Factors associated with the discontinuation of modern contraceptive methods among lactating women in nine West African high-fertility countries: findings of the most recent demographic and health surveys

Abstract

Introduction

Contraception discontinuation is a concern, especially if it occurs in breastfeeding women, thereby exposing them to a high risk of close and unwanted pregnancies. Our study aimed to measure the prevalence and identify the individual and community-level factors associated with the discontinuation of modern contraceptives among breastfeeding women.

Methods

This was a secondary analysis of retrospective data of the most recent Demographic and Health Surveys (DHS) data from nine high-fertility rate countries, conducted mostly between 2018–2021. We reported weighted frequencies of modern contraceptives discontinuation (binary variable, coded 1 and 0). The independent variables included individual-level variables, including sociodemographic characteristics, female reproduction and family planning history, the women and their households exposure to media, and community-level ones such as place of residence (urban and rural) and country. Multilevel-modified Poisson regression was used to identify associated factors at the 5% threshold.

Results

The overall prevalence of modern contraceptives discontinuation was 13.1% among 5,599 lactating mothers, with wide variations between countries (prevalence ranging from 8.2% in Sierra Leone to 33.6% in Guinea). Women were more likely to discontinue contraception if they were the head of the household (adjusted prevalence ratio (aPR) = 1.71; 95% CI [1.17–2.50]; p = 0.006). In addition, compared to implant users, women using pills (aPR = 3.06; 95% CI [2.24–4.16]; p < 0.001), those using injectables (aPR = 2.80; 95% CI [2.16–3.62]; p < 0.001), and women whose partners used condoms (aPR = 2.30; 95% CI [1.47–3.59]; p < 0.001) were more likely to discontinue contraception. Moreover, women who were not sexually active (aPR = 2.11; 95% CI [1.75–2.54]; p < 0.001) and those who wanted children within two subsequent years (aPR = 1.84; 95% CI [1.36–2.48]; p < 0.001) were more likely to discontinue contraception. Finally, method discontinuation varied by country, with women in Gambia, Guinea, Mauritania, and Mali more likely to discontinue a modern contraceptive method than those living in Burkina Faso.

Conclusion

To improve the retention of women using contraceptive, high-fertility rate countries need to focus on contraceptive education, communication about side effects, dissemination of family planning messages through the media, and regular monitoring of women taking contraceptives.

Résumé

Introduction

L’abandon d’utilisation de la contraception chez les femmes allaitantes est préoccupant pour les politiques de santé. L’objectif de notre étude était de mesurer la prévalence et d’identifier les facteurs individuels et communautaires de l’abandon de la contraception moderne chez les femmes allaitantes.

Méthodes

Il s’est agi d’une analyse secondaire des données les plus récentes enquêtes démographiques et de santé de neuf pays à forte fécondité. Nous avons rapporté les fréquences pondérées de l’abandon d’utilisation de la contraception moderne (variable binaire, codée 1 et 0). Les variables indépendantes étaient les variables du niveau individuel (caractéristiques sociodémographiques, variables liées à la reproduction des femmes et à planification familiale et variables liées l’exposition des femmes et de leurs ménages aux médias) et du niveau communautaire (milieu de résidence et pays). Une régression de Poisson modifiée multiniveau a été réalisée afin de rechercher les facteurs associés au seuil de 5%.

Résultats

Notre étude a rapporté une prévalence globale d’abandon de la contraception moderne de  13,1% d’abandon sur les 5599 femmes allaitantes avec des disparités énormes entre les pays (prévalences allant de 8,2% en Sierra-Leone à 33,6% en Guinée). Les femmes qui étaient chef de ménage (Ratio de prévalence ajusté (RPa) = 1,71 ; IC à 95% [1,17-2,50] ; p=0,006), utilisatrices des pilules (RPa = 3,06 ; IC à 95%  [2,24-4,16] ; p<0,001), utilisatrices des injectables (RPa = 2,80 ; IC à 95% [2,16-3,62] ; p<0,001) et les femmes dont les conjoints utilisaient les condoms (RPa = 2,30 ; IC à 95% [1,47-3,59] ; p<0,001), étaient plus susceptibles d’abandonner la contraception. En outre, les femmes qui n’étaient pas sexuellement actives et celles qui avaient un désir d’enfant dans les deux années subséquentes (PR ajusté = 1,84 ; IC à 95% [1,36-2,48] ; p<0,001)  étaient plus susceptibles d’abandonner la contraception.

