Background

Family planning allows couples and individuals to exercise their fundamental right to freely and responsibly choose the number and timing of their children [1, 2]. This practice is supported by demographic, health, and human rights considerations and plays a crucial role in empowering women, alleviating poverty, and promoting sustainable development [3]. Modern contraceptive methods have been shown to effectively prevent unintended pregnancies, leading to a decrease in maternal and child mortality rates, a reduction in unsafe abortions, and an improvement in the overall health of women and families [4, 5]. These methods are products or medical procedures that prevent reproduction from sexual intercourse [6]. It includes male and female sterilization, injectables, intrauterine devices (IUDs), contraceptive pills, implants, condoms, the standard days’ method, the lactational amenorrhoea method, and emergency contraception [7].

Unintended pregnancy and births are becoming a significant public health problem that causes illness during pregnancy, complications at birth, and maternal deaths [8]. Roughly in the globe, 121 million unintended pregnancies occurred each year between 2015 and 2019, and of these unintended pregnancies, 61% (73 million) ended in abortion per year [9]. Considerably, unintended pregnancy and childbirth are higher in developing than in developed regions [8, 10]. Nevertheless, evidence suggests that implementing effective strategies such as family planning could help prevent unintended or unwanted pregnancies in developing nations [11].

Around the world, 966 million women in the reproductive age group are currently utilizing a form of contraception, with about 874 million of them opting for modern contraceptive options. However, despite these numbers, there remain approximately 164 million women who wish to postpone or prevent pregnancy but are not practicing any form of contraception [12]. In Sub-Saharan Africa, only 56% of women who desire to avoid pregnancy utilize modern contraceptive methods, marking the region with the lowest usage rate [12]. Additionally, many of the most impoverished countries in sub-Saharan Africa continue to exhibit a substantial unmet need for family planning services [13]. Based on the findings of the 2016 Ethiopian Demographic and Health Survey (EDHS), the adoption of modern contraceptives among reproductive-age women stands at 35% for currently married individuals and 55% for sexually active unmarried women [7], figures that fall significantly short in effectively addressing the prevalent unmet needs and unintended pregnancies in Ethiopia.

To date, Ethiopia has been striving to enhance the adoption of contraception and bolster maternal, neonatal, and child health outcomes. Despite the combined endeavors of the Ethiopian government, NGOs, and the evident advantages of contraceptives, the uptake of modern contraceptive methods still remains low, with a rising unmet need reaching 22% [14] and lags behind global family planning benchmarks and objectives [15].

However, evidence has indicated that meeting all unmet needs for modern contraception in Ethiopia could lead to a 90% decrease in unintended pregnancies, subsequently resulting in a reduction of unplanned births from 1.2 million to 128,000 and a drop in abortions from 571,000 to 59,000 [16].

Sustainable Development Goal (SDG) goal 3 also emphasized universal access to sexual and reproductive health care services for healthy lives and well-being for all ages by 2030 [11]. Subsequently, Ethiopia is working towards achieving this goal and improving momentum in reaching the HSTP II and SDGs, mainly target 3.7, which focuses on ensuring access to family planning information and education by 2030 [14, 17].

Several studies elsewhere in the country revealed many reasons for low modern contraceptive use [18–21]. Most previous studies focus only on individual factors and neglect contextual factors. In addition, unmarried women, who are at a higher risk of unintended pregnancy, are frequently overlooked in research, which tends to focus on married women [22, 23], Although monitoring and reporting on modern contraceptive prevalence rates for all women have been shown to encourage the use of contraception [24].

Hence, this current study considered all reproductive-age women to represent the study population best. The primary objective of this study was to examine the individual and community factors influencing the utilization of modern contraceptives among women of reproductive age in Ethiopia, based on data from the 2019 Performance for Action Ethiopia (PMAET-2019) survey.

Moreover, previous studies’ analysis only utilized the odds ratio, the proportion of change in variance (PCV), and intra-cluster correlation (ICC) to describe multilevel results. However, this current research incorporated summary measures such as Interval Odds ratio (IOR) and proportion of opposed odds ratio (POOR), as well as measures of components heterogeneity like Median odds ratio (MOR), to enhance the informativeness and explanatory power of our findings. The insights gained from this study will be valuable for policymakers, healthcare organizations, communities, and other stakeholders working towards enhancing maternal, newborn, and child health outcomes through the effective promotion of modern contraceptive methods.

