Utilization of institutional delivery service and associated factors in Bench Maji zone, Southwest Ethiopia: community based, cross sectional study

Background At the end of Millennium development goals, Ethiopia was included among 10 countries which constitutes about 59% of maternal deaths due to complications of pregnancy and/or childbirth every year globally. Institutional delivery, which is believed to contribute in reduction of maternal mortality is still low. Hence this study was conducted in order to assess utilization of institutional delivery and related factors in Bench Maji zone, Southwest Ethiopia. Methods Cross sectional study was employed from September 1st – 30th, 2015 in Bench Maji Zone, Southwest Ethiopia where 765 mothers who deliver 2 years preceding the study provided data for this research. Data were collected by enumerators who were trained. In addition to descriptive statistics, binary and multivariate logistic regression analyses were performed. Statistical significance was considered at a p-value < 0.05. Strength of association was also assessed using odds ratios with a 95% confidence intervals. Results About 800 mothers were approached but 765 of them who gave birth 2 years preceding the survey participated and gave consent to the data included in the analysis. About 78.30% delivered their last child in health institution while rest gave birth at home. Factors such as maternal age, religion, occupation, availability of information source as TV/Radio, income quartile, residence, knowledge of problems during labor and antenatal follow up had association with institutional delivery which was significant. Conclusion In Bench Maji Zone institutional delivery was shown to be comparatively good compared to other studies in the region and in Ethiopia in general even though it is below the health sector transformation plan of Ethiopia which aimed to increase deliveries attended by skilled health personnel to 95%. Empowering women, increasing awareness about institutional delivery and proper scaling up of antenatal care services which is an entry point for institutional delivery are recommended. Electronic supplementary material The online version of this article (doi:10.1186/s12913-017-2057-y) contains supplementary material, which is available to authorized users.


Background
Each year, still at the end of Millennium development goals (MDG), hundreds of thousands of women are dying due to complications of pregnancy and/or childbirth. In 2015, about 99% of the estimated global maternal deaths occurred in developing with Sub -Saharan Africa alone accounting for roughly 66% of deaths [1].
The first hospital based maternal mortality ratio (MMR) estimation in Ethiopia was done in Addis Ababa in 1980 [2]. However till today, Ethiopia is still included among 10 countries which hosts nearly 59% of maternal deaths ranking 4th next to Nigeria, India and Democratic Republic of the Congo with 11,000 (3.6%) maternal deaths [1].
The proportion of maternal death due to different causes changes from year to year. In Ethiopia the main causes of maternal deaths include infection, hemorrhage, obstructed labor, abortion and hypertension in pregnancy [3].
Traditional birth attendants, what ever trained or untrained, are not in a position to predict or get by with solemn complications [4]. Hence; increasing the number of mothers delivering at health facilities with the help of trained professional, which is the main goal of safe motherhood and child survival movements [5,6] is essential for the betterment of both child and maternal health.
The percentage of institutional deliveries continues to be low in Ethiopia. Nevertheless there has been significant advancement in the past 15 years in which the number of deliveries at health facility is higher, from 5% in 2000 to 10% births reported in 2011 [7].
Different studies in Ethiopia showed socio-demographic, obstetric, and health services related factors as determinants for institutional delivery service utilization. Among socio demography related factors residence, access to radio (media exposure), maternal education, wealth, partner and family involvement in decision making on place of delivery, age during first marriage, ethnicity and occupation [9][10][11][12][13][14]16] were shown to be significantly associated with delivery at health institution. Regarding obstetric factors parity, experience of a pregnancy and/or delivery related problems, past history of intra partum complication or prolonged labor [9,10,12,14], were reported to be associated. Health service related factors as; time it takes to the health facility, knowledge of the number of recommended ANC visits, good perception about quality of health services given to mother, gestational age during first ANC visit, antenatal care follow up during pregnancy, access to transport, being birth prepared and ready for its complication, perception of distance to health institutions [9][10][11][12][13][14]16] were also seen to be associated institutional delivery.
Working towards Sustainable Development Goal (SDG) number 3.1 which aimed to cut down the global MMR below 70 per100, 000 live births and stopping maternal mortality which is preventable by 2030 requires magnifying efforts and advancements catalyzed by MDG 5. Institutional delivery which is believed to contribute to the success of this plan was about 14.9% in Southern Nations Nationalities and Peoples Regional State (SNNPRs) [7] and 38.1% in the current study area (Bench Maji Zone) [17] which is still low.
This study was carried to assess utilization of institutional delivery service and associated factors in Bench Maji zone, Southwest Ethiopia. The finding of this study could help to understand institutional delivery and associated factors and reveals areas that need further study.
Results may give useful insights to health care providers, Health Extension Workers (HEWs) and promoters of community health to realize the situation in the survey area so as to give priority and directing their efforts. Moreover, since there is scarcity of similar study in the survey area, the present study can be used as a bench mark for future research. So that the present study was conducted to assess institutional delivery service utilization and associated factors in Bench Maji zone, Southwest Ethiopia.

