Beyond knowledge acquisition: factors inuencing family planning utilization among women in conservative communities in Rural Burundi.

With a fertility rate of 5.4 children per woman, Burundi has been ranked as seventh highest country with the highest fertility rate in the world. Family planning is known to allow couples to achieve the desired family size, appropriate space birth and the limitation of pregnancies. Also, family planning can contribute to mitigating some health issues such as unintended pregnancies and abortions all of which, are often associated with multi-parity. In conservative community in rural Burundi, knowledge on family planning is high and such services are free yet utilisation is low. Employing a mixed methods, this study rst quanties contraceptive prevalence and second, explore the contextual multilevel factors associated with low family planning utilisation among married women. Methods The eighteen Quantitative with and data and


Abstract Background
With a fertility rate of 5.4 children per woman, Burundi has been ranked as seventh highest country with the highest fertility rate in the world. Family planning is known to allow couples to achieve the desired family size, appropriate space birth and the limitation of pregnancies. Also, family planning can contribute to mitigating some health issues such as unintended pregnancies and abortions all of which, are often associated with multi-parity. In conservative community in rural Burundi, knowledge on family planning is high and such services are free yet utilisation is low. Employing a mixed methods, this study rst quanti es contraceptive prevalence and second, explore the contextual multilevel factors associated with low family planning utilisation among married women.

Methods
An explanatory sequential mixed study was conducted. Five hundred and thirty women in union were interviewed using structured and pre-tested questionnaire. Next, 11 focus group discussions were held with community members composed of married men and women, administrative and religious leaders (n=132). The study was conducted in eighteen collines of two health districts of Vyanda and Rumonge in Bururi and Rumonge provinces in Burundi. Quantitative data was analysed with SPSS and qualitative data was coded and deductive thematic methods were applied to nd themes and codes.

Results
The overall contraceptive prevalence was 22.6%. From logistic modelling analysis, it was found that women aged 25 to 29, those completed secondary school and having four or less children was signi cantly associated with use of family planning. Among factors why family planning was unused included experience with side effects and costs associated with its management in the health system. Religious conceptualisation and ancestral negative beliefs of family planning had also shaped how people perceived it. Furthermore, at the household level, gender imbalances between spouses had resulted in break in communication, also serving as a factor for non-use of family planning.

Conclusion
Given that use of family planning is rooted in negative beliefs emanating mainly from religious and cultural practices, engaging local religious leaders and community actors may trigger positive behaviours change needed to increase its use.

Plain English Summary
In the context of rural Burundi, community members agree that large family are di cult to sustain but contraception coverage remains consistently low. This study explored the factors behind this low utilisation of family planning in two health districts located in Vyanda and Rumonge communes of Burundi. The ndings suggest that fear of side effects is the main reason of family planning nonutilization or discontinuation. The culture and religious beliefs also favour large family size and among men, this is conceived as a sign of wealth, power and respect. Lack of spousal communication and unequal gender relations in household also impedes women from contributing decisions on family planning. Therefore, the onus lies on men, whom, have limited understanding of how family planning works to make nal decision on initiation.
In improving coverage of family planning in these communities, rst, the health system, which serves a primary contact, should be capacitated to provide quality, timely and people-driven family planning services to those who need it. At the community level, use of community health workers to deliver family planning services could signi cantly increase uptake. Men and religious leaders' involvement in promoting family planning use can contribute to reducing the impact of cultural and religious barriers, increasing uptake to desired levels.

