Prevalence and factors associated with unsafe abortion among women in reproductive age attending Remera Rukoma Hospital, Rwanda

Abstract

Background

Unsafe abortion remains a critical global health issue, particularly affecting women in low and middle-income countries including Rwanda. In Africa, 99% of abortions are unsafe. Between 2019 and 2023, there has been a reported increased of abortion-related hospitalizations at Remera Rukoma Hospital (RRH), highlighting an urgent need for investigation. Therefore, this study was conducted to determine the prevalence and factors associated with unsafe abortion among women of reproductive age attending RRH.

Methods

This cross-sectional study was conducted among 384 women of reproductive age (15–49 years) attending at RRH in Rwanda. The sample size was obtained using the Cochrane formula and systematic random sampling. Data analysis utilized SPSS Version25, employing descriptive statistics, bivariate and multivariate analyses to determine the factors associated with unsafe abortion at 95% CI, and p-value < 0.05 was considered statistically significant.

Results

The majority of participants were married (51.8%), Catholic (32.8%), attained primary education (35.4%), and were housewives (42.2%). The prevalence of unsafe abortion was 35.2%. Multivariate analysis showed unsafe abortion was higher than twice among women married before 18 years (aOR = 2.277, 95% CI: 1.247–4.157, p = 0.007), those with 3 sexual partners in the last 12 months (aOR = 2.285, 95% CI: 1.031–5.066, p = 0.042). Women who experienced gender-based violence had higher odds of engaging in unsafe abortion (aOR = 1.965, 95% CI: 1.128–3.424, p = 0.017).

Conclusion

This study revealed unsafe abortion as a significant health concern among women of reproductive age, with over one-third of participants reporting unsafe abortions. Early marriage before 18 years, multiple sexual partners, and gender-based violence emerged as key factors associated with unsafe abortion practices. Addressing this issue requires a multifaceted approach, community education, and targeted initiatives to combat gender-based violence and early marriage in Rwanda.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12978-025-02212-w.

Background

Unsafe abortion remains a critical global health issue, particularly affecting women in low and middle-income countries [1]. The World Health Organization (WHO) defines unsafe abortion as a procedure for terminating a pregnancy performed by persons lacking the necessary skills or in an environment not conforming to minimal medical standards, or both [2].

Globally, 45% of all abortions are unsafe, with 97% of these occurring in developing countries [2]. Between 2015 and 2019, there were 73.3 million abortions annually worldwide, with 7 million women in developing countries admitted to hospitals each year due to complications from unsafe abortions [3]. Unsafe abortions contribute significantly to maternal mortality, accounting for 4.7–13.2% of maternal deaths worldwide [4]. Unsafe abortion is a leading but preventable cause of maternal deaths and morbidities [2]. Abortion, especially unsafe can lead to physical and mental health complications and social and financial burdens for women, communities and health systems [2].

In Africa, the situation is particularly severe. The continent has the highest risk of dying from an unsafe abortion, at 520 deaths per 100,000 unsafe abortions [5]. Unlike in Latin America and Africa, where the reports indicate that the majority (approximately 3 out of 4) of all abortions are unsafe [2]. About 99% of all abortions carried out in Africa are unsafe [6]. A study by Bankole et al. (2015) found that in Africa, an estimated 8.3 million induced abortions occurred in 2010–2014, with only 24% considered safe [7]. The majority of unsafe abortions in Africa occur in Eastern Africa, where Rwanda is situated. As of 2010–2014, 77% of abortions in the region are unsafe, compared with the global average of 45% [8]. Maternal death associated with abortion, especially the unsafe, accounts for 37 deaths per 100,000 live births in Sub-Saharan Africa [9]. In most Sub-Saharan countries, legal restrictions and stigma still compel women to undergo clandestine abortions, the safety of which cannot be ensured [8].

In East Africa, unsafe abortion significantly contributes to maternal morbidity and mortality. A study in Tanzania found that 60% of women admitted to a hospital for post-abortion care had undergone unsafe abortions [10]. In Uganda, unsafe abortion accounts for nearly one-third of maternal deaths [11]. Up to 14% of pregnancies end in unsafe abortion in Kenya [12], and the prevalence of lifetime-induced abortion in female sex workers (FSWs) in Kenya was previously estimated between 43% and 86% [13].

In Rwanda, despite improvements in maternal health care, unsafe abortion remains a pressing issue. A study estimated that 60,000 abortions occur annually in Rwanda, with an abortion rate of 25 per 1,000 women aged 15–44 [14]. Many of these abortions are considered unsafe due to restrictive abortion laws and limited access to safe abortion services [15]. Overall, half of all abortions in Rwanda are performed by untrained individuals and are considered to be very high risk-an estimated 34% are provided by traditional healers and 17% are induced by the women themselves [16]. In 2022, 40% of Rwandan women were subjected to complications related to self-induced abortions that required medical attention [17].

