Trend of Unintended Pregnancy, Induced Abortion and Associated Factors among Adolescents in Ethiopia: Evidence from the 2000, 2005, 2011 and 2016 EDHS Data

Abstract


Introduction
Unintended pregnancies are pregnancies that are mistimed, unplanned or unwanted at the time of conception [1].Unintended pregnancies may also result from rape, incest or various other forms of forced or unwanted sex [2,3].
Unintended pregnancy is a global public health problem and its sequels are major causes for maternal morbidity and mortality with its effect to maternal illness as well [4][5][6].Globally, unintended pregnancy has a critical public, clinical and social health concern because it commonly results in induced abortion whether it's safe or unsafe leads to complication and maternal death.It has been reported that of the estimated 210 million pregnancies that occur each year worldwide, approximately 38% are unplanned and 22% are terminated [7].
Induced abortion is the termination of a fetus that brought about intentionally using different options safe or unsafe based on quality of care.It doesn't include spontaneous and other post abortion cares that comes accidentally or due to other causes when the pregnancy is intended [7,8].Globally, the estimated rate in the period 2010-2014 was 35 abortions per 1000 women (aged 15-44 years) and an estimated 21.6 million unsafe abortions took place worldwide in 2008 almost all in developing countries [5].Abortion rates have declined signi cantly since 1990 in the developed world but not in the developing world [7,9].Induced abortion and unintended pregnancy rates are increasing in developing regions than developed regions [9].Ensuring access to sexual and reproductive healthcare could help millions of women avoid unintended pregnancies and ensure access to safe abortion [7].
Adolescent is a transitional phase of growth and development between childhood and adulthood [10] between the age of 10 and 19 years [11].Adolescents are highly at risk of unintended pregnancy due to physiological, sexual, social and psychological growth [12].The pregnancy may end with early childbirth, induced abortion and its complications [13].In Ethiopia, an estimated 620,300 induced abortions were performed in 2014.The annual abortion rate among women aged 15-49 was 28 per 1,000 that showed an increase from 22 per 1,000 in 2008, and was highest in urban than rural areas [14].
Each year, about 79 million unintended pregnancies, excluding miscarriage, occur worldwide.More than half of these unintended pregnancies end in abortion [15].Data on abortion rates are inexact but can be used to explore trends.
Unintended pregnancy rates have declined worldwide but vary widely across regions.In many areas of the world where rates of unintended pregnancy are high [16,17], and unsafe abortions have also been shown to be high [18].Of all pregnancies worldwide in 2008, 41% were reported as unintended or unplanned, and approximately 50% of these ended in abortion.Of the estimated 21.6 million unsafe abortions occurring worldwide in 2008 (around one in 10 pregnancies), approximately 21.2 million occurred in developing countries, often due to restrictive abortion laws and leading to an estimated 47,000 maternal deaths and untold numbers of women who will suffer long-term health consequences [16].
In Africa, the importance of knowing the trends of unintended pregnancy and induced abortion is very vital.Because different literatures show us, in developing countries there is high number of unintended pregnancy and abortion, whether safe or unsafe.In Malawi, 53% pregnancies are unintended and that of 30% ends with abortion which is 2015 [19].Similarly, higher rate of abortion was reported in Nigeria [20] and in Congo [21].
Ethiopia's parliament amended the criminal code of the federal democratic republic of Ethiopia the (Proclamation No. 414/2004) to expand the circumstances in which abortion is legal in 2005 and the country has expanded abortion care service and post abortion family planning [22].One in three birth is unintended (32%) and about two third of these were mistimed [23].Unintended pregnancy has a multiple determinates and leads to an induced abortion and affects maternal health, and may end with morbidity and mortality [4,5].
Results from EDHS 2016 reported 22% unmet need for family planning of which 9% was for limiting and spacing.The contraceptive prevalence rate trend has shown a signi cant change from 6%in 2000 to 35% in 2016.But there are factors associated with it and needs further studies [24].Information on trends of unintended pregnancy and induced abortion has a great role in developing strategies and prevention modalities for complications due to unintended pregnancy and unsafe abortion where most maternal injuries and deaths are related to unsafe abortion specially in developing regions including Ethiopia.Thus, this study aimed to determine the trend of unintended pregnancy, induced abortion and associated factors among adolescents in Ethiopia from 2000 to 2016.

Data source
The study utilized data from cross sectionally conducted four consecutive Ethiopian Demographic and Health Surveys of 2000, 2005, 2011 and 2016 EDHS women's data (IR data is the EDHS women's data le code) set in Ethiopia.According to the Ethiopian census report; by the year 2018 the estimated population of Ethiopia were 106.8 million of this 24.2% were adolescent (from 9 to 18 years) and 11.7% of them were 15 to 19 year old.Adolescent girls were 5.8% from a total population.The fertility rate was 56.61 births per 1000 women aged 15-19 years old [25].Until now, four surveys were conducted in Ethiopia where reports are available in CSA and online [26].The rst survey was conducted in 2000/1 followed by the 2005/6 EDHS which is part of the worldwide measure DHS project funded by USAID and was conducted under the auspices of the ministry of health and implemented by the then Population and Housing Census Commission O ce (PHCCO).The EDHS surveys of 2011 and 2016 are also the same surveys like other years.The objective of EDHS is for planning, policy formulation, monitoring, and evaluation of population and health program in the country [24].