Background
Burundi, a small, landlocked country in East Africa is a home to 11,175,378 individuals [1]. The population density is estimated at 435 people per square kilometre thereby, making Burundi second most populated African nation. [1] Despite the scarcity of cultivable land, 87% of the rural population depend solely on subsistence farming [2]. With a fertility rate of 5.4 children to a woman, Burundi has been classi ed as among the top ten countries with the highest fertility, placing just seventh after Niger, Somalia, the Democratic Republic of the Congo, Mali, Chad and Angola [24]. High fertility rate has been identi ed as a major contributor to poor health outcomes such as increased maternal and child morbidity and mortality. Consistently, previous studies have suggested that grand multiparity (parity >=5) is associated with adverse pregnancy outcomes such as caesarean delivery, foetal macrosomia, diabetes mellitus and pregnancy induced hypertension. [4]. Additionally, Hendrik et.al (2014) found that short birth intervals negatively affect perinatal, neonatal and child health by increasing the incidence of preterm birth, low birth weight and perinatal death. [5] On the other hand, contraceptive use prevents pregnancies and associated risks of miscarriage, stillbirth and postpartum haemorrhage among others [6]. Research suggests that preventing births in mothers with ve or more children could reduce maternal deaths by 58% and family planning could prevent around 272,000 maternal deaths in the world every year [7]. Recent evidence suggests that family size could also be a determinant to child nutritional status, morbidity and mortality [8]. Moreover, socially, contraceptive use among women is an indicator for autonomy which is also accomplishment of fundamental human rights of decisions on how to use their bodies [9] and in ful lment of sustainable development goals 3 [10].
Despite that modern contraceptive methods have shown to be effective in spacing births and avoiding unintended pregnancies, its low uptake is still a concern in many parts of the developing world particularly in Sub-Saharan Africa. 21% percent of the 214 million women of childbearing age in developing countries who would like to avoid pregnancy but are not using any modern contraceptive method reside in Sub-Saharan Africa [7]. According to the DHS, only 23% of women in Burundi living with a partner and aged between 15-49 years use modern contraceptive methods of which 30% of these have an unmet need of family planning [3]. Although 97% have comprehensive knowledge on family planning, 40% of women using contraceptives in the last ve years discontinued with 33% of women citing side effects as reasons for discontinuation [3].
This nding is consistent with other studies conducted in Sub-Saharan Africa and South Asia. A study conducted in Pakistan highlights that more than 80% of married women reported that the experience of side effects is the main reason of discontinuation of the last contraceptive method used [11].Another study by Bekele et al in South East Ethiopia has suggested that fear of side effects accounts for 48% of discontinued contraceptive method use among married women [12].
While Burundi is one of the countries with the least contraceptive prevalence, little research has been conducted from community's perspective to understand the factors driving utilisation especially in communities (collines) with conservative traditional values. Little is known about how the sociocultural space facilitates or hinders the utilisation and continuation of family planning in the context of Burundi.
This study aims to investigate the factors affecting FP utilisation in conservative communities in rural Burundi using an explanatory sequential mixed methods study. Through acquisition of quantitative and qualitative information, we sought to bring an in-depth comprehension of this complexity in this setting.

Study Design
An explanatory sequential mixed design study was carried out from 1 st May to 28 th June 2019, commencing with quantitative information on a mass scale then using focus group discussions (qualitative methods), we explored and described patterns/relationships that emerged from the quantitative strand. To allow comparison with national gures, we adopted and adapted quantitative questionnaires from the DHS to understand contraceptive prevalence disaggregated by background characteristics. Questionnaires were deployed via an electronic system and administered by appointed enumerators and teams. Due procedures to ensure respondents' rights and privacy were respected.
We learned from family planning records from Kigutu Health centre which serve the community in the eighteen collines that women represented more than ninety per cent of the people who came for family planning service during the previous year. Having that background knowledge, this study has focused on married women during the quantitative phase and questionnaire was administered to a sample of this target population.
Data from quantitative phase was collected and analysed and those interesting relationships were further explored using focus-group discussions. These groups comprised of married community members, opinion, and religious leaders. Those community representatives in the focus group discussion were selected because they qualify for family planning utilization or were in the position to in uence family planning uptake. They were selected in all 18 collines, however in case of collines with similar characteristics, representatives of two or three collines were invited to join one discussion group. This qualitative phase used an open-ended questionnaire that assessed different topics including family planning. Although questions had been preconceived, it was general, and the direction of discussions were driven by study participants. An experienced and trained moderator was appointed and ensured that all respondents were given equal opportunity to speak and their views were respected.

