Unveiling 10‐Year Dynamics of Pregnancy Termination Across Sub‐Saharan Africa: A Multilevel Study

Abstract

ABSTRACT

Background and Aims

Sub‐Saharan African (SSA) women face significant reproductive health challenges, including unwanted pregnancies and unsafe abortions. Despite the high prevalence of pregnancy termination in the continent, multilevel determinants and recent data trends remain understudied. This study addresses these gaps by leveraging recent Demographic and Health Survey (DHS) data and advanced statistical techniques.

Methods

This study involves secondary analysis using DHS data collected between 2010 and 2021 from 33 countries in Sub‐Saharan Africa. The analysis focused on pregnancy termination among women aged 15–49, comprising a weighted sample of 470,330 individuals. The data underwent a weighting process, considering sampling weight, primary sampling units, and strata. We utilized a multilevel binary logistic regression model to evaluate the correlation between individual and community‐level variables and the probability of pregnancy termination. Given the nested structure of the models, comparisons were made using the deviance statistic (−2 log‐likelihood ratio). All analyses were performed using STATA version 17. Variables with a p‐value ≤ 0.2 in the bivariable multilevel analysis were included in the multivariable model. The final results are presented as adjusted odds ratios (AORs) with 95% confidence intervals (CIs) to indicate the strength and statistical significance of associations.

Results

The overall prevalence of pregnancy termination among reproductive‐age women in SSA was 6.96% (95% CI: 6.89%, 7.03%), with the highest (13.59%) and lowest (3.83%) prevalence reported in Ghana and Ethiopia, respectively. The odds of pregnancy termination among rural resident women were 8% lower (AOR = 0.92, 95% CI: 0.89, 0.95) compared to urban residents. Women in Southern Africa had 9% decreased odds of pregnancy termination than women in Eastern Africa. Compared to women from East Africa, women in Western and Central Africa were 1.21 (AOR = 1.21, 95% CI: 1.18, 1.25) and 1.40 (AOR = 1.40, 95% CI: 1.35, 1.44) times higher odds of pregnancy termination, respectively.

Conclusion

The study reveals a notably high rate of pregnancy termination in SSA, which is particularly worrisome due to the legal limitations on abortion services in many SSA countries. Expanding access to contraception and comprehensive sexual health education is crucial to reducing unintended pregnancies across the region. Additionally, a reassessment of the strictness of abortion service restrictions is critical to encourage women to obtain these services from qualified professionals.

Background

Globally, an estimated 73 million abortions occur annually, with over half resulting from unintended pregnancies, representing nearly one-third of all pregnancies. In the period between 2010 and 2014, almost 44% of all global pregnancies were unplanned, equivalent to 62 unintended pregnancies [1] per 1000 women aged between 15 and 44. Termination of pregnancy—whether spontaneous or induced—often occurs in unsafe conditions in low- and middle-income countries (LMICs), where access to care is limited and abortion laws are restrictive [1, 2]. A staggering 97% of all fatal pregnancy terminations happen in LMICs. Over half of unsafe abortion procedures occur in Asia, particularly in the southern and central areas. In Latin America and Africa, approximately 75% of abortions are conducted under unsafe conditions, with nearly half in Africa occurring in especially high-risk settings [3–5].

An estimated 4.7%–13.2% of maternal deaths annually are attributed to unsafe abortions, particularly in low-resource settings [6]. The mortality rate from such unsafe procedures is substantially higher in developing regions, with a rate of 220 deaths per 100,000 unsafe pregnancy terminations, in contrast to the rate of 30 deaths per 100,000 unsafe terminations in developed regions [5]. It was estimated that, in 2012, seven million women in developing countries were hospitalized due to complications arising from unsafe abortions [7]. Health systems in developing nations spend approximately $553 million each year on treating complications from unsafe abortions, with households losing a total of $922 million due to long-term disabilities caused by abortions [8].

In Sub-Saharan Africa (SSA), unintended pregnancies and unsafe abortion practices continue to pose major reproductive health concerns [9–11]. Various factors, such as marital status, age, religious beliefs, educational attainment, economic status, ethnicity, employment status, and familial factors, have been identified as being associated with pregnancy termination in SSA [12–21]. However, the prevalence of unsafe abortions has been attributed to multiple factors including poverty, social inequality, and denial of women's human rights [22]. Restrictive abortion laws disproportionately endanger the health of young, low-income, and less-educated women, especially where access to safe procedures is limited [23, 24]. Unsafe abortions are often carried out using various risky methods such as oral or intravenous administration, vaginal preparations, insertion of foreign bodies into the uterus, and traumatic abdominal methods [25]. In LMICs, abortion practices are shaped by deep socioeconomic and healthcare disparities. Poverty, limited access to quality reproductive health services, restrictive abortion laws, and pervasive sociocultural stigma all converge to create an environment where unsafe abortion is a leading cause of maternal mortality. In these settings, women from marginalized communities are disproportionately affected, often resorting to unsafe methods due to a lack of access to skilled providers or fear of legal repercussions [5, 9, 24]. Furthermore, health system inadequacies, including shortages of trained personnel and essential supplies, exacerbate the risks associated with pregnancy termination, contributing to adverse outcomes at both individual and population levels.

Ba et al. [26] recently explored the determinants of pregnancy termination among women of reproductive age across 36 LMICs. However, several limitations were identified in the study, including the use of binary logistic regression without accounting for the hierarchical nature of Demographic and Health Survey (DHS) data, the failure to incorporate recent DHS data from certain countries such as Liberia, and the usage of the broad outcome variable “ever had termination of pregnancy.” To address these limitations, this study applies multilevel logistic modeling with random effects, drawing on updated DHS data covering 2010–2021, and employing the outcome variable “pregnancy termination that occurred 5 years before the survey.” This variable is considered to have notable implications for clinical and statistical outcomes and will likely provide a more up-to-date understanding of the factors connected to pregnancy termination in the region.