Methods and materials

Sample

First, 9107 women of reproductive age 15–49 years were identified from the PMA-ET 2019 survey to identify the eligible study population for this study. Then, we dropped incomplete household and female interviews. The following Stata command criteria were used to identify the female respondents for this study. Firstly, keep if the woman is part of a household that completed the household survey (HHQ_result = = 1) and completed the female survey (FRS_result = = 1). So, incomplete household and female questionnaire interviews were deleted using the Stata command keep if HHQ_result = = 1 & FRS_result = = 1 [27]. Accordingly, 8928 remained. Then, pregnant women were excluded as they were not eligible for modern contraceptive method use. Sexually active women were also selected by excluding those who never had sexual intercourse. Finally, after excluding the menopause and infecund women, a total of 6117 eligible women were identified and included in the analysis [28]. Details are indicated in (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Selection of eligible women for modern contraceptive utilization among reproductive-age women in Ethiopia, data from PMA-ET, 2019.

https://doi.org/10.1371/journal.pone.0303803.g001

Individual-level variables

The level 1 (individual women level) variables included age, educational status, marital status, marriage history, place of birth, parity, Educational Status, religion, the intention to have the next birth, household wealth quintile, knowledge of contraceptive methods, whether visited a health facility in the past 12 months and discussed family planning, whether visited by a health worker in the past 12 months (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptions of the dependent and independent variables considered in this study.

https://doi.org/10.1371/journal.pone.0303803.t001

Results

Socio-demographic and reproductive (Individual-level) characteristics of the respondents

This study focused on a sample of 265 clusters (Enumeration areas)in Ethiopia. We included unweighted 6117 or weighted 6065 reproductive-age women in the 265 clusters (Enumeration areas). The prevalence of modern contraceptive method use among the respondents was 37% (34.3–39.8). However, more than 97% of the respondents have heard of at least one modern contraceptive method. About 1588 (43.4%) of non-current users of contraceptive methods have a future intention to use modern contraceptive methods. Around 2,812.9 (46.4.%) of the respondents have no education. Of these, 2480(96%) could not read and write a simple sentence. More than half, 3343(55.13%) of the respondents were aged 20–34. About 5007(82.56%) respondents were married/living with a partner. About 1403.6 (23.14%) respondents had six or more children. The proportion of respondents who preferred wait duration from now before another childbirth was soon/now was 2711(68.15%) (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Socio-demographic and reproductive (individual-level) characteristics of respondents, data from PMA-Ethiopia, 2019 (n = 6117).

https://doi.org/10.1371/journal.pone.0303803.t002

Community-level characteristics (higher-level variables) of the respondents

Two hundred sixty-five clusters (Enumeration areas) were considered a unit of analysis. Accordingly, 52% and 59.6% of the clusters had lower community literacy levels and exposure to FP information, respectively. Over half (52.8%) of the clusters have empowered women and girls about FP. Half (49.5%) of the clusters have a non-supportive attitude/norm toward family planning. More than half (54.2%) of the clusters had poor knowledge of modern contraceptive methods, and 15.2% of the clusters’ community way of life belonged to pastoralist communities (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Community-level variables of respondents, data from PMA-Ethiopia, 2019 (n = 6117).

https://doi.org/10.1371/journal.pone.0303803.t003

Multivariable multilevel logistic regression analyses

Model building and selection.

Of the four models (I-IV) that we fitted, model IV best fits the data. The random effect results were estimated using intra-class correlation (ICC), variance partition coefficient (VPC), a proportional change in variance (PCV), shrinkage factor, Interval odds ratio (IOR), and proportion opposed odds ratio (POOR), Median odds ratio (MOR). Thus, a two-level mixed-effects logistic regression model was considered for further analysis to consider all the above.