Outcome Variable
The main outcome variable for investigation was current use of contraceptives which a binary variable (Yes/No). The de nition of this variable was in line with that of the Burundian Demographic Health Survey [3]. To acquire information for this indicator, women in union were asked if they or their partners were doing something or using any method to avoid or delay pregnancy [13]. The population base used as a denominator was all women in union at the time of interview.

Sampling Strategy
The quantitative phase included a two-stage sampling strategy which comprised of cluster and systematic sampling. During the rst stage, all collines, their population were considered as a possible sampling frame. A probability proportion to size (PPS) was applied to select a desired cluster size of 30. The calculated sample size was divided by this number to determine the sampling interval. Enumerators upon arrival at every colline, received household list from the colline administrator. The sampling interval applied to derive the interval of households to be selected systematically. Given the study objectives, the following criteria were applied to ensure enumerators were interviewing the right households: Inclusion Criteria 1. Household with women aged 15 and 49 also classi ed as those in reproductive age 2. Head of household and caregiver in the right state of mind to actively respond to questions 3. Households whose head signed off the consent form Exclusion Criteria 1. Households that did not meet the stated inclusion criteria 2. Households whose women of reproductive age were currently pregnant Sample Size Calculation The sample size was calculated based on predetermined methodology of the Demographic Health Survey. Using this rigorous methodology, a sample size of 952 households were selected and distributed equally among the clusters. Details of the sample size calculation has been published elsewhere [3].
The qualitative phase employed 11 focus groups with an average size of 12. Members of these groups were purposively selected to represent a mix of community representatives and members.

Data Analysis
Quantitative analysis was initially cleaned via quality check for completeness and consistency.
The analysis in this paper is based on knowledge of contraceptive methods and current use of contraceptive methods. Women who never had child or wanted another child within 2 years were excluded in the analysis regarding current use of contraceptive methods. These indicators were analysed quantitatively using IBM software-SPSS Statistic version 20 [13]. Chi-squares were used to assess relationships between outcomes and exposure variables. A logistic regression controlling for cofounders such as place of residence, wealth quintile and religion was used to predict the adjusted odds of a woman using contraceptives. Signi cance of statistical relationships was considered at 95%CI (two tailed).
For the qualitative analysis, focus group discussions were recorded, transcribed and translated from Kirundi to French then to English. In ensuring consistency in translation, all transcripts were reviewed by two native Kirundi and French speakers with an advanced uency in English. When there was a disagreement on translated word, a third advice was sought from a third person. After this, deductive mechanisms consisting of desired number of children, facilitators and barriers of family planning were employed to detect emerging themes and codes. The generated codes were applied to all transcripts for quotes which were in line to themes.

Ethical Considerations
A signed letter approval was granted for this study by the local health authorities in the Bururi and Rumonge province which is a representative body of the National Ethics committee at the province level. Study participants signed a consent form ahead of interview and careful steps were taken to ensure respondent's rights and privacy were respected at all times.

Socio Demographics Characteristics
A total of 530 women in union at the time of data collection were interviewed from 930 households, accounting for a response rate of 95.4%. The mean age of the respondents was 30.76(± 6.872 SD) with a minimum age of 18 years (Table 1). More than half percent of women (51.7%) were between the age of 25 and 34 years. The mean age at the rst pregnancy was 20.17(±3.53SD) years. More than a quarter (26.6%) of women in union has not attended formal education while only 12.3% of those who attended formal education had completed secondary education. The average number of children a woman had was 4.3(± 2.4 SD). 216 women (42%) had ve or more children. Additionally, 37.7% of women spaced the last two children less than 24 months. 14% of the participants reported that they had ever lost a child less than 5 years old.