Rwanda has ratified the Maputo Protocol (Protocol to the African charter on human and peoples’ right on the rights of women in Africa) and allows access to safe legal abortion under five circumstances in its reformed 2019 penal code and changes as per the ministerial order N^o 002/MOH/2019 [18]. The abortion law in Rwanda was amended in 2012 and 2018 on the conditions for women who become pregnant due to rape, child pregnancy, a second-degree kinship or if the pregnancy can harm a mother or a fetus [19]. At Remera Rukoma Hospital (RRH) in Kamonyi district, Rwanda, a concerning trend in abortion-related hospitalizations has emerged, as evidenced by Health Management Information System (HMIS) data from 2019 to 2023. The hospital has witnessed a steady increase in post abortion care cases, rising from 454 in 2019 to 615 in 2020, 637 in 2021, and 668 in 2023, with a slight dip to 587 cases in 2022. This overall upward trajectory, marked by a 47% increase over five years, highlights a growing public health concern and underscores the urgent need for comprehensive investigation and recommending targeted interventions to address the root causes of this alarming trend.

Various factors contribute to the prevalence of unsafe abortions, as highlighted by recent research. Restrictive abortion laws are strongly correlated with higher rates of unsafe procedures [5]. Limited access to effective contraception leads to unintended pregnancies, a key driver of unsafe abortions [20]. Socioeconomic disparities play a significant role, with women from lower-income backgrounds more likely to resort to unsafe methods [21]. Social stigma and lack of information about safe abortion services further exacerbate the issue [9]. Additionally, healthcare system limitations, including inadequate infrastructure and a shortage of trained providers, restrict access to safe abortion services [22]. Factors associated with abortion in some districts in Rwanda include lack of access to contraception, poverty, and social stigma surrounding unintended pregnancies [23].

The increasing trend of abortion-related hospitalizations (among women who come for post abortion care) at RRH from 2019 to 2023 emphasizes the urgency of this problem in the local context. This study on factors associated with unsafe abortion among women of reproductive age attending RRH was crucial. It provided valuable insights into the local dynamics contributing to unsafe abortions, informed targeted interventions and policies to address this pressing public health concern.

Methods

Study design

This study employed a cross-sectional study design with a quantitative approach among women of reproductive age (15–49 years) attending RRH for obstetric services. A cross-sectional design was chosen as it is appropriate for assessing the prevalence of unsafe abortion practices and identifying associated factors at a specific point in time, providing a snapshot of the burden and characteristics of the problem in the study population. RRH was selected as the study site because it is a district hospital serving a large catchment area with a high volume of obstetric cases, including post-abortion care services, making it an ideal setting to capture diverse experiences and perspectives on unsafe abortion. Additionally, RRH has established reproductive health services and maintains comprehensive patient records, facilitating data collection and ensuring the inclusion of women who have experienced or are at risk of unsafe abortion.

Study setting

This study was conducted in RRH, a faith-based public hospital in Rukoma Sector, Kamonyi district, Southern Province of Rwanda, 9 km from the main road of Kigali- Muhanga. This hospital was founded by Eglise Presbyterienne au Rwanda (EPR) in 1927 as an infirmary, and it was licensed as a hospital by the government of Rwanda in 1971. According population census of 2022, the population of Kamonyi district was 450,849, with 233,279 female [24]. RRH serves the community from 14 health center in its catchment areas from 14 sectors with 3 private clinics, 4 dispensaries, and 44 health posts. According to Rwanda Demographic and Health Survey 2019–2020 (RDHS 2019–2020), it was found that 98% of the women aged 15–49 years in Kamonyi District received antenatal care services from the skilled health care providers, 95% delivered at the health facilities with the assistance of a health care providers [25]. RRH provides post abortion care to the women in reproductive age seeking care. On average, 12 daily spontaneous vaginal deliveries, and 5 C-section were conducted at RRH as of 2023.

Target population

The participants were the women of reproductive age (15 to 49 years) who attended RRH for post-abortion care. Included were women who attended the facility, from 4 months prior to the study up to the last day of data collection, and provided informed consent. Women who were critically ill, had cognitive impairments preventing informed consent, or were unable to communicate were excluded from the study; post-abortion care encompasses medical treatment for complications arising from spontaneous or induced abortions (including management of incomplete abortion, hemorrhage, infection, and counseling); and the 4-month recall period was selected to balance the need for adequate sample size while minimizing recall bias, as it captures recent experiences that participants can reliably remember while including sufficient cases for meaningful analysis.