In addition, the initial target of Sustainable Development Goal 3 is to reduce the global maternal mortality ratio to fewer than 70 deaths per 100,000 live births by the year 2030 [27]. To help achieve this objective and develop interventions tailored to the factors contributing to and facilitating pregnancy termination in SSA, this study aims to offer a comprehensive understanding of the current factors associated with pregnancy terminations in this region. This study primarily seeks to assess the prevalence of pregnancy termination, with a secondary objective of identifying its associated factors. The findings are expected to inform the revision of current laws and policies related to pregnancy termination in the participating countries. Notably, policy initiatives that enhance access to contraception and reduce the stigma surrounding abortion—such as those implemented in Rwanda—have been linked to declines in maternal mortality rates.

Methods

Data Source

This study utilized data from DHS conducted between 2010 and 2021 in 33 SSA nations. These nationally representative, cross-sectional surveys are widely acknowledged for their methodological rigor in collecting demographic and health-related information. Only countries with DHS datasets available from 2010 onward were included. The analysis was based on data extracted from the Birth Record (BR) files, resulting in a weighted sample of 470,330 women. To enhance representativeness and account for the survey design, weighting adjustments were applied using variables for sampling weights (v005), clusters (v021), and strata (v023). The DHS employs a stratified two-stage cluster sampling strategy, using the most recent national census as a sampling frame. Regions were stratified into urban and rural areas. In the first stage, enumeration areas (EAs) were selected proportional to their size; in the second stage, households within those EAs were chosen systematically. Further details on the sampling methodology are provided in the respective DHS country reports. Country-specific sample sizes ranged from 5323 in Comoros to 41,821 in Nigeria (Table 1).

Table 1 Number of study participants included by country.

Country (year)	Sample (470,330)	Percentage
Angola (2015/2016)	14,379	3.06
Benin (2017/2018)	15,928	3.39
Burundi (2016/2017)	17,269	3.67
Burkina Faso (2010)	17,084	3.63
Cameroon (2018)	13,616	2.89
Chad (2014/2015)	17,683	3.76
Comoros (2012)	5323	1.13
Congo (2011/2012)	10,819	2.30
Congo DR (2013)	18,822	4.00
Ethiopia (2016)	15,683	3.33
Gabon (2012)	16,831	3.58
Gambia (2019/2020)	11,865	2.52
Ghana (2014)	9396	2.00
Guinea (2018)	10,874	2.31
Kenya (2014)	14,623	3.11
Lesotho (2014)	6621	1.41
Liberia (2019/2020)	8065	1.71
Madagascar (2021)	18,869	4.01
Malawi (2015/2016)	24,562	5.22
Mali (2018)	10,519	2.24
Cote d'Ivoire (2021)	10,052	2.14
Namibia (2013)	9173	1.95
Niger (2012)	11,159	2.37
Nigeria (2018)	41,821	8.89
Rwanda (2019/2020)	14,634	3.31
Senegal (2010/2011)	15,688	3.34
Sierra Leone (2019)	15,574	3.11
Tanzania (2015/2016)	13,265	2.82
Uganda (2016)	18,506	3.93
South Africa (2016)	8514	1.81
Togo (2013/2014)	9474	2.01
Zambia (2018)	13,683	2.91
Zimbabwe (2015)	9955	2.12

Study Variables

Outcome Variable

Pregnancy termination status was the outcome variable. Reproductive-age who had a history of pregnancy termination was labeled as “Yes” and those who had no history of pregnancy termination were labeled as “No”. In DHS, this was asked as ever having termination of pregnancy and those who had history of pregnancy termination were asked the date of termination. Then, for our study, we have considered those who had a history of termination of pregnancy in the 5 years preceding the survey as having the event. Several studies have also used this variable in measuring pregnancy termination, while acknowledging that considering the question posed to the women, there is a tendency that some of the responses may be spontaneous termination of pregnancies though this proportion might be insignificant [28–31].

Independent Variables

Due to the hierarchical structure of the DHS data and the study's analytical goals, independent variables were categorized into two levels: individual-level (level-one) and community-level (level-two) factors. The individual-level variables included maternal age, marital status, educational attainment, household wealth status, employment, number of children ever born (parity), exposure to media (radio, newspapers, and television), contraceptive use, and the sex of the household head. At the community level, the analysis considered place of residence (urban or rural) and SSA region.

Data Management and Analysis

Data processing and statistical analyses were conducted using STATA version 17 and R version 4.3. To account for the complex survey design and adjust for nonresponse, the data set was weighted using the DHS-provided variables: sampling weight (v005), primary sampling unit (v021), and stratification variable (v023). These adjustments ensured that the findings were nationally representative. All reported estimates in this study are based on weighted data. The prevalence of pregnancy termination was presented alongside 95% confidence intervals, and data visualization was supported by tables and figures.

Given the hierarchical structure of DHS data—where individual respondents are nested within sampling clusters—there is a likelihood of shared characteristics among women within the same cluster. This clustering violates the independence assumption underlying standard logistic regression, necessitating the use of advanced statistical techniques to account for intra-cluster correlation. To address this, we employed a multilevel binary logistic regression model to explore associations between individual- and community-level predictors and the likelihood of pregnancy termination.

To identify statistically significant factors, a series of multilevel logistic models were fitted. This approach allowed for accurate estimation while addressing the hierarchical nature of the data. A total of four models were compared using the deviance statistic (−2 log-likelihood). We evaluated the extent of clustering and model performance using indicators such as the likelihood ratio (LR) test, the intra-cluster correlation coefficient (ICC), and the median odds ratio (MOR) [32, 33].