Family Planning
In this study, 94.3% of women, reported to know at least one type of contraception with 94.2% knowing at least a source of acquiring contraceptives. Only 22.6% of women who desired to space their next birth more than 2 years were using a contraception method at the time of interview. The most prevalent modern contraceptive method was the injectable accounting for 40%, followed by implants used by 24.6% of women. The male preservative, Intra Uterine Device (IUD), pills and sterilization represented 10.8%, 6.2%, 3.1% and 1.5% respectively. 13.8% of women who adhered to any form of contraception reported to be using natural contraceptive methods. Among women not on any form of contraceptives, the following reasons were cited: side effects (51%), absence of menses since delivery which is considered as postpartum or postpartum amenorrhea in case it exceeds six weeks (18.8%), religious beliefs (12.9%), partner opposition (8.4%), partner absenteeism (6.4%) and lack of awareness about family planning services (2.5%). Among women on modern contraception, 68% reported that they were not informed about eventual side effects at the time they started utilization. There was a statistically signi cant negative association between last two children birth intervals and family planning utilisation (p-value <0.001). Those who had longer intervals between the two last births were less likely to adopt any family planning method as spacing had been considered as a natural family planning method.

Findings from Focus-Group Discussions
The participants of the focus group discussion were women of reproductive age (between 15 to 49 years old), men whose spouses' age was within 15 to 49 years, administrative and religious leaders. Each of the 18 collines was represented by an even number of women and men selected from the above group of people.
The results of the quantitative phase showed low uptake of the family planning methods. The qualitative phase was undertaken to understand the reasons justifying the quantitative results. Large family size emerged as a key theme which resonated with most participants irrespective of the colline of residence.
Considering the desired number of children expressed by each group, the discussion was directed towards the different barriers that hamper the achievement of the ideal family size as many families were having more children than they wished to have.

Desired number of children
Both men and women agreed that it was very hard to sustain a large family in the present socioeconomic situation. Most participants suggested that at most four children was ideal to achieve a decent standard of living. From the ndings, fear of contraceptives' side effects was reported as the main reason for underutilisation and discontinuation modern contraceptive methods. Rumours regarding the side effects of family planning which included bleeding, cancer and infertility propagated by other community members had largely contributed to this communal fear.
"I know one lady who started using modern contraceptive after her second pregnancy. Later on she abandoned to get the third child but failed to conceive." (FGD Woman, Migera colline) This climate of fear had been used as an avenue exploited by religious leaders and other anti-family planning people to make contraceptives unpopular: A male participant said, "Church leaders often instruct the faithful that the use of modern contraception is the cause of the increase of cancer cases that occur nowadays." (FGD man, Gashasha colline) In addition, some participants had shared their personal experience with side effects and how this was conceived by their spouses and the community as a whole. Personal experiences had deeply shaped continuation of family planning and although there was wide community consensus on the adverse effects of family planning, side effects created preconception which encouraged family planning discontinuation as soon as side effects emerged: "I have used injectable and I experienced continuous bleeding. I went back to the hospital and obtained some medicines and I have since abandoned the modern contraceptives method." (FGD woman, Cabara colline) In addition, the issue of side effect management at the health system level was raised during the discussions.
"There should be quali ed personnel to deliver quality health care to follow those side effects cases closely and inform the people who are coming for family planning services about the side effects that may occur. If the health practitioner cannot give medicines to resolve the side effects issue, the patient will give up the use of family planning methods." (FGD man, Mushishi colline) Further, participants were concerned about increased fees to treatment of side effects despite receiving the family planning for free. This concern had metamorphosed into a fear which served as a disincentive for not only initiating family planning but also, continuation for those that subscribed: "I know a woman who have used an IUD but experienced side effects. This woman had to pay a lot of money for her treatment and then discouraged other women to consider any modern contraceptive methods." (FGD man, Mushishi colline)

Religious beliefs
In conservative communities, religion forms an integral section of the life and in the case of deciding the use and continuation of family planning, religion is a major determinant. From the focused group discussions, we learned that natural family planning which consists in identifying the signs and symptoms of fertility during a menstrual cycle and practising sexual abstinence during the fertile period to avoid pregnancy was the only method that most of the religions recommended. This in uence from religious leaders and conceptualization of modern family planning as a sin served as a deterrent for most women: "Our community health workers always sensitize about family planning but their teachings con ict with church teachings which say that modern family planning is a sin of killing. When it is known that a church follower has adhered to one of those methods, she will be suspended from the church services. That is the reason many individuals have given up the use of modern contraceptive methods." (FGD man, Kabwayi colline)