Variables

In this study, the dependent (outcome) variable was the prevalence of unsafe abortion, defined as abortion conducted in non-hospital settings (home). The independent variables are categorized into three groups: Sociodemographic factors (age, marital status, education level, occupation, partner’s education level, economic level, forced marriage); reproductive health factors (family planning use, infections, parity, pregnancy complications, age at marriage, number of sexual partners, number of live births, unwanted pregnancy, pregnancy with twin, and time between birth of the recent children); and lifestyle factors (alcohol intake, smoking, gender based violence experience). The conceptual framework also acknowledged the intervening factors, including hospital policy on abortion management and government policy on abortion prevention. (Fig. 1) 

Fig. 1.

Conceptual framework for the factors associated with unsafe abortion

Sample size calculation

The sample size was calculated using Cochran’s (1977) formula [26] with a 95% confidence level (Z = 1.96), unknown proportion of unsafe abortion (p = 0.5), and margin of error of 5% (d = 0.05). Using the formula n = Z²p(1-p)/d², a sample size of 384 participants was determined to be necessary for this study.

Sampling technique

This study employed a systematic sampling method with a random start [27] among the women of reproductive age who attended the RRH for post abortion services from 4 months prior, up to the period of data collection. To address the systematic sampling method with a random start, we first obtained the total number of eligible women in reproductive age attending the RRH for post-abortion or abortion services during the 4-month study period. This approach ensured that each woman in the reproductive age group had an equal and predetermined chance of being included in the study, while eliminating potential selection bias [28].

Data collection and material

Semi-structured, interviewer-administered questionnaire from a similar study [29], which was adopted to the Rwandan context, was used in data collection (supplementary file 1). Overall, the instrument used for data collection covered sociodemographic characteristics, factors related to reproductive health, and lifestyle factors. The validity of this data collection tool was verified.

Data collection was conducted among women of reproductive age who sought post-abortion care at RRH and provided informed consent. Data were collected by three data collectors, including 2 midwives and one 1 nurse working in the maternity department at RRH. Data were collected from September to November 2024. To ensure confidentiality and protect participants’ privacy, all responses were recorded anonymously.

Reliability and validity of the instrument

Reliability refers to the consistency of a measure, by giving the same result when repeated [30]. This research on unsafe abortion factors at RRH ensured reliability through a pilot study. A preliminary investigation using 10% of the total sample size was used to check if all study elements were thoroughly covered and if questions were accurately formulated. The internal consistency of the questionnaire was assessed using Cronbach’s alpha. The overall alpha coefficient was 0.76, which is acceptable [31, 32].

Validity refers to the extent to which the score from a measure represents the variable they are intended to [30]. To ensure the instrument’s validity, a pilot study was conducted [33]. During piloting, 38 participants (10% of the total sample size) were interviewed. This helped verify the relevance of all questions and confirmed that all important aspects of the research were addressed.

To determine the content validity index, we followed a systematic approach involving expert review. First, we selected a group of five subject matter experts in reproductive health, research methodology, and clinical practice. This group of experts comprise two obstetricians and gynecologists, one medical director, one experienced midwife, and one experienced researcher. Each expert independently reviewed the research instrument, evaluating each item for its relevance, clarity, and representativeness of the research objectives. The experts rated independently each item on a 3-point scale: 1 (not necessary), 2 (useful but not essential), 3 (essential) [33, 34]. The response from all experts were pooled, and the number indicating “essential” for each item was determined. To establish content validity of an individual items we used an internationally reorganized method (content validity ratio (CVR) which is useful in rejection or retention of individual items [34]. The Content validity index (CVI) is the mean CVR for all the items included in the final instrument [34, 35]. The more experts (beyond 50%), perceiving an item as ‘‘essential’’, the greater the extent or degree of its content validity [34].

The CVR was calculated as follows [36]:

where:

ne is the number of panelists identifying an item as “essential” and.

N is the total number of panelists (N/2 is half the total number of panelists).

When all panelists agree an item is ‘‘essential,’’ the CVR is 1.00 (adjusted to 0.99 for ease of manipulation according to Lawshe [36]. When the number of panelists rating an item ‘‘essential’’ is more than half, but less than all, the CVR is somewhere between 0 and 0.99, and if none of the raters marks the item as ‘‘essential,’’ the CVR would be 0 [34].