As shown in the table below for the empty model, 21.9% of the total variance in the odds of using modern contraceptives was attributed to the between-cluster variation (ICC = 0.219). The result of the random effect model showed that the intra-class correlation coefficient (ICC) was 21.9%, indicating that the correlation between women in the same EA on modern contraceptive use was 0.219, i.e., 21.9% of the variation in the utilization of modern contraceptives among reproductive age women could be attributed to EA difference.

In the final model, the proportional change of the variance (PCV) of the combined factors was 20.5% (τ = 0.732). This indicated that individual and community-level predictors simultaneously explained 20.5% of the variation in utilizing modern contraceptives across communities. Likewise, a shrinkage factor was calculated better to interpret the intra-cluster association in the cluster level factor. Model III’s shrinkage factor value was 0.893, which is nearly one. This indicates that the population-average regression coefficients (and corresponding population-average odds ratios) for the four clusters/community-level factors are very close to the conditional or cluster-specific regression coefficients and corresponding odds ratios. The 80% Interval odds ratio (IOR) for the four community-level factors in model III were (0.48, 10.7) for pastoralist vs. agrarian community way of life and (0. 308, 6.87) for lower vs. higher community literacy level. Besides, it was (0. 31, 6.927) for un-empowered vs. empowered women and girls about FP, (0.29, 6.46) for community non-supportive attitude/norm vs. supportive attitude/norm towards FP. Overall, as the interval contains 1, the cluster variability is large compared to the effect of the cluster-level variable.

In addition, the proportion opposed odds ratio (POOR) of the community factors value was 0.248 for the community way of life, 0.378 for the community literacy level, 0.375 for women and girls’ empowerment about FP, and 0.397 for community attitude/norms towards FP. Thus, in 37.5% of comparisons between the community with non-empowered and empowered women and girls about FP, the OR for this comparison would be different in direction from that of the overall OR at the community level. Moreover, in a 37.5% pairwise comparison, the odds of modern contraceptive utilization would be higher in the community with empowered women and girls about FP than in non-empowered. Further, in the model, the Median Odds ratio (MOR) was 2.5 (empty model), 1.65 (individual factors), 2.19 (community factors), and 2.26 (combined model). The MOR for individual factors was 1.65, with a reciprocal of 0.6. Three out of four individual-level factors had odds ratios falling within the interval of 0.6 to 1.65, indicating consistent effects across different clusters. These factors did not significantly deviate from one, suggesting a limited impact on variation in modern contraceptive utilization between clusters. Therefore, individual-level factors do not strongly influence differences across clusters, and cluster effects play a more dominant role in explaining variability. Additionally, clustering notably affected the final model, as the MOR for unexplained community variation decreased from 2.5 in the null model to 2.26 in the final model when all predictors were considered (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Measures of the effects and heterogeneity of community-level variables (Variance, a proportional change in variance, Shrinkage factor, interval odds ratio, proportion of opposed odds ratio, and the median odds ratio), data from PMA-Ethiopia, 2019 (n = 6117).

https://doi.org/10.1371/journal.pone.0303803.t004

Individual and community-level determinants of modern contraceptive utilization

Effect of individual-level characteristics on modern contraceptive utilizations.

After adjusting for the individual and community factors, reproductive-age women with the marital status of divorced/widowed were less likely to report modern contraceptive use (AOR:0.18, 95% CI 0.13,0.23) than never-married women. The data also showed that the odds ratio of modern contraceptive utilization increases with an increasing level of education. For instance, women who attended primary school, secondary school, and more than secondary were more likely to report modern contraceptive utilization (AOR: 1.47, 95% CI 1.20, 1.79), (AOR: 1.73, 95% CI 1.33, 2.26) and (AOR: 1.58, 95% CI 1.16, 2.17), respectively than with no education. Moreover, women who reported having facility visits for family planning discussions in the last 12 months were more likely to report modern contraceptive use (AOR:1.84, 95% CI 1.52, 2.23) compared to never having had a visit in the last 12 months.

Effect of community-level factors on modern contraceptive utilizations.