ICC = $$\partial $$²/($$\partial $$² + π²/3) [29], but captures the degree of variability in outcomes between clusters and is defined as the median value of the odds ratio when comparing a randomly selected cluster with a higher risk of pregnancy termination to one with a lower risk [30]. $$\text{MOR}\,=\text{exp}\,(\sqrt{2\,\ast \,{\partial }^{2}\,\ast \,0.6745})\unicode{x0007E}\text{MOR}=\text{exp}\,(0.95\,\ast \,\partial ),$$

$$\partial $$² indicates that cluster variance.

To assess model performance, four hierarchical logistic regression models were estimated. These included [1]: an empty model with no predictors to assess the extent of variability in pregnancy termination across clusters [3]; a model incorporating only individual-level (level-one) variables [2]; a model including only community-level (level-two) variables; and [4] a full model that combined both individual and community-level factors. The model with the lowest deviance value was selected as the best-fitting model.

Before the multivariable analysis, bivariate analyses were conducted to identify candidate variables. Those with a p-value ≤ 0.2 were included in the multivariable model. In the final analysis, adjusted odds ratios (AORs) along with 95% CIs were reported to determine the strength and significance of associations. Multicollinearity was assessed using the variance inflation factor (VIF), and all included variables had VIFs below 5 and tolerances above 0.1, indicating no significant multicollinearity issues.

Ethical Consideration

For this study, permission to use the data was obtained through an official authorization letter from the DHS Program. As the DHS datasets are publicly available and fully deidentified, additional ethical approval was not required.

Results

Overall Prevalence of Pregnancy Termination in SSA

The overall prevalence of pregnancy termination among women aged 15–49 years in SSA was 6.96% (95% CI: 6.89%, 7.03%). This equates to roughly one out of every 14 women reporting a pregnancy termination within the last 5 years. The prevalence varied widely between countries, with Ghana reporting the highest prevalence at 13.59% and Ethiopia reporting the lowest prevalence at 3.83% (Figure 1).

[IMAGE OMITTED. SEE PDF]

Regional Differences

The prevalence of pregnancy termination exhibited marked regional disparities. Among the 32,723 terminations reported, Western Africa accounted for the largest share (42.29%), while Southern Africa had the lowest proportion (3.08%). Urban women reported a higher proportion of pregnancy terminations (54.21%) compared to rural women (45.79%). These regional trends are detailed in Table 2, which also highlights variations across other demographic characteristics such as maternal age, education level, and wealth status.

Table 2 Descriptive characteristics of women aged 15–49 in sub-Saharan African countries and those who had terminated a pregnancy in the five years before the survey.

Variables	Number of women who gave birth in the last 5 years (N = 470,330)	Women who reported having a termination of pregnancy in the last 5 years (N = 32,723)
Residence (n = 470,330)
Rural	192,283 (40.88)	14,984 (45.79)
Urban	278,047 (59.12)	17,739 (54.21)
Sub-Saharan Africa region (n = 470,330)
Eastern Africa	166,373 (35.37)	9739 (29.76)
Southern Africa	24,307 (5.17)	1008 (3.08)
Western Africa	187,499 (39.87)	13,839 (42.29)
Central Africa	92,151 (19.6)	8137 (24.87)
Maternal age (in years) (n = 470,330)
15–24	186,982 (39.76)	9454 (28.89)
25–34	149,395 (31.76)	14,006 (42.80)
≥ 35	133,953 (28.48)	9263 (28.31)
Maternal education (n = 470,330)
No education	142,856 (30.37)	9648 (29.48)
Primary	142,906 (30.38)	9857 (30.12)
Secondary	159,458 (33.90)	11,209 (34.25)
Higher	25,110 (5.34)	2009 (6.14)
Household wealth status (n = 470,330)
Poorest	81,608 (17.35)	5308 (16.22)
Poorer	86,904 (18.48)	5881 (17.97)
Middle	90,793 (19.30)	6239 (19.06)
Richer	99,337 (21.12)	7021 (21.45)
Richest	111,688 (23.75)	8275 (25.29)
Reading newspaper (n = 470,330)
No	367,845 (78.21)	25,246 (77.15)
Yes	102,485 (21.79)	7477 (22.85)
Listening to radio (n = 470,330)
No	195,158 (41.49)	12,267 (37.49)
Yes	275,172 (58.51)	20,456 (62.51)
Watching television (n = 470,330)
No	255,244 (54.27)	16,002 (48.90)
Yes	215,086 (45.73)	16,721 (51.10)
Parity (n = 470,330)
Nulliparous	128,733 (27.37)	4944 (15.11)
Primiparous	68,113 (14.48)	6566 (20.07)
Multiparous	160,597 (34.15)	13,494 (41.24)
Grand multiparous	112,887 (24.00)	7718 (23.59)
Perceived distance to health facility (N = 454,269)
Not a big problem	286,621 (63.09)	20,396 (64.10)
A big problem	167,648 (36.91)	11,425 (35.90)
Maternal working status (n = 470,330)
No	190,285 (40.46)	11,015 (33.66)
Yes	280,045 (59.34)	21,708 (66.34)
Sex of household head (n = 470,330)
Male	342,736 (72.87)	24,879 (76.03)
Female	127,594 (27.13)	7844 (23.97)
Marital status (n = 470,330)
Never in union	133,681 (28.42)	3404 (10.40)
Married	296,201 (62.98)	26,501 (80.99)
Divorced, windowed, separated	40,448 (8.60)	2818 (8.61)
Contraceptive use (n = 470,330)
No	104,247 (22.16)	7356 (22.48)
Yes	366,083 (77.84)	25,368 (77.52)
Healthcare decision-making autonomy (n = 470,330)
Respondent alone	52,419 (11.14)	4807 (14.69)
Jointly with husband/partner	110,980 (23.60)	9788 (29.91)
Othersa	306,931 (65.26)	18,128 (55.40)

Individual-Level Factors Associated With Pregnancy Termination

Maternal Age

Maternal age was found to be a significant predictor of pregnancy termination. Women aged 25–34 had 1.37 times greater odds of reporting a termination compared to those aged 15–24 (AOR = 1.37, 95% CI: 1.33–1.42). Those aged 35 and above also had elevated odds (AOR = 1.08, 95% CI: 1.03–1.12). These higher odds among older women may reflect factors such as completed desired family size, financial limitations, or health-related risks.