To the content validity of entire instruments we calculated a CVI which is simply the mean of the CVR values for all items meeting the CVR thresholds of >0.7 and retained for the final instruments [34]. The overall CVI of 0.74 was obtained by averaging the CVR scores, indicating good content validity [34]. Specifically, items rated 3 or 4 by 4 or more out of 5 experts were considered valid, demonstrating the tool’s accuracy and appropriateness for measuring the intended research constructs [33].

Data analysis

SPSS version 25 was used for data analysis, and descriptive statistics using univariate analysis were used to describe the characteristics of the study population. Bivariate analysis with chi -square test was conducted to determine the relationship between unsafe abortion with each independent variable. The significant variables in bivariate analysis (with p value < 0.05) were entered into a multivariate analysis to adjust for the other potential confounders. Multivariable analysis was used to present the adjusted odds ratio (aOR) of having an unsafe abortion with the independent variables, and a to adjust the confounders. The odds ratio was determined at a 95% confidence interval (CI), and a p-value < 0.05 was considered statistically significant [37, 38]. Multicollinearity diagnostics were conducted using variance inflation factor (VIF), with values less than 5 considered acceptable to ensure the absence of multicollinearity among independent variables.

Ethical consideration

This research prioritized ethical considerations throughout its execution. Prior to commencement, ethical approval was sought from Mount Kenya University’s institutional review board (REF: MKU/ETHICS/23/01/2024) and Remera Rukoma Hospital’s research and ethics committee. Participants were fully informed about the study’s purpose, procedures, risks, and benefits through a detailed consent form, ensuring voluntary participation with the right to withdraw at any time. Written informed consent to participate was obtained from the participants, and the parents or legal guardians of any participant under the age of 18 years old. To maintain confidentiality, data were anonymized using unique codes and stored securely on password-protected devices with encrypted files. Though participants didn’t receive direct benefits, the research aims to contribute valuable insights for improving reproductive health services. This study was conducted in accordance with the ethical standards of our institutional review board and with the 1964 Helsinki Declaration and its later amendments.

Results

Socio-demographic characteristics of the study respondents

Table 1 presents socio-demographic characteristics of participants, where; majority, 199 (51.8%) were married. Regarding religion, majority of 135 (32.8%) were catholic. The majority of respondents, 136 (35.4%), had attained primary education. Regarding occupation of the participant, the majority, 162 (42.2%) were housewives. Concerning monthly income, more than half of the respondents, 196 (51.0%). In regard to the education level of the spouse, the majority, 140 (36.5%). Regarding age group, the largest age group was 19–30 years, comprising 166 (43.2%) of respondents.

Table 1.

Socio-demographic characteristics of the study respondents

Variable	Frequency (N = 384)	Percent
Marital status
Married	199	51.8
Single	81	21.1
Divorced	18	4.7
Separated	32	8.3
Widowed	54	14.1
Religion
Catholic	135	35.2
Muslim	92	24.0
EPR	29	7.6
Adventist	91	23.7
Others religion	37	9.6
Education level
Uneducated	47	12.2
Primary education	136	35.4
Vocational training	48	12.5
Secondary education	115	29.9
University	38	9.9
Occupation
Government employed	76	19.8
Housewife	162	42.2
Student	49	12.8
Private employed	67	17.4
Others	30	7.8
Monthly income
< 100k Rwf	196	51.0
100-300k Rwf	115	29.9
> 300k Rwf	73	19.0
Education level of spouse
Uneducated	50	13.0
Primary education	140	36.5
Vocational training	43	11.2
Secondary education	115	29.9
University	15	3.9
I haven’t spouse	21	5.5
Age group of the participants
<=18 years	22	5.7
19–30 years	166	43.2
31–40 years	134	34.9
>=41 years	62	16.1
Total	384	100

The prevalence of unsafe abortion

Figure 2, shows the prevalence of unsafe and safe abortion, which is 35.2% (135 cases), compared to safe abortion being more common at 64.8% (249 cases).

Fig. 2.

Prevalence of unsafe abortion among the study participants

Factors associated with unsafe abortion among women of reproductive age attending RRH

Table 2, presents a bivariate analysis of factors associated with unsafe abortion among women of reproductive age attending RRH. The analysis examined various demographic variables such as marital status, religion, education level, occupation, monthly income, spouse’s education level, and age group. Notably, none of the variables examined showed statistically significant associations with unsafe abortion, as indicated by the p-values all being > 0.05.

Table 2.

