores was found to be reliable and deemed to be good. Both data collectors and supervisors got a one-day intensive training on the objectives of the study, data collection techniques, and procedures. Finally, supervisors reviewed and checked questionnaires for completeness at the end of each data collection day, and appropriate feedback was given to data collectors the very next day. The privacy of study participants was maintained to encourage genuine participation and information sharing as sources for the study. The outcome variable was the use of safe abortion services. Those who have terminated their pregnancy in a health facility by a certified health care provider were considered to have had a safe abortion (YES = 1), while those who sought abortion outside of a health facility were considered to have had an unsafe abortion (NO = 0). Insecurely housed (street) women: Women who live or spend most of their time on the street and rely on it for their livelihood [33, 34]. On-the-street women: were those women who had no formal homes (insecurely housed) and sleep on streets, verandas, balconies, etc. at night [34, 35]. Off-the-street women: are those women who have houses to go to for sleep at night while making their lives on the street life [34, 35]. Knowledge of abortion service: Eight questions were used to assess respondents’ level of knowledge of abortion services, and those who scored at or above the mean were considered knowledgeable unless they were not [36]. The perceived waiting time for abortion services: is the average time from initial referral to procedure reception, and it is considered prolonged if it exceeds 2 hours and short if it is less than 2 hours. The data were entered into EpiData version 3.1 and exported to SPSS version 21 For further analysis. Inconsistencies, completeness, and outliers were checked using running frequency, cross-tabulations, and sorting. To check for the distribution of variables throughout the population, descriptive statistics such as frequency distribution, proportion, mean, and standard deviation were computed. A binary logistic regression analysis was used to find factors associated with safe abortion service utilization. A bivariable logistic regression was used to assess the relationship between each explanatory variable and the response variable. As a result, with a P-value <0.25, eight of the eleven variables showed an association and were entered into a multivariable logistic regression model. Finally, multivariable logistic regression was performed, and three statistically significant variables at p-value <0.05 were identified. The adjusted odds ratio (AOR) of each significant variable with the corresponding 95 percent confidence interval was used to report the strength and direction of the association. To determine whether the model is powerful enough to identify any significant effects that do exist on the dependent variables, a Post-hoc power analysis was performed by using a Post-hoc Statistical Power Calculator for Multiple Regression. The following parameters were considered during the calculation: the number of predictors in the final model = 3, observed R2 = 0.28, probability level = 0.05, and a sample size = 122. Finally, the observed statistical power was found to be 0,9998, which is much greater than the minimum requirement of 0.8. So, enough statistical power existed to detect a significant effect. To assess model fitness, the effect size of the final model was estimated by correlating the predictive probabilities of each case (explanatory variables) with the outcome of interest. By doing so, the correlation value(the predicted probability) was 0.575, possibly an indicator of a good model fit. Furthermore, model fitness was assessed using the Nagelkerke R Square and the Omnibus test, the results of which were 0.466 and 0.001, respectively, indicating a good model fit. Multicollinearity between independent variables was also checked by estimating the variance inflation factor and no multicollinearity was detected (VIF<10). Ethical clearance was obtained from Jimma University Health Institute’s Human Research Ethics Committee. A letter of cooperation was also obtained from the Epidemiology department for the health offices of three selected towns. A letter of permission was obtained from each health office to proceed with the study. The participants were informed about the purpose of the study, and their right to refuse participation and discontinue the interview. Written Informed consent was obtained from each participant before the interview. By avoiding any identifiers of the study participants, the information obtained from them was kept confidential throughout the study. Personal protective equipment was provided for data collectors because the study takes place amid a global pandemic, COVID-19.