Education

Educational attainment was strongly associated with pregnancy termination. Compared to women with no formal education, those with primary, secondary, and higher education had increased likelihoods of termination, with AORs of 1.19 (95% CI: 1.15–1.27), 1.26 (95% CI: 1.22–1.31), and 1.23 (95% CI: 1.15–1.31), respectively.

Household Wealth Status

Women from wealthier households were less likely to report pregnancy termination compared to those from the poorest households. Specifically, women from middle-income, richer, and richest households had 5% (AOR = 0.95, 95% CI: 0.92, 0.99), 7% (AOR = 0.93, 95% CI: 0.89, 0.97), and 12% (AOR = 0.88, 95% CI: 0.84, 0.92) lower odds, respectively.

Media Exposure

Media exposure, including radio and television, was linked to an increased likelihood of pregnancy termination. Women who reported listening to the radio had 1.17 times higher odds (AOR = 1.17, 95% CI: 1.15–1.21), while those who watched television had 1.20 times higher odds (AOR = 1.20, 95% CI: 1.16–1.23), compared to those with no such exposure.

Parity

Parity showed a nonlinear association with pregnancy termination. Primiparous women had higher odds of termination (AOR = 1.16, 95% CI: 1.11, 1.22) compared to nulliparous women. However, multiparous and grand multiparous women had reduced odds (AOR = 0.80, 95% CI: 0.77, 0.84 and AOR = 0.70, 95% CI: 0.66, 0.74, respectively).

Community-Level Determinants of Pregnancy Termination

Residence

Residence was a significant community-level determinant, with rural women being 8% less likely to terminate pregnancies than urban women (AOR = 0.92, 95% CI: 0.89, 0.95).

Region

The odds of pregnancy termination varied significantly across regions. Compared to Eastern Africa, women in Western Africa had 1.21 times higher odds (AOR = 1.21, 95% CI: 1.18, 1.25), and those in Central Africa had 1.40 times higher odds (AOR = 1.40, 95% CI: 1.35, 1.44). Conversely, women in Southern Africa had 27% lower odds (AOR = 0.73, 95% CI: 0.68, 0.78).

Random Effect Analysis Results

The random effect analysis revealed significant clustering effects in pregnancy termination outcomes. The null model's MOR of 1.14 indicated notable between-cluster variation. The final model, which included both individual- and community-level variables, demonstrated the best fit based on deviance values and likelihood ratio tests (Table 3).

Table 3 Random effect analysis results and model comparison parameters.

Parameters	Null model	Model 1	Model 2	Model 3
Cluster level variance	0.02	0.019	0.017	0.017
ICC	< 10%	< 10%	< 10%	< 10%
MOR	1.14	1.13	1.12	1.12
LLR	−115,971.3	−111,119.7	−115,506.8	−110,810.8
Deviance (−2LLR)	231,942.6	222,239.4	231,013.6	221,621.6
LR test	< 0.001	< 0.001	< 0.001	< 0.001

Summary of Factors Associated With Pregnancy Termination

The multivariable analysis revealed that individual factors such as age, education, wealth status, and parity, alongside community-level factors like residence and region, play significant roles in influencing pregnancy termination. These findings highlight the complex interplay between personal and environmental determinants and underscore the need for targeted interventions addressing both individual and systemic factors (Table 4).

Table 4 Factors associated with pregnancy termination in SSA.

Variable	Null model (model with no independent variable)	Model with level-one variables (AOR with 95% CI)	Model with level-two variables (AOR with 95% CI)	Model with level-one and -two variables (AOR with 95% CI)
Residence
Rural			0.91 (0.89, 0.93)	0.92 (0.89, 0.95)*
Urban			1	1
Sub-Saharan Africa region
Eastern Africa			1	1
Southern Africa			0.91 (0.60, 0.69)	0.73 (0.68, 0.78)**
Western Africa			1.21 (1.18, 1.25)	1.18 (1.15, 1.22)*
Central Africa			1.40 (1.35, 1.44)	1.41 (1.36, 1.46)**
Maternal age (in years)
15–24		1		1
25–34		1.35 (1.31, 1.40)		1.37 (1.33, 1.42)**
≥ 35		1.05 (1.01, 1.09)		1.08 (1.03, 1.12)**
Maternal education
No education		1		1
Primary		1.16 (1.13, 1.20)		1.19 (1.15, 1.27)**
Secondary		1.29 (1.25, 1.34)		1.26 (1.22, 1.31)**
Higher		1.24 (1.16, 1.32)		1.23 (1.15, 1.31)**
Household wealth status
Poorest		1		1
Poorer		1.01 (0.97, 1.04)		1.01 (0.96, 1.04)
Middle		0.96 (0.93, 0.99)		0.95 (0.92, 0.99)**
Richer		0.95 (0.91, 0.98)		0.93 (0.89, 0.97)**
Richest		0.90 (0.87, 0.94)		0.88 (0.84, 0.92)**
Reading newspaper
No		1		1
Yes		1.00 (0.97, 1.04)		1.03 (0.91, 1.07)
Listening to radio
No		1		1
Yes		1.12 (1.09, 1.15)		1.17 (1.15, 1.21)**
Watching television
No		1		1
Yes		1.28 (1.24, 1.32)		1.20 (1.16, 1.23)**
Parity
Nulliparous		1		1
Primiparous		1.17 (1.12, 1.23)		1.16 (1.11, 1.22)**
Multiparous		0.82 (0.79, 0.86)		0.80 (0.77, 0.84)**
Grand multiparous		0.74 (0.70, 0.78)		0.70 (0.66, 0.74)**
Maternal working status
No		1		1
Yes		1.18 (1.15, 1.25)		1.15 (1.12, 1.18)*
Sex of household head
Male		1		1
Female		1.04 (1.00, 1.07)		1.04 (1.01, 1.08)*
Marital status
Never in union		1		1
Married		4.90 (4.66, 5.16)		4.65 (4.41, 4.90)**
Divorced, widowed, separated		3.35 (3.15, 3.56)		3.30 (3.10, 3.52)**
Use contraceptive
No		1.21 (1.17, 1.24)		1.11 (1.08, 1.14)**
Yes		1		1
Health care decision-making autonomy
Respondent alone		1		1
Jointly with husband/partner		1.00 (0.96, 1.03)		0.97 (0.94, 1.01)
Othersa		1.05 (1.01, 1.09)		0.98 (0.94, 1.02)