Bivariate analysis of the sociodemographic factors associated with unsafe abortion among the participants

Variable	Unsafe abortion		Total (%)	X2	P-value
	At home (unsafe abortion)n (%)	At health facility (Safe abortion) n (%)
Marital status				1.043	0.903
Married	73(36.7)	126(63.3)	199(51.8)
Single	26(32.1)	55(67.9)	81(21.1)
Divorced	5(27.8)	13(72.2)	18(4.7)
Separated	12(37.5)	20(62.5)	32(8.3)
Widowed	19(35.2)	35(64.8)	54(14.1)
Religion				0.19	0.996
Catholic	48(35.6)	87(64.4)	135(35.2)
Muslim	32(34.8)	60(65.2)	92(24)
EPR	10(34.5)	199(65.5)	29(7.6)
Adventist	33(36.3)	58(63.7)	91(23.7)
Others religion	12(32.4)	25(67.6)	37(9.6)
Education level				3.269	0.514
Uneducated	21(44.7)	26(55.3)	47(12.2)
Primary education	43(31.6)	93(68.4)	136(35.4)
Vocational training	18(37.5)	30(62.5)	48(12.5)
Secondary education	38(33)	77(67)	115(29.9)
University	15(39.5)	23(60.5)	38(9.9)
Occupation				2.237	0.692
Government employed	26(34.2)	50(65.8)	76(19.8)
Housewife	57(35.2)	105(64.8)	162(42.2)
Student	15(30.6)	34(69.4)	49(12.8)
Private employed	23(34.3)	44(65.7)	67(17.4)
Others	14(46.7)	16(53.3)	30(7.8)
Monthly income				1.436	0.488
< 100k Rwf	67(34.2)	129(65.8)	196(51)
100-300k Rwf	38(33)	77(67)	115(29.9)
> 300k Rwf	30(41.1)	43(58.9)	73(19)
Education level of spouse				6.175	0.29
Uneducated	19(38)	31(62)	50(13)
Primary education	54(38.6)	86(61.4)	140(36.5)
Vocational training	8(18.6)	35(81.4)	43(11.2)
Secondary education	41(35.7)	74(64.3)	115(29.9)
University	5(33.3)	10(66.7)	15(3.9)
I haven’t spouse	8(38.1)	13(61.9)	21(5.5)
Age group of the participants				2.208	0.53
<=18 years	5(22.7)	17(77.3)	22(5.7)
19–30 years	63(38)	103(62)	166(43.2)
31–40 years	45(33.6)	89(66.4)	134(34.9)
>=41 years	22(35.5)	40(64.5)	62(16.1)
Forced marriage				1.176	0.675
yes	29(37.2)	49(62.8)	78(20.3)
no	106(34.6)	200(65.4)	306(78)

EPR Eglise Presbyterienne au Rwanda

Table 3 presents a bivariate analysis of reproductive health factors associated with unsafe abortion among women of reproductive age attending RRH. It examined various reproductive characteristics such as age at first intercourse, age at marriage, number of pregnancies, living children, sexual partners, unintended pregnancy, live births, pregnancy complications, twin pregnancies, birth spacing, family planning use, and reproductive infections. Among these factors, two variables showed statistically significant associations with unsafe abortion: age at marriage (χ² = 7.084, p = 0.029), and sexual partners in the last 12 months (χ² = 7.913, p = 0.048). The other reproductive factors examined did not show statistically significant associations with the occurrence of unsafe abortions in this study population.

Table 3.

Bivariate analysis of the reproductive factors associated with unsafe abortion among the participants