Cultural beliefs
Culture as expected, was found as a determinant of family planning and this emerged as a theme from the focus group discussions. For some participants they cannot limit the family size when they have only girls on the other hand, they prefer many children because some of their offspring may eventually die and in that case, they hope that at least some will survive and support the family. Other women perceived increased family size of a man as a security to their marriage: "In Burundian culture we are afraid of having few children. For instance, if we go for vasectomy and death takes all the children, it will not be possible to reproduce again." (FGD man, Kanenge colline) "In our community, women with many children think that their husband will not seek extra marital children and therefore do not adhere to family planning methods." (FGD woman, Kanenge colline)

Spousal communication gap
Both men and women agreed that they do not openly discuss the optimal number of children and how family resources could support child upbringing.
"I could say that there is a lack of communication between husband and wife. Otherwise, if they were communicating effectively, they could convince each other and reach a common understanding on how to achieve family planning." (FGD woman, Kabwayi colline)

Unbalanced power and gender roles
Group discussions also re ected the fact that men do not participate in family planning sensitization and do not take any responsibility towards family planning yet they have a predominant role in family decision-making including childbearing.
"Women are victims of men who do not understand family planning policies. They spend most of the time in bars and when they come at home, they are drunk and force us into sexual intercourse while we are in our fertile period and we get unplanned pregnancies in that way." (FGD woman, Karagara colline) Discouragement from family planning adherence by health practitioners Some participants shared that in their health facility, they have encountered health practitioners who were religious and discouraged patients in adopting family planning methods. They advise them to consider only natural methods and emphasize the fact that modern contraceptive methods have many side effects.
"Some medical staff are against modern contraceptive methods because of their side effects. If the health practitioners doubt on those methods, we will be more doubtful about adopting any form of modern contraception methods." (FGD man, Migera colline)

Discussion
This mixed-methods study assessed the factors that hindered uptake of family planning methods among women of reproductive age (15 to 49 years) in rural collines of Vyanda and Rumonge districts of Burundi.
The study found that knowledge on contraception methods was high with 94.3% of respondents able to cite at least one modern contraceptive method whereas utilization was low (22.6%). This con rms the BDHS nding which suggested that 97% of men and women know at least one contraceptive method but only 29% used any family planning method at the time of the survey [3].
To understand the increased gap between the knowledge and the utilization of contraceptive methods, a qualitative study involving focus group discussions with men, women, community, and church leaders was conducted. Five recurring themes emerged from the qualitative analysis. Firstly, the fear of side effects from use of contraceptives and its management which is not handled appropriately by health workers. Second, religious beliefs and in uence of religious actors which signi cantly impede uptake of modern contraceptives also emerged as a major determinant affecting uptake. Third, cultural norms and social constructs on childbearing was found to be another contributory factor. At the household level, lack of communication between spouses in having common understanding of how planning a family also served as a barrier. Also, low family planning methods uptake is explained by unbalanced power and gender roles which give men prerogatives to control woman's procreative power while the men are less knowledgeable about reproductive health. Finally, some health practitioners discourage utilization of modern contraceptive methods, citing reasons of religion and/or side effects.
This study highlights key lessons which should be considered for future program interventions. Fear of perceived side effects emerged as the rst and most important factor affecting utilisation of family planning methods. After experiencing unpleasant side effects especially from using modern contraceptives, a signi cant fraction of women discontinue utilisation. These negative experiences are shared with social networks which, as a result, heighten the worries of side effects, serving as a deterrent for others to adopt these modern methods. Other studies in similar contexts in Africa have found consistent ndings [14,15]. In many instances, these side effects are also misconstrued especially by religious actors as 'payment for the sin of using family planning' and discontinuing it would be the only way to relief and freedom. This belief is held in other African contexts and this has been captured by academic literature [16,17].
Spousal miscommunication, unbalanced power and gender roles and non-involvement of male in family planning awareness program hinder the success of family planning interventions in such conservative communities. This is exacerbated by cultural norms in favour of large families and patriarchal society where family planning decisions must be approved by men whom are considered as head of the family despite having limited knowledge in this area. Generally, patriarchy in rural communities especially in sub-Saharan Africa have always been a determinant of reproductive decisions inclusive of family planning. This has become a political issue which remains unattended even at national and international levels [18,19]. In our communities, men who accepted family planning could not own up and defend their decision publicly as it was recognised as a sign of weakness and contravention to the cultural norms which strongly upheld increased household size as a wealth and power.
Our work suggests that family planning uptake can be increased in conservative rural communities via overall health systems strengthening including health workers' capacity to administer family planning methods with dignity and without blemish. Communication of side effects at the onset of utilisation of family planning is key, as such, health staff especially those at primary healthcare should be well-trained to advise accordingly. Out-of-pocket payments associated with management of side effects are hindering factors, therefore implementing interventions to reduce the burden of payment could be critical and essential to uptake and continuation of modern contraceptives.
Accessibility to family planning in this setting is key and evidence from elsewhere has shown the importance of using community health workers in offering doorstep service especially in hard to reach places. One study in Uganda proved that in addition to pills and condoms, community health workers in Uganda could effectively provide injectable services and offer side effects counselling services effectively [21]. Successful uptake of family planning in Ethiopia, Malawi and Rwanda all underscores the distinct contribution of health extension workers and overall health system strengthening strategy [22].
Men and religious leaders at the community level should be involved in family planning awareness initiatives. An assessment of a male motivator project in Malawi provided evidence that men involvement in family planning program signi cantly increased spousal communication on family planning and effectively promote contraceptive uptake [20]. Moreover, religion plays a critical role in shaping people's ideas, views and thoughts. Therefore, incorporating aspects of religion into family planning interventions could yield some positive impacts [23]. Adedini et al. [23] shared evidence of Nigerian urban Reproductive Health Initiative which suggests that there is a signi cant association between contraceptive uptake and exposure to family planning messages delivered by religious leaders.