Conclusion

Dans le but d’améliorer la rétention des femmes utilisatrices de la contraception, les pays ouest africains doivent mettre l’accent sur l’éducation sur la contraception, la communication sur les effets secondaires, la diffusion de messages de planification familiale à travers les médias et un suivi régulier des femmes sous contraception.

Text box 1. Contributions to the literature
• Studies of contraceptive use worldwide indicate that women typically stop using contraception for various reasons, including the desire to become pregnant.
• However, contraceptive discontinuation has not been studied in women with children under 24 months.
• This is why this study was conducted in nine West African countries with high fertility rates.
• We found that the prevalence varied across countries, ranging from 8.2% in Sierra Leone to 33.6% in Guinea.
• To ensure that breastfeeding women continue to use contraception, West African countries should focus on effective communication about contraceptive side effects and regularly monitor women who are using contraception.

Introduction

Family planning (FP) includes all practices that enable individuals or couples to avoid unwanted pregnancies, regulate the time between two pregnancies, and decide when to have children and how many to have, depending on their age and socioeconomic characteristics [1]. Appropriate use of family planning not only helps to avoid unwanted pregnancies but also reduces maternal mortality due to unsafe abortions and prevents sexually transmitted diseases [2, 3]. Modern contraception is, therefore, an important mechanism for controlling high fertility and improving the physical and economic well-being of women and their families [4, 5]. In addition to these benefits, it contributes to healthy and productive families, food security, and sustainable development in low- and middle-income countries [5].

Despite these benefits, contraceptive use among women of childbearing age is low in low-income countries [6, 7]. This is reflected in the high fertility rate in this part of the world. In 2021, the World Bank identified ten countries in sub-Saharan Africa with fertility rate above 5, higher than the world and African averages of 2.47 and 4.44, respectively [8].

The situation is the most worrying because, despite these low levels of use, a significant proportion of women discontinue modern contraception prematurely [9, 10]. The word "discontinuation" is used to refer to women who have started a contraceptive method and stopped using it for any reason while they are still at risk of getting pregnant [11]. According to Family Planning 2020, more than one-third of women who started using a modern contraceptive method stopped using it within the first year, and more than half stopped before two years [11]. According to the WHO, after three months of use, 40% of women in Egypt, 51% in Kenya, 73% in Malawi, 56% in Tanzania, and 47% in Zimbabwe were at risk of becoming pregnant [12].

The issue of women discontinuing contraception, in general, is so worrying that several authors have studied the topic. In 2016, Fekadu et al. concluded that majority of women in Ethiopia had stopped using contraception after three years [13]. Ouédraogo et al. reported similar findings in Burkina Faso, Mali, and Niger in a multi-country study in 2021 [14]. The problem is even more worrying for breastfeeding women with a live child under 24 months. This group of women is exposed to the potential risk of an imminent pregnancy, with all the possible severe complications. This means that stopping contraception exposes these women to a new pregnancy with a birth interval of less than 33 months, which is below the WHO recommendations of at least 24 months to avoid a very high risk of morbidity and mortality for the woman and her child [15, 16].

To the best of our knowledge, no study has examined contraceptive discontinuation among breastfeeding women with a child under 24 months of age. The rationale for conducting this study in breastfeeding women is that, firstly, particularly in sub-Saharan Africa, discontinuation of modern contraception in women of childbearing age or young women has been widely studied in the literature [12, 15, 17,18,19,20]. Secondly, one of the reasons women give for stopping contraception is the desire to have a child, and finally, according to the World Health Organisation, a woman who becomes pregnant less than 24 months after her last birth is exposed to several health risks [15, 16]. We therefore wanted to study the factors associated with contraceptive discontinuation in breastfeeding women with an under 24-months child.

The interest of our study is, therefore, to examine modern contraceptive methods discontinuation among breastfeeding women and its associated factors using the DHS data from nine West African countries with high fertility rates.

Methods

Study settings

Nine low- and middle-income countries in sub-Saharan Africa were included: Burkina Faso, The Gambia, Guinea, Mali, Mauritania, Niger, Nigeria, Senegal, and Sierra Leone. These countries were selected because their fertility rates were among the highest in the world [21]. Table 1 shows the fertility rates by country in 2021.

Table 1 Fertility rates by country in 2021 and year of the DHS

Operational definitions

Contraceptive discontinuation is defined as a woman who has started a contraceptive method and stopped using it for any reason while at risk of becoming pregnant [11]. We did not include method switching.