Discussion

This study explores the prevalence and contributing factors of pregnancy termination in SSA, enhancing our understanding of this issue based on previous studies from Ethiopia, Nigeria, and other East African countries [2, 34, 35]. Building upon an earlier study [26], which focused on data up to 2018, this study extends the knowledge base to 2021. Additionally, while previous studies presented evidence at the regional level, this study advances the analysis by examining the phenomenon at the sub-regional level (East, West, Central, and Southern Africa). Our study reports a 6.96% prevalence rate of pregnancy termination, a modest increase from the 5% reported in earlier studies in the same region but still lower than the 7.79% seen in East African countries [2, 36]. Despite differences in legal frameworks regarding pregnancy termination across SSA, a shared societal stigma around unplanned pregnancies remains, particularly those occurring outside of marriage. This stigma often leads unmarried, sexually active young people in the region to face the difficult challenge of managing unplanned pregnancies [37].

Pregnancy termination has far-reaching consequences beyond the individual. At a personal level, the procedure can affect reproductive health, increasing the risk of preterm birth, infertility, and mental health issues [38, 39]. Additionally, the financial burden is significant, as families and governments must spend resources on treating fertility complications or hormonal changes resulting from unsafe termination methods [40]. Therefore, pregnancy termination should be carefully considered, and women who do not intend to have children must use reliable contraceptives that ensure maximum protection. Governments and nongovernmental organizations in the region must focus on educating women about the consequences of pregnancy termination and the importance of contraception when they do not wish to conceive.

Furthermore, societal responses to unplanned pregnancies may be shaped by the restrictive abortion laws prevalent in many SSA countries. Severe restrictions on abortion services, resulting in limited or no access to legal abortion options, contribute to unsafe termination practices. The incidence of pregnancy termination in SSA, as observed in this and previous studies, is closely linked to these legal restrictions [2, 21, 34, 36]. Such restrictions often lead to reproductive health complications or fatalities due to the use of unsafe, unregulated abortion services or crude termination methods [24, 41].

Our study found significant correlations between pregnancy termination and various individual-level factors, including marital status, maternal age, household wealth, education level, media exposure (radio and television), gender of the household head, contraceptive use, and number of children. Community-level characteristics, including place of residence (urban vs. rural) and geographic region, also influenced pregnancy termination. Additionally, married women had a higher likelihood of reporting a termination compared to those who were divorced, widowed, separated, or never married. This finding is consistent with results from East Asian countries, where married women have also been found to be more likely to terminate pregnancies than unmarried women [2]. Given that the data in this study, as well as in previous studies, is self-reported, the societal stigma surrounding out-of-wedlock pregnancies may contribute to underreporting of terminations among unmarried women. It is also plausible that unmarried young women may be less sexually active, thus reducing the likelihood of pregnancy. Moreover, prevailing socio-cultural norms that emphasize premarital abstinence may prompt unmarried women to take proactive measures to avoid pregnancy [37, 42].

Our analysis further revealed that women aged 25–34 were nearly twice as likely to report pregnancy termination compared to women aged 15–24. Similarly, women aged 35 and older had 1.08 times higher odds of reporting pregnancy termination than their younger counterparts. These findings align with previous research suggesting higher termination rates among older women compared to younger ones in East African countries [2]. This difference may be attributed to the fact that younger individuals, especially adolescents and young adults, might be less sexually active or less confident in reporting pregnancy or termination due to societal stigma. Additionally, adolescents from conservative or traditional cultures often place greater emphasis on preventing pregnancy during premarital sexual activities than on avoiding sexually transmitted infections [37]. As a result, socio-cultural expectations of premarital abstinence may reduce sexual activity or discourage reporting of pregnancy termination among young women [42]. The underreporting of pregnancy termination in this age group highlights the need for a reassessment of socio-cultural norms and underscores the importance of providing care and support to help young women navigate the challenges associated with this developmental stage.

Contrary to some earlier studies, our analysis revealed that women with formal education were more likely to report pregnancy termination than those without any formal education [2]. This pattern aligns with findings from a study conducted in Nepal [43]. Although this may appear unexpected—given that educated women generally have greater access to information and services related to contraception—it is plausible that women with higher education levels may prioritize career commitments, influencing their reproductive decisions. Such career-oriented women might be more inclined to terminate unplanned pregnancies to maintain their professional development. Our study also aligns with previous research showing that working women are more likely to report pregnancy termination than nonworking women [2].