Variable	Unsafe abortion		Total, n (%)	X2	P-value
	At home (unsafe abortion), n (%)	At health facility (Safe abortion), n (%)
Age at first vaginal intercourse				1.015	0.798
<=14 years	10(30.3)	23(69.7)	33(8.6)
15–19 years	40(38.1)	65(61.9)	105(27.3)
20–24 years	56(35.7)	101(64.3)	157(40.9)
>=25 years	29(32.6)	60(67.4)	89(23.2)
Age at marriage				7.084	0.029
I have not married	32(49.2)	33(50.8)	65(16.9)
< 18 years	51(30.9)	114(64.1)	165(43)
> 18 years	52(33.8)	102(66.2)	154(40.1)
Total Pregnancies at lifetime				2.205	0.332
1 time	27(29.3)	65(70.7)	92(24)
2 times	64(38.6)	102(61.4)	166(43.2)
>= 3 times	44(34.9)	82(65.1)	126(32.8)
Living children				0.682	0.877
1 child	31(35.6)	56(64.4)	87(22.7)
2 children	40(32.5)	83(67.5)	123(32)
3 children	41(36)	73(64)	114(29.7)
4 children	23(38.3)	37(61.7)	60(15.6)
Life time sexual partner				7.167	0.127
1 partner	37(31.9)	79(68.1)	116(30.2)
2 partners	38(33.6)	75(66.4)	113(29.4)
3 partners	26(36.1)	46(63.9)	72(18.8)
>= 4 partner	23(52.3)	21(47.7)	44(11.5)
I prefer not to say	11(28.2)	28(71.8)	39(10.2)
Sexual partner in the last 12 months				7.913	0.048
1 partner	91(39.7)	138(60.3)	229(59.6)
2 partners	27(34.6)	51(65.4)	78(20.3)
3 partners	9(21.4)	33(78.6)	42(10.9)
4 partners	8(22.9)	27(77.1)	35(9.1)
Unintended pregnancy				3.075	0.079
Yes	60(30.9)	134(69.1)	194(50.5)
No	75(39.5)	115(60.5)	190(49.5)
Live births				1.613	0.656
No Live birth	20(42.6)	27(57.4)	47(12.2)
1 live birth	35(32.1)	74(67.9)	109(28.4)
2 live births	53(34.6)	100(65.4)	153(39.8)
3 live births	27(36)	48(64)	75(19.5)
Pregnancy complications				0.57	0.812
Yes	45(34.4)	86(65.6)	131(34.1)
No	90(35.6)	163(64.4)	253(65.9)
Pregnancy with twin				0.432	0.516
Yes	6(28.6)	15(71.4)	21(5.5)
No	129(35.5)	234(64.5)	363(94.5)
Time between birth of the recent child				2.55	0.769
< 6 months	10(28.6)	25(71.4)	35(9.1)
6 to 11 months	12(40)	18(60)	30(7.8)
12 to 17 months	25(34.2)	48(65.8)	73(19)
18 to 23 months	25(30.5)	57(69.5)	82(21.4)
>=24 months	51(38.3)	82(61.7)	133(34.6)
This my first pregnancy	12(38.7)	19(61.3)	31(8.1)
Current family planning use				1.578	0.209
Yes	88(37.6)	146(62.4)	234(60.9)
no	47(31.3)	103(68.7)	150(39.1)
Reproductive infections				0.119	0.73
yes	49(36.3)	86(63.7)	135(35.2)
no	86(34.5)	163(65.5)	249(64.9)

Table 4, presents a bivariate analysis of lifestyle factors associated with unsafe abortion among women of reproductive age attending RRH. The analysis examined factors such as forced marriage, alcohol consumption, smoking status, chronic health conditions, and gender-based violence (GBV). Among these factors, only gender-based violence showed a statistically significant association with unsafe abortion (χ² = 7.116, p = 0.029). Other lifestyle factors such as alcohol consumption, smoking status, and chronic health conditions did not demonstrate statistically significant associations with the occurrence of unsafe abortions in this study population.

Table 4.

Bivariate analysis of the lifestyle factors associated with unsafe abortion among the participants

Variable	Unsafe abortion		Total (%)	X2	P-value
	At home (unsafe abortion), n (%)	At health facility (Safe abortion), n (%)
Alcohol consumption				0.208	0.648
yes	120(34.8)	225(62.2)	345(89.8)
no	15(38.5)	24(61.5)	39(10.2)
Smoking status				0.001	0.974
No, I’ve never smoked	28(35)	52(65)	80(20.8)
Yes, I smoke occasionally (not every day)	107(35.2)	197(64.5)	304(79.2)
Chronic health conditions				0.016	0.901
Yes	28(34.6)	53(65.4)	81(21.1)
No	107(35.3)	196(64.7)	303(78.9)
Gender based violence				7.116	0.029
Yes	32(26.9)	87(73.1)	119(31)
No	49(35)	91(65)	140(36.5)
Prefer not answer	54(43.2)	71(56.8)	125(32.6)

The multivariate analysis of factors associated with unsafe abortion among women of reproductive age attending RRH revealed several significant findings (Table 5). Women who married before 18 years of age had significantly higher odds of unsafe abortion (aOR = 2.277, 95% CI: 1.247–4.157, p = 0.007). Regarding sexual partners in the last 12 months, women with 3 partners had significantly higher odds of unsafe abortion (aOR = 2.285, 95% CI: 1.031–5.066, p = 0.042). In terms of gender-based violence, women who responded “yes” had significantly higher odds of unsafe abortion (aOR = 1.965, 95% CI: 1.128–3.424, p = 0.017).

Table 5.