Accordingly, after controlling the effect of individual and community factors, clusters with a high number of agrarian community ways of life were reported to have better modern contraceptive utilization (AOR: 2.27, 95% CI 1.5, 3.44) and higher community literacy levels were reported to have better modern contraceptive utilization (AOR: 1.46, 95% CI 1.09, 1.94). Similarly, clusters that empowered women and girls about FP were more likely to report FP use (AOR: 1.47, 95% CI 1.11, 1.93) than non-empowered clusters. Besides, clusters with supportive community attitudes/norms toward FP were associated with higher odds of modern contraceptive use (AOR: 1.37, 95% CI 1.04, 1.81) than clusters with non-supportive community attitudes/norms towards FP (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Multilevel mixed-effects logistic regression modeling of individual and community level factors associated with modern contraceptive utilization among reproductive-age women in Ethiopia, data from PMA-Ethiopia 2019 (n = 6117).

https://doi.org/10.1371/journal.pone.0303803.t005

Discussion

This research indicated that the use of modern contraceptives among women of reproductive age was low, with only 37% utilizing such methods. This figure is consistent with the findings of the EDHS 2016 report [7]. However, it is higher than the rates reported in Sub-Saharan Africa’s demographic and health surveys [41]. Despite this, Ethiopia has not met the ambitious targets set by FP2020 to increase the contraceptive prevalence rate to 55% and the total fertility rate to three [15]. To achieve SDG target 3.7, significant efforts are required to improve access to and utilization of modern contraceptive methods in the country, as there is still a long way to go in meeting its family planning objectives.

The current study found that modern contraceptive use is influenced by individual factors such as marital status, education level, and discussions about family planning at health facilities. Community factors such as supportive norms, empowerment of women and girls in family planning, agrarian lifestyle, and higher literacy levels were also associated with increased modern contraceptive use. Together, these individual and community factors accounted for around 20.5% of modern contraceptive utilization among women of reproductive age in Ethiopia.

In this research, it was observed that women who are divorced or widowed were less likely to use modern contraceptives in comparison to those who have never been married. Studies conducted in sub-Saharan Africa highlighted a higher prevalence of contraceptive use among never-married women than divorced/widowed or married women [42]. Additionally, research indicated that married adolescent girls exhibit the lowest median prevalence of modern contraceptive use across various global regions [43]. This trend could be attributed to the fact that divorced/widowed women and married adolescents may not have reached their desired family size, while the infrequent sexual activity among divorced/widowed women might lead to a lower likelihood of contraceptive utilization. This collective data suggests that being married does not necessarily pledge to higher contraceptive utilization among women. As a result, family planning initiatives need to target reproductive-age women comprehensively rather than solely focusing on married women.

We found that the odds ratio of modern contraceptive utilization rises with higher levels of education, peaking at secondary education. However, there was a slight decrease in contraceptive use among women with education beyond the secondary level compared to those with secondary education. A study done in Nigeria revealed that increasing levels of education rise modern contraceptive utilization [44]. Many studies also show that a woman’s educational attainment is positively associated with contraception use [45–47]. Moreover, education alone was indicated to have had a significant effect on the utilization of contraception [48]. Women were less likely to utilize modern contraceptives without education [41]. The possible reason for the decrement in modern contraceptive utilization in those above secondary education might be due to late marriage after the completion of higher education and the desire for more children at later ages.

Having family planning discussions at a health facility was positively associated with modern contraceptive utilization. This aligns with a study done in Uganda, which discussed FP with a field or health worker and remained statistically significant [49]. Women were less likely to utilize modern contraceptives if they were not told of family planning at a health facility [41]. Hence, as women have more contact with healthcare providers, their chances of receiving information and using contraceptives will increase.

Thus, the women who get an excellent opportunity to be aware of FP would be vital to fill the missed opportunity [7]. In addition, it creates a conducive environment that helps mitigate the myths and misconceptions about FP and enhances the relationship between the health care providers and women, thereby increasing FP utilization. Such collective effort would contribute to increasing women’s actual modern contraceptive utilizationCommunity norms/attitudes towards FP utilization were a community-level factor influencing contraceptive utilization. Our finding aligns with another study done in Ethiopia and 73 low-and middle-income countries [33, 40]. In this study, women from clusters with supportive attitudes/norms toward family planning were associated with modern contraceptive utilization. Moreover, another study revealed that social pressure on women to bear and not bear children incurs social disapproval [50]. So, this was indicated to have a major influence on the intentions and behaviors of women regarding spacing and contraceptive use. Hence, it is noted that effort should be put into overcoming and changing the community attitude/norms regarding FP with the focus of promoting FP use may encourage women to utilize contraceptives.