Similar to studies conducted in East African countries, our findings indicate that women exposed to media were more likely to report pregnancy termination than those with no media exposure [2, 44]. This presents a nuanced situation, as media exposure is generally linked to improved awareness of contraceptive methods, which would typically be expected to lower the incidence of pregnancy termination. However, women in urban areas, who have greater access to media, also reported higher rates of pregnancy termination, suggesting that the type of health-related content being disseminated through these platforms may influence reproductive behaviors. Given these findings, we recommend that mass media platforms prioritize content that promotes contraceptive use over termination, particularly in light of the legal restrictions on abortion services in SSA, which often force women to resort to unsafe, unregulated services, putting their health at risk.

The analysis also revealed that women residing in female-headed households had a higher likelihood of reporting pregnancy termination than those in male-headed households. This finding underscores the importance of decision-making autonomy in women's reproductive health. The implications of this observation suggest that increasing women's autonomy in health-related decisions could lead to greater use of contraceptives, ultimately reducing the need for pregnancy termination in contexts where access to safe abortion services is limited.

Consistent with previous research, we observed that women who had given birth multiple times were less likely to report pregnancy termination than women who had never given birth [2, 44]. This result is logical, as women with multiple children may be less inclined to terminate pregnancies, while those who have never given birth may be unmarried and younger—two intersecting factors that could increase the likelihood of pregnancy termination due to societal stigma surrounding out-of-wedlock pregnancies [37]. However, the lack of sufficient contraceptive services for sexually active, unmarried adolescents and young women in SSA is a cause for concern. Without access to safe abortion services, these young women may resort to unsafe methods of termination, which pose significant risks to their reproductive health and lives [41, 45].

In summary, our study confirms previous findings that women who have had multiple children are less likely to report pregnancy termination than those who have never given birth [2, 44]. This observation is understandable, as a decreased inclination toward terminating pregnancies may lead to multiple pregnancies. It is also plausible that women who have never given birth may be younger and unmarried, factors that can increase the likelihood of pregnancy termination due to societal stigma associated with unplanned pregnancies outside marriage [37]. The limited availability of contraceptive services for sexually active, unmarried young women in SSA, where safe abortion access is restricted, further exacerbates this issue. Young women facing unplanned pregnancies may resort to unsafe and potentially dangerous termination methods, endangering their reproductive health and lives [41, 45].

Strengths and Limitations

This study leverages a comprehensive and high-quality data set derived from the DHS, a globally recognized database renowned for its rigorous data collection methods and standardized protocols across countries. The use of DHS data ensures robust representativeness, allowing for nuanced analysis of pregnancy termination at both regional and sub-regional levels across SSA. Moreover, advanced statistical techniques, including multilevel modeling, have been employed to enhance the reliability and validity of the findings by accounting for individual- and community-level factors.

However, certain limitations must be acknowledged. The reliance on self-reported data introduces the possibility of social desirability bias, where participants may underreport sensitive behaviors such as pregnancy termination due to societal norms or stigma. This limitation is particularly relevant in contexts where unplanned pregnancies or terminations are heavily stigmatized, potentially leading to underestimation of prevalence rates. Additionally, recall bias may influence data accuracy, as participants were required to recall events that occurred up to 5 years before the survey. These biases, inherent in the study design, underscore the challenges of investigating sensitive reproductive health issues in such settings.

Despite these limitations, the study provides valuable insights into the factors influencing pregnancy termination and highlights the unmet needs in sexual and reproductive health services across SSA. It contributes significantly to the body of evidence necessary for informing policy and programmatic interventions. The findings emphasize the importance of improving access to and quality of contraceptive services while advocating for a reconsideration of restrictive pregnancy termination policies. By addressing these gaps, the study seeks to bolster the sexual and reproductive health of women, particularly adolescent girls and young women, who face disproportionate risks in contexts of limited resources and entrenched societal challenges.

Conclusion

To summarize, this study reveals an alarmingly high rate of pregnancy termination in SSA, a problem exacerbated by severe legislative restrictions on abortion providers in many SSA regions. The observed high rate of abortion reflects a profoundly ingrained reproductive health condition that requires prompt and comprehensive response. The cornerstone of these initiatives should be the major improvement of contraceptive services and the reinforcement of sexual and reproductive health education. Such preemptive measures are critical in reducing unwanted pregnancies in these areas.

Furthermore, this study emphasizes the crucial need for a full evaluation of the rigidity connected with abortion service laws. A reevaluation could create a climate in which women feel safe and encouraged to seek these services from certified medical experts, lowering the possibility of unsafe abortion practices. The current legislative climate, which provides restricted access to safe abortion services, has the potential to push women into risky practices, putting their health at risk. As a result, an early and comprehensive plan that addresses these obstacles while also advocating for reproductive freedom is critical for the protection of women's health in SSA.

Author Contributions

Getayeneh Antehunegn Tesema: data curation, formal analysis, investigation, visualization, writing – original draft, writing – review and editing. Sylvester R. Okeke: data curation, formal analysis, writing – original draft, writing – review and editing. Michael Sarfo: formal analysis, writing – original draft, writing – review and editing. Edward K. Ameyaw: data curation, formal analysis, investigation, writing – original draft, writing – review and editing. Olanrewaju Oladimeji: validation, writing – original draft, writing – review and editing. Sanni Yaya: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, supervision, validation, writing – original draft, writing – review and editing.

Acknowledgments

The authors thank the MEASURE DHS project for their support and for free access to the original data.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data for this study were sourced from Demographic and Health surveys (DHS) and available here: .

Transparency Statement

The lead author Sanni Yaya affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

References

G. Sedgh, J. Bearak, S. Singh, et al., “Abortion Incidence Between 1990 and 2014: Global, Regional, and Subregional Levels and Trends,” Lancet 388, no. 10041 (July 2016): 258–267.