Modern contraceptive methods include female and male sterilisation, intrauterine devices (IUDs), injectables, implants, pills, and condoms [22].

Study design and period

This was a secondary analysis of the DHS data collected from nationally representative households. The data used are from the women’s reproductive calendar. The DHS reproductive calendar comprises a month-by-month history of reproductive events for a woman of childbearing age, typically for five years preceding the interview. Most calendar datasets cover pregnancies, births, pregnancy terminations, contraception use and reasons for its discontinuation [23]. Except for Niger (2012), the data were collected between 2018–2021 for all the countries.

Study population

The study population consisted of breastfeeding women with children under 24 months who lived in the nine countries listed above at the time of the different surveys.

Inclusion criteria

Women were included if they were breastfeeding, had a child less than 24 months old, and had used at least one modern contraceptive method since their last birth.

Exclusion criteria

Pregnant women were excluded during the analysis process.

Data source and sampling

DHS data are collected periodically in low- and middle-income countries using standardised, pre-tested, and validated questionnaires and follow a similar sampling, data collection, and coding procedure that allows cross-country analysis. The DHS uses direct interviews with women for questions related to reproductive health. In each country, the sample is generally representative at the national level, residence level (urban–rural), and regional level (departments, states). The sample is usually based on a stratified two-stage cluster design. In the first stage, enumeration areas are generally drawn from Census files, and in the second stage, in each selected enumeration area, a sample of households is drawn from an updated list of households. A household questionnaire collects informations on household characteristics and the usual residents and visitors to identify eligible women of childbearing age. Eligible respondents are then interviewed using an individual women's questionnaire (for women of childbearing age) covering a non-exhaustive list of topics, including sociodemographic characteristics, childbearing behaviour and intentions, contraception, antenatal care, childbirth, and postnatal care. Data quality checks were conducted continuously to improve instruments.

Variables

The dependent variable in this study was contraceptive method discontinuation. It is a binary variable, with a value of '1' for women who have discontinued contraception and '0' for those who have not.

The independent variables at the individual level that influence contraceptive discontinuation were grouped into sociodemographic variables, variables related to women reproduction and family planning, and variables related to women and their households exposure to the media.

The community-level independent variables are the area of residence (urban and rural) and country.

The different variables and their modalities are listed in Table 2.

Table 2 Variables used

Data analysis

Categorical variables were described using weighted frequencies. The "svy" command was used to adjust for sampling weights and cluster effects from the Demographic and Health Survey (DHS). Multilevel modified Poisson regression with robust variance ('mepoisson' command) was used to identify individual and community factors associated with contraceptive discontinuation. Four models were fitted in sequence: (i) model 1 or the null model included only the dependent variable, (ii) model 2 included the dependent variable and the individual-level variables, (iii) model 3 included the dependent variable and the second-level (community) variables, and finally (iv) model 4 included the dependent variable and the individual-level and community-level factors whose associations were considered relevant in the univariate analysis. Model 2 and Model 3 are presented in the Supplementary file. Fixed effects were presented as adjusted prevalence ratios (aPR) with their 95% confidence intervals. The goodness of fit of the models was assessed using the Akaike information criterion (AIC), with a lower AIC indicating a better fit. All analyses were performed using Stata version 18 [25], and the significance level was set at p < 0.05.

Ethical considerations

Data for these countries were obtained from the official DHS program database after permission was granted via an online request explaining the study aim. The downloaded databases do not identify the survey participants, so the anonymity and confidentiality of the participants were maintained. Approval from each country's ethics and research committees was required before implementing the various surveys.

Results

Flow chart

This study included 5,599 women breastfeeding a child under 24 months. Figure 1 shows the study flowchart.

Fig. 1

Flowchart of the study

Prevalence of modern contraceptive discontinuation

A total of 733 lactating mothers with a child under 24 months of the 5,599 discontinued their contraceptive method. Therefore, the pooled prevalence was 13.1% (95% CI = 12.1–14.2). The weighted prevalence varied across the countries, ranging from 8.2% in Sierra Leone to 33.6% in Guinea (Fig. 2).

Fig. 2

Contraceptive discontinuation (proportion) among breastfeeding women

Sample description

Women's sociodemographic characteristics

The contraceptive discontinuation rate was 22.4% among women whose head of household was female, 27.5% when the respondent was the head of household herself, 20.8% when the head of household was under 25, and 15.0% among women with no occupation. These results are shown in Table 3.