Multivariate analysis of the factors associated with unsafe abortion among the participants

Variables	Unsafe abortion
	aOR	95% CI	p-value
Age at marriage
I have not married	Ref
< 18 years	2.277	[1.247–4.157]	0.007
> 18 years	1.827	[0.998–3.344]	0.051
Sexual partner in the last 12 months
1 partner	Ref
2 partners	1.173	[0.678–2.027]	0.569
3 partners	2.285	[1.031–5.066]	0.042
4 partners	1.931	[0.815–4.576]	0.135
Gender based violence
I prefer not to say	Ref
Yes	1.965	[1.128–3.424]	0.017
No	1.316	[0.792–2.185]	0.289

Discussion

The present study investigated factors associated with unsafe abortion among 384 women of reproductive age attending RRH. The sociodemographic characteristics of this study reveal interesting patterns among women of reproductive age attending RRH in Rwanda. The majority of respondents were married (51.8%), which aligns with findings from a study by Bell et al. (2022) in Burkina Faso, where 58.6% of women seeking post-abortion care were married [39]. Regarding education, 35.4% had primary education, while 29.9% had secondary education, suggesting a moderate level of literacy among participants. This distribution differs somewhat from a study in Ethiopia by Gebrehiwot et al. (2016), where 41.2% of women had no formal education, highlighting potential regional variations in educational access [40].

The occupational data shows that 42.2% of respondents were housewives, followed by 19.8% in government employment. This high proportion of housewives is notable and may have implications for economic independence and decision-making power regarding reproductive health. The monthly income distribution reveals that over half of the participants (51%) earn less than 100,000 Rwandan francs, indicating a significant proportion living in lower income brackets. This economic profile may influence access to safe abortion services and contraception, as observed in a study by Chae et al. (2017) in Ghana, where economic factors were strongly associated with unsafe abortion practices [41]. These sociodemographic findings provide crucial context for understanding the factors that may contribute to unsafe abortion practices in this Rwandan district.

The prevalence of unsafe abortion among women of reproductive age in this study was 35.2% (135/384). The findings of this study are contrary to a study conducted in Nepal by Khatri et al. (2019), the prevalence of unsafe abortion was reported at 23%, lower than our findings [29]. However, a higher prevalence of 34.6% was reported in a study from Nigeria [42]. Moreover, a multi-country study across sub-Saharan Africa by Bankole et al. (2022) reported varying prevalence rates of unsafe abortion, ranging from 18% to 40%, placing our findings within this range of broader regional context [8]. The variation in these prevalence rates could be attributed to differences in healthcare access, legal frameworks, cultural norms, and socioeconomic factors across these different settings. Our finding of 35.2% prevalence indicates a significant public health concern in the Kamonyi district of Rwanda, suggesting a need for improved access to safe abortion services and comprehensive reproductive health education.

Multivariate analysis shows that early marriage (before 18 years) was found to be significantly associated with increased odds of unsafe abortion (aOR = 2.277, 95% CI: 1.247–4.157, p = 0.007). This finding aligns with several studies conducted in other regions. A study by Gelaye et al. (2014) in Ethiopia reported that women who married before the age of 18 were at higher risk of unsafe abortion [43]. Similarly, research by Chae et al. (2017) across 21 countries found that adolescent marriage was associated with increased risk of unsafe abortion [44]. In addition, a study by Kabiru et al. (2016) in Kenya found that early sexual debut was associated with increased likelihood of unsafe abortion among young women [45]. These findings suggest that early marriage may limit women’s reproductive autonomy and access to family planning services, potentially leading to unintended pregnancies and unsafe abortions. Furthermore, these findings highlight the importance of addressing early marriage and its health implications.

This study also found that women with multiple sexual partners, specifically three partners in the last 12 months, had significantly higher odds of unsafe abortion compared to those with one partner (aOR = 2.285, 95% CI: 1.031–5.066, p = 0.042), pointing to the need for comprehensive sexual education and family planning services. This result is consistent with findings from other studies. Palermo et al. (2014) reported that women with multiple sexual partners in sub-Saharan Africa were more likely to experience unintended pregnancies and seek unsafe abortions [46]. Moreover, a study by Mekie et al. (2021) in Ethiopia also found that having multiple sexual partners was a risk factor for unsafe abortion [47]. In addition, a study by Kabiru et al. (2016) in Kenya found that multiple sexual partnerships were associated with increased likelihood of unsafe abortion among young women [45]. These findings suggest that multiple sexual partnerships may increase the risk of unintended pregnancies and subsequent unsafe abortions, possibly due to inconsistent contraceptive use or higher exposure to sexual activity.