S. Hailegebreal, E. B. Enyew, A. E. Simegn, et al., “Pooled Prevalence and Associated Factors of Pregnancy Termination Among Youth Aged 15–24 Year Women in East Africa: Multilevel Level Analysis,” PLoS One 17, no. 12 (December 2022): e0275349.

J. Bearak, A. Popinchalk, B. Ganatra, et al., “Unintended Pregnancy and Abortion by Income, Region, and the Legal Status of Abortion: Estimates From a Comprehensive Model for 1990–2019,” Lancet Global Health 8, no. 9 (September 2020): e1152–e1161.

World Health Organization, “World Health Organization Fact Sheet Details Abortion,” 2021, https://www.who.int/news-room/fact-sheets/detail/abortion.

B. Ganatra, C. Gerdts, C. Rossier, et al., “Global, Regional, and Subregional Classification of Abortions by Safety, 2010–2014: Estimates From a Bayesian Hierarchical Model,” Lancet 390, no. 10110 (November 2017): 2372–2381.

L. Say, D. Chou, A. Gemmill, et al., “Global Causes of Maternal Death: A WHO Systematic Analysis,” Lancet Global Health 2, no. 6 (June 2014): e323–e333.

S. Singh and I. Maddow‐Zimet, “Facility‐Based Treatment for Medical Complications Resulting From Unsafe Pregnancy Termination in the Developing World, 2012: A Review of Evidence From 26 Countries,” BJOG: An International Journal of Obstetrics & Gynaecology 123, no. 9 (August 2016): 1489–1498.

M. Soleimani Movahed, S. Husseini Barghazan, F. Askari, and M. Arab Zozani, “The Economic Burden of Abortion and Its Complication Treatment Cares: A Systematic Review,” Journal of Family & Reproductive Health 14, no. 2 (June 2020): 60–67.

A. Bankole, S. Singh, and T. Haas, “Characteristics of Women Who Obtain Induced Abortion: A Worldwide Review,” International Family Planning Perspectives 25, no. 2 (June 1999): 68.

I. F. Adewole, N. C. Orji, and B. A. Oye‐Adeniran, “Unsafe Abortion and Development: A Strategic Approach,” IDS Bulletin 39, no. 3 (July 2008): 47–54.

F. Okonofua, “Adolescent Reproductive Health in Africa: Future Challenges/Santé reproductive de l'adolescent en Afrique: Défis de l'avenir,” African Journal of Reproductive Health 4, no. 1 (April 2000): 7.

B. O. Ahinkorah, A. A. Seidu, J. E. Hagan, et al., “Predictors of Pregnancy Termination Among Young Women in Ghana: Empirical Evidence From the 2014 Demographic and Health Survey Data,” Healthcare 9, no. 6 (June 2021): 705.

K. S. Dickson, K. S. Adde, and B. O. Ahinkorah, “Socio – Economic Determinants of Abortion Among Women in Mozambique and Ghana: Evidence From Demographic and Health Survey,” Archives of Public Health 76, no. 1 (December 2018): 37.

G. Gilano and S. Hailegebreal, “Determinants of Abortion Among Youth 15‐24 in Ethiopia: A Multilevel Analysis Based on EDHS 2016,” PLoS One 16, no. 3 (2021): e0248228.

A. Weitzman, J. S. Barber, Y. Kusunoki, and P. England, “Desire for and to Avoid Pregnancy During the Transition to Adulthood,” Journal of Marriage and Family 79, no. 4 (August 2017): 1060–1075.

J. I. B. Adinma, E. D. Adinma, L. Ikeako, and C. Ezeama, “Abortion Treatment by Health Professionals in South‐Eastern Nigeria,” Journal of Obstetrics and Gynaecology 31, no. 6 (August 2011): 529–532.

A. J. Taft and L. F. Watson, “Depression and Termination of Pregnancy (Induced Abortion) in a National Cohort of Young Australian Women: The Confounding Effect of Women's Experience of Violence,” BMC Public Health 8, no. 1 (December 2008): 75.

Y. Mekonnen, D. S. Telake, and E. Wolde, “Adolescent Childbearing Trends and Sub‐National Variations in Ethiopia: A Pooled Analysis of Data From Six Surveys,” BMC Pregnancy and Childbirth 18, no. 1 (December 2018): 276.

A. Oyefabi, A. Nmadu, and M. Yusuf, “Prevalence, Perceptions, Consequences, and Determinants of Induced Abortion Among Students of the Kaduna State University, Northwestern Nigeria,” Journal of Medicine in the Tropics 18, no. 2 (2016): 86.

G. Adjei, Y. Enuameh, K. P. Asante, et al., “Predictors of Abortions in Rural Ghana: A Cross‐Sectional Study,” BMC Public Health 15, no. 1 (December 2015): 202.

F. I. Onukwugha, M. A. Magadi, A. M. Sarki, and L. Smith, “Trends in and Predictors of Pregnancy Termination Among 15–24 Year‐Old Women in Nigeria: A Multi‐Level Analysis of Demographic and Health Surveys 2003–2018,” BMC Pregnancy and Childbirth 20, no. 1 (December 2020): 550.

U. Rehnström Loi, K. Gemzell‐Danielsson, E. Faxelid, and M. Klingberg‐Allvin, “Health Care Providers’ Perceptions of and Attitudes Towards Induced Abortions in Sub‐Saharan Africa and Southeast Asia: A Systematic Literature Review of Qualitative and Quantitative Data,” BMC Public Health 15, no. 1 (December 2015): 139.

D. A. Grimes, J. Benson, S. Singh, et al., “Unsafe Abortion: The Preventable Pandemic,” Lancet 368, no. 9550 (November 2006): 1908–1919.