Study aim, design, area and period
Community based cross sectional study was taken place in September, 2015 in Bench Maji zone, Southwest Ethiopia to measure institutional delivery service utilization and related factors. Bench Maji zone is one of the 16 zones in SNNPRs located 561 km away from Addis Ababa, the center of Ethiopia, in Southwest direction with an estimated population of 829,493 out of which 418,213 are women, 129,500 are children under five and 26,462 are under 1 year. The expected number of households in the zone is about169,284 and the primary health service coverage of the zone is 92.6% [18] covering a total catchment area of 19965.8 Km2 with majority 86% (1,061,120) of the population living in the rural areas [19]. The zone includes a city administration (Mizan Teferi), 10 Woredas (districts), 246 kebeles (smallest administrative units) (229 rural and 17 urban). There are two Hospitals (Bachuma and Maji Primary) one under construction and one functional (Aman General) Hospital. The zone has about 40 functional health centers and five under construction (Koka, Aroge Birhan, Gabisa, Zozo and Kuju). Additionally, there are 182 functional health posts and 9 under construction, one University and one Health sciences college. The zone has 12 physicians' and 511 of health professionals of different ranks and 476 health extension workers [19].

Sample Size determination and sampling procedures
The study participants were selected from child bearing age women that gave birth 2 years preceding the survey and regular residents of the selected Woreda (districts) of the zone. A sample size of 800 was calculated using the Statistical Package OpenEpi, Version 2, open source calculator. Owing to use of multistage cluster sampling, prevalence of women giving birth in the health institutions 61.6% (14) taken from a study area which shares the same population characteristics and provides the largest sample size, error margin of 5%, a 95% confidence level and 10% contingency for non response. Multi stage sampling technique was implemented to identify the study participants. From 10 districts and a city administration, two districts and one city administration were randomly selected. Then three Kebeles from each selected district and city administration were selected randomly using lottery method.
Households which host mothers who delivered 2 years preceding the survey were recorded in a census made by resident data collectors and the allocated sample for selected Kebeles was calculated using probability sampling proportional to the number of households in the Kebele. A total of 800 samples were selected using systematic sampling from the selected nine kebeles. For households with more than one eligible respondent, face-to-face interview was performed by choosing a respondent by lottery method. Two revisits were made when eligible mothers were not available during the visit. Eight female data collectors who had a minimum of Diploma were deployed after 2 day training on the objectives and methods of data collection. The data collection was closely monitored by supervisors and the principal investigator.

Data processing and analysis
The completed questionnaire was entered to a computer using EpiData version 3.0 and exported to SPSS version 20.0 for analysis. To check for consistency of the data entry frequencies were calculated and 5% of the questionnaire was selected randomly and re entered. Frequency, percentage, chi-square with p-value was used for description of the study population in relation to relevant variables. Analysis was performed employing a two step logistic regressions [bivarate and multivariable] since the dependent variable is dichotomous to look for the effect of the independent variables on the dependent variable by controlling confounders. Statistical significance was evaluated at 95% levels of confidence and p value less than 0.05 was considered significant.

Variables in the study
Data were collected on mothers' age, ethnicity, religion, maternal and husbands level of education and occupation, marital status, average monthly income, availability of communication media such as TV/radio and residence. Obstetric history and decision making related questions like age at first pregnancy, number of deliveries in the last 2 years, previous pregnancy outcomes, husbands desire to have more children and by whom decision was made to decide on the number of children, knowledge about institutional delivery, problems that can occur during pregnancy and child birth and problems that occur during previous pregnancy were also included.
The outcome variable focused on institutional delivery services utilization, by identifying identified mothers who gave birth with the assistance of health professionals trained on skills which are essential to carry off normal delivery, diagnose and refer obstetric complications.
Face-to-face interviews were made to collect data with structured and pre-tested instrument during house to house survey (Additional file 1). Most of questions were prepared in close ended forms including the list of alternatives from different studies including the DHS. An alternative "Other" was used to write responses for each closed ended question. Questions which include age, household income, number of deliveries, number of ANC visits and age at first pregnancy were prepared in open ended form and the responses were recorded and later coded during data analysis. For both open and closed ended questions the list of alternatives was not red to respondents and only clarification of unclear questions to the respondents were made.