Table 3 Modern contraception discontinuation by women's sociodemographic characteristics among lactating women in nine West African high-fertility countries

Women's reproductive and family planning characteristics

The contraceptive discontinuation rate was 18.8% among pill users, 16.0% among users of injectables, and 26.7% among women who had not been sexually active in the four weeks prior to the survey. These results are presented in Table 4.

Table 4 Modern contraception discontinuation by women's reproductive and family planning characteristics

Variables related to women's and households' exposure to the media

In this section, we looked at women and households exposure to the media. The modern contraception discontinuation rate was 15.6% for women whose households did not have a radio, 14.5% for women who did not listen to the radio, and 14.4% for women who had not heard anything about family planning in the last few months. These results are presented in Table 5.

Table 5 Modern contraceptive discontinuation by women's and households' media exposure

Contextual factors

Contraceptive discontinuation was not related to place of residence (13.7% in urban compared to 12.6% in rural areas; p = 0.323). However, it varied across countries, as shown in Fig. 2.

Reasons for not using modern contraception

The most common reason why women stopped modern contraception, as reported by women, was the side effects (29.1%), followed by lack of a partner or infrequent sex (24.7%). The third most common reason was that the woman had stopped because she wanted a more effective method (7.9%), followed by women who wanted to get pregnant (7.5%). We also found that 47 women, or 6.7% of the total, stopped using the method because their husbands/partners objected. Other equally important reasons were the difficulty of using the method (4.0%), and accessibility or availability problem (3.7%).

Factors associated with modern contraceptive discontinuation

After adjusting for contextual variables, women were more likely to discontinue modern contraception if they were the head of the household (adjusted Prevalence ratio (aPR) = 1.71; 95% CI [1.17–2.50]) compared with women who were not the head of the household. In addition, compared with implant users, women using pills (aPR = 3.06; 95% CI [2.24–4.16]; p < 0.001), injectables (aPR = 2.80; 95% CI [2.16–3.62]; p < 0.001) and women whose partners used condoms (aPR = 2.30; 95% CI [1.47–3.59]; p < 0.001) were more likely to stop using contraception. Finally, women who were not sexually active (aPR = 2.11; 95% CI [1.75–2.54]; p < 0.001) and those who wanted to have a child within two years (aPR = 1.84; 95% CI [1.36–2.48]; p < 0.001) were more likely to discontinue contraception. Finally, method discontinuation varied by country, with women in Gambia, Guinea, Mauritania, and Mali more likely to discontinue a modern contraceptive method than those living in Burkina Faso. All these results are presented in Table 6.

Table 6 Factors associated with contraceptive discontinuation in multilevel multivariate analysis among lactating women in nine West African high-fertility countries

Discussion

We identified factors associated with modern contraceptive discontinuation among breastfeeding women with a live child under 24 months of age. Overall, the prevalence of modern contraceptive discontinuation was 13.1%. We find this prevalence quite worrying because of the consequences that discontinuation can have on women’s reproductive health. Contraceptive discontinuation among breastfeeding women, who would be exposed to short birth intervals if they became pregnant, is associated with high rates of unintended pregnancies, leading to unsafe abortion and maternal and infant morbidity and mortality [26,27,28,29,30,31]. Our finding may be explained by the fact that breastfeeding women's specific family planning needs are not considered. Also, the methods available may not be adapted to their needs, and the follow-up of women who are using modern contraception is ineffective in our context. Women do not receive enough information about the possible side effects of the methods they choose [32]. All this contributes significantly to contraceptive discontinuation. In addition, several studies have previously reported a large number of women who stopped using contraception when they were at risk of getting pregnant [13, 14, 27, 28, 33].

This study found that side effects were the most common reason for discontinuation (29.1%). This worrying result means that a significant proportion of women stopped using contraception for reasons other than wanting to become pregnant. In other words, these women will become pregnant early and have unwanted pregnancies. However, the findings on side effects as the most common reason for discontinuation aligns with previous studies [30, 34, 35]. On the other hand, other studies have shown that wanting to become pregnant is the most common reason for stopping modern contraception [28, 36, 37]. This difference could be explained in our context by the prevalence of misconceptions about contraception, as reported by Tran et al. in Burkina Faso and the Democratic Republic of Congo in 2018 [38] and Alvergne et al. in Ethiopia in 2017 [39]. A literature review of factors influencing contraceptives use in sub-Saharan Africa between 2005 and 2015, published in 2017, also found that the negative factors reducing contraceptives use were women misconceptions about the side effects of contraception [40]. Awareness campaigns should be conducted to minimise the extent of these misconceptions about contraception. Husbands' reluctance to women's contraception use is also an important reason why breastfeeding women do not use contraception (6.7%). Several studies have found similar results [31, 38, 40]. It is therefore necessary to develop awareness-raising and information campaigns aimed at men so that they accept their wives use of contraceptive methods. In fact, a systematic review has shown that men related interventions are effective in increasing contraceptive use [41].