M. Gebremedhin, A. Semahegn, T. Usmael, and G. Tesfaye, “Unsafe Abortion and Associated Factors Among Reproductive Aged Women in Sub‐Saharan Africa: A Protocol for a Systematic Review and Meta‐Analysis,” Systematic Reviews 7, no. 1 (December 2018): 130.

R. Hussain, “Abortion and Unintended Pregnancy in Kenya,” Issues in Brief (Alan Guttmacher Institute) (2012): 1–4.

D. M. Ba, Y. Zhang, O. Pasha‐Razzak, et al., “Factors Associated With Pregnancy Termination in Women of Childbearing Age in 36 Low‐and Middle‐Income Countries,” PLOS Global Public Health 3, no. 2 (February 2023): e0001509.

United Nations General Assembly, Transforming Our World: The 2030 Agenda for Sustainable Development (United Nations General Assembly, 2015).

M. A. Lamina, “Prevalence of Abortion and Contraceptive Practice Among Women Seeking Repeat Induced Abortion in Western Nigeria,” Journal of Pregnancy 2015 (2015): 1–7.

J. K. Ganle, B. Obeng, J. Y. Yeboah, E. T. D. Tagoe‐Darko, and C. M. Mensah, “Disparities in Abortion Experience and Access to Safe Abortion Services in Ghana: Evidence From a Retrospective Survey,” African Journal of Reproductive Health 20, no. 2 (2016): 43–52.

P. G. C. Ilboudo, S. M. A. Somda, and J. Sundby, “Key Determinants of Induced Abortion in Women Seeking Postabortion Care in Hospital Facilities in Ouagadougou, Burkina Faso,” International Journal of Women's Health 6 (2014): 565.

K. S. Adde, K. S. Dickson, E. K. Ameyaw, and J. Amo‐Adjei, “Contraception Needs and Pregnancy Termination in Sub‐Saharan Africa: A Multilevel Analysis of Demographic and Health Survey Data,” Reproductive Health 18, no. 1 (December 2021): 177.

G. Rodríguez and I. Elo, “Intra‐Class Correlation in Random‐Effects Models for Binary Data,” Stata Journal: Promoting Communications on Statistics and Stata 3, no. 1 (2003): 32–46.

J. Merlo, B. Chaix, H. Ohlsson, et al., “A Brief Conceptual Tutorial of Multilevel Analysis in Social Epidemiology: Using Measures of Clustering in Multilevel Logistic Regression to Investigate Contextual Phenomena,” Journal of Epidemiology and Community Health 60, no. 4 (2006): 290–297.

G. Gilano and S. Hailegebreal, “Determinants of Abortion Among Youth 15–24 in Ethiopia: A Multilevel Analysis Based on EDHS 2016,” PLoS One 16, no. 3 (March 2021): e0248228.

F. I. Onukwugha, M. A. Magadi, A. M. Sarki, and L. Smith, “Trends in and Predictors of Pregnancy Termination Among 15–24 ‐Year‐Old Women in Nigeria: A Multi‐Level Analysis of Demographic and Health Surveys 2003–2018,” BMC Pregnancy and Childbirth 20, no. 1 (December 2020): 550.

B. O. Ahinkorah, “Socio‐Demographic Determinants of Pregnancy Termination Among Adolescent Girls and Young Women in Selected High Fertility Countries in Sub‐Saharan Africa,” BMC Pregnancy and Childbirth 21, no. 1 (December 2021): 598.

S. R. Okeke, “‘It Was Protected, Except, It Wasn't [With] a Condom’: A Mixed‐Methods Study of BBVs/STIs Protective Practices Among International University Students in Sydney, Australia,” BMC Public Health 22, no. 1 (November 2022): 2168.

J. M. Thorp, “Public Health Impact of Legal Termination of Pregnancy in the US: 40 Years Later,” Scientifica 2012 (October 2012): 1–16.

B. L. Scholten, G. C. M. L. Page‐Christiaens, A. Franx, C. W. P. M. Hukkelhoven, and M. P. H. Koster, “The Influence of Pregnancy Termination on the Outcome of Subsequent Pregnancies: A Retrospective Cohort Study,” BMJ Open 3, no. 5 (January 2013): e002803.

P. Bhide, P. Gund, and A. Kar, “Prevalence of Congenital Anomalies in an Indian Maternal Cohort: Healthcare, Prevention, and Surveillance Implications,” PLoS One 11, no. 11 (November 2016): e0166408.

A. Bankole, L. Remez, L. Owolabi, J. Philbin, and P. Williams, “From Unsafe to Safe Abortion in Sub‐Saharan Africa: Slow but Steady Progress,” Guttmacher Institute (2020).

S. R. Okeke, “How Perceived Australian Sexual Norms Shape Sexual Practices of East Asian and Sub‐Saharan African International Students in Sydney,” BMC Public Health 21, no. 1 (December 2021): 395.

A. Yogi and K. C. P. Neupane, “Prevalence and Factors Associated With Abortion and Unsafe Abortion in Nepal: A Nationwide Cross‐Sectional Study,” BMC Pregnancy and Childbirth 18 (December 2018): 1.

A. A. Seidu, B. O. Ahinkorah, E. K. Ameyaw, et al., “What Has Women's Reproductive Health Decision‐Making Capacity and Other Factors Got to Do With Pregnancy Termination in Sub‐Saharan Africa? Evidence From 27 Cross‐Sectional Surveys,” PLoS One 15, no. 7 (July 2020): e0235329.

M. Romero, R. Gomez Ponce de Leon, B. Carroli, H. Mehrtash, J. Randolino, and G. Carroli, “Abortion‐Related Morbidity in Six Latin American and Caribbean Countries: Findings of the WHO/HRP Multi‐Country Survey on Abortion (MCS‐A),” BMJ Global Health 6 (2020): e005618.