Socio-demographic profile of respondents
A total of 800 respondents were approached and 765 mothers who deliver 2 years preceding the survey were included in the analysis which yields a non response rate of 4.37%. About 400 (52.2%) of the women were from urban (Zonal Town), 166 (21.7%) semi urban (Woreda Town) and the rest 200 (26.1%) were rural residents. The mean age of respondents was 25.47 ± 4.4 with the Mean average monthly income of 1146.19 ± 1328.1 Ethiopian Birr (ETB) while the average family size was 4 ± 1.5 (ranging from 2 to 11). Majority of mothers were from Bench ethnic group (24.6%) followed by Amhara (20.3%). Majority of mothers were Protestant (48.2%) followed by Orthodox Christian (36.7%). Concerning literacy majority of mothers cannot read and write (40%) with majority of their husband attended primary school (42.5%), about 73.2% of mothers were housewives with 36.3% of their husbands were farmers followed by merchants (34.3%) and majority of families have television (TV)/radio (63.1%) in their house (Table 1).

Obstetric history of respondents
Three hundred seventy six (49.2%) of mothers gave birth to their first child between the age of 20-24 with the mean age during first child birth of 20.2 ± 2.75. Teenage pregnancy (15)(16)(17)(18)(19) which have association with high morbidity and mortality for both mother and child was also significantly higher among participants (43.7%). Among the participants 3.7% have delivered children less than a year apart. Concerning the outcomes of pregnancy, 10 (1.31%) of mothers never gave live birth, 4.9 and 4.0% of mothers also have a history of still birth and any forms of abortion. Majority of respondents (67.1%) jointly decided the number of children they have ( Table 2).

Knowledge of danger signs of pregnancy
Three hundred eighty eight (50.7%) of mothers replied that they can identify danger signs that may indicates problems that occur during pregnancy, of whom about 81.19% mentioned vaginal bleeding and only 5.93% mentioned drowsiness as a danger sign of pregnancy Among all mothers, 14.9% replied they had a problem during last pregnancy where hand/leg swelling occurred in about 22.81% of mothers (Fig. 1).

Institutional delivery related knowledge and service utilization
Majority of respondents (95.6%) replied that they have heard about institutional delivery. Health institutions and health care providers were sources of information for majority (68.9%) of the respondents. About 81.0% replied that they know a health problem that can occur during childbirth in which severe bleeding was mentioned by majority (71.8%) of respondents.
Among mothers who were asked about their place of delivery, majority (78.30%) delivered their last child in health institution and about 21.70% delivered at home. The main reason mentioned for institutional delivery was that "insisted by health professionals and others" to  Who joined secondary education, technical schools, colleges and university deliver in health institution (82.1%). and the widely mentioned reason for home delivery was "the labor was going well" (62.9%) and some of mothers were also forced to deliver at home since the labor started suddenly. Concerning the final decision on place of delivery, majority (61.70%) replied by joint discussion with husband and about one third (34.20%) replied they decided on their own (Table 3).

Factors associated with institutional delivery service utilization
Different factors have been identified to have a positive or negative association with utilization of institutional delivery service. On bivariate analysis of the selected factors with institutional delivery; age of the mother, religion, maternal education, maternal occupation, availability of TV/radio, income quartiles, residence, knowledge of problems that may occur during labor, ANC follow up were found to be significantly associated. After adjustment of confounding factors using multivariable logistic regression analysis maternal education lost its association with institutional delivery (Table 4).
This study found that mothers in the age groups of 25-34 and 35-44 were less likely to deliver in health institution (AOR = 0.4 and 0.3) respectively in comparison with those mothers in the age group of 15-24. Mothers who are followers of Orthodox Christianity were about two times more likely to deliver in health institution (AOR = 2.2) in comparison with Protestants while Merchants were about 12 times more likely to deliver in health institution (AOR =12.3) in comparison with House wives.
Availability of information source as TV/Radio had shown a significant association in which those mothers who replied they have TV/radio at home were more likely to deliver in health institution (AOR = 2.6) in comparison with those mothers who replied they don't have TV/Radio at home. Concerning their income, in comparison with those mother's in the lowest income quartile, those in the middle (AOR = 0.5) and upper income quartile (AOR = 0.2) were less likely to deliver in health institution. In comparison with mothers residing in urban areas, mothers from semi urban (AOR = 0.1) and rural areas (AOR = 0.03) were less likely to deliver in health institution.  Knowledge of problems during labor also had a significant association with place of delivery. Mothers who said they know problems which may occur during labor were 11 times more likely to deliver in health institution (AOR = 11.0) in comparison to those mothers who replied they don't know problems during labor and in comparison to mothers who had no ANC follow up at all, those mothers who had ANC visits (i.e. less than the recommended four visit or within recommended four, and above visits) were more likely to deliver in health institution (AOR =7.6 and 26.2) respectively (Table 4).