Limitation of the study
Studies that employ responses from retrospective activities could result in recall bias. Recall bias could result in either overestimating or underestimating results presented here. To reduce the impact of recall bias, an exploratory approach in the form of focus group discussion was undertaken to validate and explain results from the quantitative research phase. Focus group discussions, if not moderated effectively could bring about differences in power dynamics among participants which, resulting in unequal contributions from study participants and skew results to only re ect the most powerful and vocal participants. This could have been our case, however, the employment of an experienced focus group facilitator for all the sessions, to a larger extent, averted all possibilities of this bias.

Conclusion
This study emphasizes that the knowledge of contraceptive methods and the desire to limit the family size do not translate in contraceptive uptake. This information contributes to the existing body of knowledge on family planning in the context of Burundi and to a lesser extent, sub-Saharan Africa. Given that Burundi is among the countries with the least use of family planning methods, this study brings more clarity to the idiosyncratic factors affecting use especially among those living in conservative rural areas.
The multifaceted impediments to increased contraceptive utilization must be considered by policy makers and program implementers in order to develop tailored interventions that are more integrated and likely to yield best results. To make strides towards universal access to family planning, government and partners should tackle the issue of side effect management and promote community-based family planning approach. The latter may include religious leaders' engagement to share scriptural messages in favour of family planning during mass campaign as it is anticipated that they can in uence behaviour change of their followers. Capacity building and equipping community health workers to provide family planning counselling services, addressing community misbeliefs around family planning and door-step delivery of modern contraceptives could be an effective strategy worth the investment.

Declarations
Ethics approval and consent to participate A signed approval was obtained for this study from the local ethics committee and health authorities in the Bururi and Rumonge province after presentation of a study protocol and questionnaires. The study was also internally commissioned and approved by senior management of the organisation after review of the study protocol and questionnaires. Study participants also signed a consent form ahead of interview and child anthropometric measurements.

Consent for publication Not Applicable
Availability of data and materials The data used for analysis in this manuscript are available from the Research, Monitoring and Evaluation Department of VHW. Data is restricted but available from the authors upon reasonable request.

Competing interests
The authors declare no competing interests Funding Not Applicable