In this study, the final multilevel multivariate model showed that the type of method used, being the head of the household, lack of sexual activity, desire for children, and country of origin were the most important factors associated with contraceptive discontinuation.

Users of long-acting contraceptive methods, such as IUDs and implants, were less likely to stop using contraception than users of short-acting methods, such as the pill and injectables. These findings are similar to those of previous studies [28, 34, 39, 42,43,44,45]. In contrast to short-acting contraceptives, the lower discontinuation rate for long-acting methods may be explained by the fact that they are more difficult to discontinue and that their discontinuation requires the intervention of health professionals and sometimes has financial implications [44, 46, 47]. Women who were heads of household are more likely to discontinue modern contraception than women who were not heads of household. According to Wayack and Moussa (2015), the conditions for a woman to become head of household are, firstly, that she is a widow (40%), then that she is married to a migrant (26%), or that she is single (22%) [48]. Furthermore, according to the same authors, married women with a cohabiting partner do not identify themselves as the head of their household [49]. The lack of sexual activity among this category of women could explain their tendency to stop using modern contraception.

This study shows that despite government efforts to achieve national family planning goals and despite low contraceptive prevalence, a significant proportion of breastfeeding women discontinue contraception. Information on contraception, communication about side effects, access to information through the media, and regular monitoring would enable women to overcome barriers to contraceptives use. This will improve the design and implementation of family planning programs for breastfeeding women.

The study's main strength is its use of information from the reproductive calendar (monthly information). This study has several limitations. The first limitation is recall bias due to the retrospective nature of reproductive calendar data collection and social desirability bias. Regarding recall bias, some authors have explored the reliability score of the reproductive calendar data [50, 51]. For example, Tumlinson and Curtis stated that when using retrospective calendar data, attention must be paid to the potential for individual reporting errors. Anglewicsz et al., used population-based longitudinal data from nine settings in seven countries (Burkina Faso, Nigeria (Kano and Lagos States), Democratic Republic of Congo (Kinshasa and Kongo Central Provinces), Kenya, Uganda, Cote d'Ivoire, and India) to evaluate the reliability score and found that overall, the reliability of the calendar is in the “moderate/substantial” range for nearly all geographies and tests (Kappa statistics between 0.58 and 0.81). These results imply that some recall bias could affect the results of the reproductive calendar data. To mitigate some data quality concerns when using the reproductive calendar data, Bradley et al. [52], suggested that consistent recall of contraceptive use earlier in the six-year calendar period is better, for example, analyzing a shorter period of calendar data (e.g., 2–3 years rather than six years). Given this suggestion, we think these data quality concerns are addressed in this study because we limited the analysis to the last 24 months.

The second limitation is the heterogeneity of the years the DHS was carried out. Unfortunately, this limits the comparability of the results between countries. Although the DHSs used in this study were conducted between 2018 and 2019, we note that the Niger DHS was conducted in 2012 and the Burkina DHS in 2021.

The third limitation of this study is the unavailability of the type of breastfeeding in the databases, which could be a confounding factor. Indeed, as Bryant et al. reported milk supply concerns among women who started hormonal contraception [53], women may be confused as to what impact hormonal contraception has on lactation and infant growth, and so could stop the contraceptive method even if evidence from randomized controlled trials on the effect of hormonal contraceptives on lactation and infant growth is not consistent across trials [54, 55].

The last limitation is related to the data collection techniques. The responses were obtained from women by direct interview and are therefore subject to social desirability biases: women may voluntarily under-report contraceptive discontinuation.

Conclusion

Our study found a contraceptive discontinuation prevalence of 13.1% among lactating women in nine countries in sub-Saharan Africa. Factors associated with contraceptive discontinuation were the type of method used, being the head of the household, lack of sexual activity, desire for children, and country of origin. To help achieve the Sustainable Development Goals and ensure that breastfeeding women continue to use contraception, West African countries should focus on contraceptives education, communication about their side effects, dissemination of family planning messages through the media, and regular monitoring of women using contraception.