This study indicates that higher women and girls’ empowerment in family planning was positively associated with modern contraceptive utilization. It was also indicated from the community-level variables that women who reside in a cluster with lower women’s and girls’ empowerment about family planning had lower odds of contraceptive utilization. These findings align with a study in Ethiopia revealing women’s empowerment as an essential determinant of contraceptive use. Moreover, empowered women were more likely to utilize contraceptives than un-empowered women [51]. Similarly, another study in Tanzania indicated that self-efficacy for contraception was associated with utilizing modern contraceptives [45]. Women who had complete control over their healthcare were more likely to use contraceptives than women who had no control over their healthcare [44]. A systematic review also revealed that women’s empowerment domains are consistently positively associated with ever use of contraceptives. In many Sub-Saharan African countries, this could also be justified as women’s roles and reproductive choices could be constrained by various cultural traditions that restrict women’s autonomy and encourage men’s dominance [52]. This disempowering of women and girls might directly impact women’s decision to utilize contraceptives.

Consistent with previous studies, this study also revealed a positive association between a community literacy level and modern contraceptive utilization. Cluster with higher women’s literacy status at the community level was positively associated with using modern contraceptives. Pieces of evidence from many individual studies also indicated that women from the community with higher women’s literacy status were more likely to be users of contraceptives than their reference group. Generally, women tend to have higher decision-making power regarding contraceptives by attaining higher literacy levels [53–57]. So, it is justified that being illiterate might make women unaware of the benefit of the optimal birth interval and could not overcome the misconceptions about contraceptive utilization. This might make clusters with low community literacy levels consider giving birth with a narrow birth interval as normal. The effect of community literacy in disseminating contraceptive information and counseling women is profound. The cumulative effect brings or predicts the actual use of contraceptives. This implies that a particular effort is needed to increase the number of women who are literate and, thereby, intend to utilize contraceptives.

Our study found clusters of being agrarian as a community way of life positively associated with modern contraceptive utilization. In line with these findings, pastoralist communities in Ethiopia showed lower contraceptive utilization compared to agrarian communities [58, 59]. One factor that could contribute to the decreased use of contraceptives within pastoralist communities is the detrimental impact of husbands, religious leaders, and clan leaders on family planning decisions. Furthermore, the lack of support for family planning methods within the pastoralist community, coupled with women’s limited access to education and low levels of community literacy, may also lead to lower utilization of contraceptives among this group [55, 60]. The disparity could also be the presence of improved infrastructure, schools, healthcare services, media exposure, and greater male participation in family planning decision-making among agrarian communities compared to pastoralist communities. As a result, agrarian communities may provide a more supportive setting for increased contraceptive use. Therefore, when implementing family planning interventions, it is crucial to consider each community’s specific lifestyle and characteristics to ensure the program’s effectiveness and relevance.

Our study also summarizes measures of the effect of cluster variables (interval IOR, POOR) and components of heterogeneity (VPC or ICC and MOR). In this case, It indicated that the clustering effect for modern contraceptive use was higher than that of individual factors. Hence, it is important to consider the community/cluster to promote the utilization of modern contraceptives. Besides, addressing and considering individual and community factors would enhance modern contraceptive utilization. Ultimately, our study results will be generalized to all reproductive-age women in Ethiopia.

Conclusion

The research revealed that modern contraceptive use among women of reproductive age in Ethiopia was lower than the country’s targeted contraceptive prevalence rates. Both individual and community factors played a significant role in understanding the factors associated with modern contraceptive utilization.

The study identified never being married, higher levels of education among women, and discussions about family planning at health facilities as individual-level predictors of modern contraceptive utilization. Additionally, clusters with greater empowerment of women and girls, supportive attitudes toward family planning, higher community literacy levels, and an agrarian way of life were identified as community-level predictors of modern contraceptive utilization.

Supporting information

Acknowledgments

We have received the data from the Performance Monitoring for Action (PMA-Ethiopia), and we are grateful to PMA for providing us with this valuable data.