Discussion
This study assessed utilization of institutional delivery service and associated factors among mothers who delivered 2 years preceding to the survey in Bench Maji zone. Among respondents, majority (78.30%) delivered their last child in health facility. This result was higher than studies done in different study areas including the previous study which was conducted in the present study area (38.1%) [17], the national DHS study in which 16% of deliveries in Ethiopia and 14.9% of deliveries in SNNPR were at a health facility [7] and studies done in Sekela district (12.1%), North Gondar (13.5%), Bako (34.21%), Dodota district (18.2%) and Goba Woreda (47%) where mothers delivered their last child in health institutions [8,[10][11][12][13].
This registered high service utilization in this study area might be attributed to the wider introduction and increased number of HEWs in Ethiopia in which more than 38,000 HEWs have been trained and deployed in agrarian, pastoralist and urban areas in more than 16,000 Kebeles [20] which improves the utilization of health services by linking community and health facilities particularly health centers.
Additionally this study was conducted after the government made delivery services to be free at all levels of health institutions and implementation of Health Development Army (HDA) has launched in 2011 with development being made in the organization and network formation over the past 3 years. According to Ministry of Health's (MOH) annual report, a total of 442,773 HDA groups with 2,289,741 one-to five networks were formed in 2014 [20].
The other possible reason for increased service utilization may be attributed to improved Health infrastructure as the government provided special concern for capacity building and included in Health Sector Development Plan (HSDP) IV strategic objective capacity building to improve health infrastructure [21] and in the present study area during the data collection period there was one Hospital, five Health centers and nine health posts under construction [18]. This was also indicated on a national service provision assessment in which almost all (99%) of government managed facilities offer normal delivery services [22].
High level political commitment and the involvement of Non Governmental Organizations and provision of ambulances on top of expansion of health facilities and  In this study maternal age, religion, maternal occupation, availability of information source as TV/radio, income, residence, knowledge of problems during labor and ANC follow had statistically significant associations independently with institutional delivery which is similar with other studies such as South West Shoa in which institutional delivery was positively associated with urban residence, wealth, knowledge of the recommended number of ANC visits and birth preparedness [9]. Results also supported as a study done in North Gondar which found that mothers who have no access to radio were less likely to deliver at a health facility than women with access to radio, rural women were less likely to use the services than urban women, mother with lower incomes (<100 ETB) were less likely to deliver at health institution than those having incomes of 500 Birr and above [10]. Results from this study were also similar with a study done in Bako, Oromia where being a house wife, private employee, ANC visit were associated with institutional delivery independently [11]. Similarly, this study found similar results with a study conducted in Dodota which found that mothers residing in urban set up were more likely to deliver at health institution in comparison with rural and semi urban women [12]. Similar with this study, age of the mother was shown to be associated with institutional delivery in studies conducted in North Gondar [10] and Sekela district [8] which might possibly younger women are educated. Older mother might also consider delivering at home by considering it is not risky since they might have delivered at home previously.

Conclusions
Utilization of institutional delivery services in Bench Maji Zone was shown to be comparatively good (78.3%) compared to other studies in the region and in Ethiopia even though it is below the HSTP of Ethiopia to increase deliveries attended by skilled health personnel to 95% [20].
Health professionals' contribution in awareness creation and increased service utilization was also acknowledgeable. To save maternal, newborn and child lives improving the quality of the existing services and enhancing access to withdraw the inequalities in Maternal and Child Health service is of major importance.
Maternal occupation and income have shown relationship with outcome variables. Hence empowering women in the aspect of economy, promoting maternal independence within the family, empowering them to earn and control over the utilization of services and decide on their own health. Relationships between presence of sources of information like TV/radio implies provision of adequate IEC and increasing awareness about institutional delivery through mass media and religious leaders and health professional by giving due emphasis to rural mothers since may have limited access to information than urban mothers.
Knowledge of problems during labor and ANC follow up has shown significant association with institutional delivery. Therefore, proper scaling up of ANC service; which is an entry point for other MCH services quality and follow up with special attention on counseling on problems that may occur during labor and safe institutional delivery. To enhance quality and avoid the inequalities of services the efforts by MCH directorate of the Federal MOH should be expanded in a sustainable way. Regional health bureaus should increase their partnership with higher educational institutions, ministry offices, professional associations and local and international partners and continue awarness creation activities for MCH services