The association between gender-based violence and higher odds of unsafe abortion (aOR = 1.965, 95% CI: 1.128–3.424, p = 0.017), emphasizing the critical intersection of women’s rights, safety, and reproductive health. Additionally, women experiencing violence may prioritize immediate safety concerns over seeking proper medical care, resort to clandestine abortion methods to hide pregnancies from abusive partners, or lack the financial resources and social support necessary to access safe abortion services. This finding is consistent with the study which were conducted in other areas, a systematic review by Hall et al. (2014) found a consistent relationship between intimate partner violence and increased risk of unsafe abortion across multiple countries [48]. Similarly, a study by Pallitto et al. (2013) in 17 countries reported that women who experienced intimate partner violence were more likely to have unintended pregnancies and abortions [49]. Moreover, a study conducted by Stöckl et al. (2012) in Tanzania also found that women who experienced intimate partner violence were at higher risk of having terminated pregnancies [50]. These consistent findings across different contexts underscore the significant impact of gender-based violence on women’s reproductive health and decision-making, highlighting the urgent need for integrated interventions that address both violence prevention and reproductive health services to protect vulnerable women.

Limitations

This study has limitations that should be considered. As a hospital-based study, it may not fully represent the broader population of women in the Kamonyi district who do not seek hospital care. Additionally, the cross-sectional design limited our ability to establish causal relationships. Moreover, this study did not assess whether none of those unsafe abortions occurred at a health facility. Despite these limitations, the study has important implications for public health policy and practice in Rwanda. The findings underscore the need for targeted interventions to address early marriage, promote sexual and reproductive health education, and combat gender-based violence. Furthermore, the results suggest that improving access to safe abortion services and comprehensive family planning could significantly reduce the prevalence of unsafe abortions. Future research should consider community-based approaches to capture a more representative sample and explore the effectiveness of interventions aimed at the identified risk factors.

Conclusion

This study found that unsafe abortion remains a significant health concern among women of reproductive age. This study revealed a high prevalence of unsafe abortions among the participants, with more than one-third of the participants having undergone abortions in non-medical settings. Key factors associated with unsafe abortion were identified as early marriage before adulthood, having multiple sexual partners, and experiencing gender-based violence. These findings collectively suggest that addressing unsafe abortion requires a multifaceted approach, including legal reforms, improved healthcare access, community education, and initiatives to combat gender-based violence and early marriage.

Abbreviations

aOR

Adjusted Odd Ratio

CI

Confidence interval

CVI

Content Validity Index

CVR

Content Validity Ratio

DHS

Demographic and Health Survey

DALYs

Disability-Adjusted Life Years

EPR

Eglise Presbyterienne au Rwanda

FSWs

Female sex workers

GBV

Gender based violence

HMIS

Health Management Information System

IPIs

Interpregnancy Interval

MKU

Mount Kenya University

MOH

Ministry of Health

NISR

National Institute of Statistics Rwanda

OHAC

One Health Approach for Conservation

RBC

Rwanda Biomedical Center

RRH

Remera Rukoma Hospital

RDHS 2019–2020

Rwanda Demographic and Health Survey 2019–2020

SSA

Sub-Saharan Africa

VIF

Variance inflation factor

WHO

World Health Organization

Authors’ contributions

ES conceptualized the study and wrote the initial draft. ES developed and wrote the final manuscript. ET reviewed the manuscript. EK co-supervised this study and reviewed the manuscript. BV reviewed the manuscript. CH participated in data collection tool validation and reviewed the manuscript. EM1 reviewed the manuscript. GS reviewed the manuscript. IZ reviewed the manuscript. JM reviewed the manuscript. JPB participated in data collection tool validation and reviewed the manuscript. PU, WN, FR, LU, EM2, PYN, JCN, EN, RM, EI, and JFK reviewed the manuscript. CN supervised the implementation of this study, participated in data collection tool validation and reviewed the manuscript. All authors critically reviewed the manuscript for important intellectual content. All authors read and approved the final version of the manuscript for publication.

Funding

This study was not funded.

Data availability

The datasets generated and analyzed during this study are available from the corresponding author upon reasonable request. The statistical analysis code used in this study is also available upon request.

Declarations

Ethics approval and consent to participate

This research prioritized ethical considerations throughout its execution. Prior to commencement, ethical approval was sought from Mount Kenya University's institutional review board (REF: MKU/ETHICS/23/01/2024) and Remera Rukoma Hospital's research and ethics committee. Participants were fully informed about the study's purpose, procedures, risks, and benefits through a detailed consent form, ensuring voluntary participation with the right to withdraw at any time. Written informed consent to participate was obtained from the participants, and the parents or legal guardians of any participant under the age of 18 years old. To maintain confidentiality, data were anonymized using unique codes and stored securely on password-protected devices with encrypted files. Though participants didn’t receive direct benefits, the research aims to contribute valuable insights for improving reproductive health services. This study was conducted in accordance with the ethical standards of our institutional review board and with the 1964 Helsinki Declaration and its later amendments.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.