Sample size determinations and sampling procedure
Our analysis was based on women's data (IR data) set for EDHS 2000, 2005, 2011 and 2016.The EDHS used a two-stage strati ed cluster sampling technique to select the study participants.In stage one, after each administrative region was strati ed into urban and rural strata, Enumeration Areas (EAs) were selected using a probability proportional to EA size.In stage two, a household listing operation was carried out in all of the selected EAs and a xed number of households from each EA were selected.All women aged 15-49 years who were permanent residents or who spend the night in the selected households the night before the survey were included in the surveys [24].In this study, a total weighted sample of 2544 adolescents (15)(16)(17)(18)(19)

Study variables
The dependent variables considered were unintended pregnancy and abortion.Unintended pregnancy is de ned as a pregnancy which is a sum of mistimed pregnancy (pregnancy wanted at a later time) and unwanted pregnancy (pregnancy which is not wanted at all [3].Abortion was taken from the EDHS question 'have you ever had a pregnancy termination?' with the response 'yes' if the woman ever had an abortion and otherwise 'no' as a binary outcome.The independent variables included in this study were socio-demographic variables like maternal age, educational status, occupation status, marital status, religion, wealth status, residence, region and household size.Reproductive health variables such as parity, birth order, age at rst sex, antenatal care visit, place of delivery, knowledge of ovulation cycle, distance from the health facility, involving in health care decision, media exposure, knowledge and practice of family planning.Variables were recoded to make suitable for the study analysis such as region; Afar, Somali, Benishangul, and Gambela as "small peripheral regions", Tigray, Amhara, Oromia, and Sothern Nations Nationalities and Peoples Region (SNNPR) as "large central regions" and Harari, Dire Dawa, and Addis Ababa as "metropolitans", based on their geopolitical features [27].

Data extraction instrument
The EDHS tool has three components: the household questionnaire, the woman's questionnaire and the man's questionnaire.These questionnaires were adapted from model survey instruments developed for the MEASURE DHS project to re ect the population and health issues relevant to Ethiopia [24].The data for this study were taken from woman's questionnaire.The variables reviewed to be important for this speci c research were collected from the SPSS lled data received from http:// www.measuredhs.comand these data were transferred into a Stata Software version 14.

Data processing and analysis
Preliminary analyses involved the description of the study participants by calculating frequencies and percentages of the study variables (socioeconomic, demographic, reproductive health services).This was followed by the estimation of the prevalence of the outcome variables (unintended pregnancy and abortion) and by socioeconomic and reproductive health variables in both years' speci c data (2000, 2005, 2011 and 2016) and in the combined dataset [24].Then, percentage point change with corresponding 95% CI of the outcome variables was calculated by each of the study factors to examine the changes over the research period from 2000 to 2016.We used the combined dataset to increase the statistical power of the study in order to detect any association between the study factors and the outcomes, as well as to examine trends in unintended pregnancy and abortion over the study period (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016).We estimated P for trends in each category of the study variables to assess whether the prevalence decreased or increased over the study period using chi-squared test for trend.
A multi-level logistic regression analysis was applied to account for the hierarchal nature of the DHS data and bivariable multilevel logistic regression analysis was carried to determine the crude odds ratios at 95% con dence interval and those variables with p-value <0.25 were considered for multivariable analysis.In the multivariable multilevel logistic analysis, those variables with p-value <0.05 were considered as signi cantly associated with outcome variable.After selecting variables for multivariable analysis, four models containing variables of interest were tted.These models were model 1 is the null model in which no factor variables were included, model 2 which examined the effects of individual-level factors, model 3 examined the effect of community level factors and model 4 which examined the effects of both individual and community level variables.The Intra-Class Correlation (ICC) was estimated to assess the cluster effect and the model tness was compared using information criteria (IC).All statistical analyses were conducted using Stata version 14.0 with 'svy' command to adjust for sampling weights, clustering effects and strati cation, and the 'melogit' command was used for the regression analysis.Finally, the results were presented with tables and graph.

Data quality and assurance
The data quality depends on EDHS data quality and the formats that are going to be use like data extraction formats.The quality of the data was maintained by checking its completeness, cleaning the missing values by running frequencies and some of the variables were re-coded.

Socio demographic and economic characteristics
In this study, a total weighted sample of 2544 women age 15-19 years who gave birth in the 5 years preceding the surveys or pregnant during the surveys were included.About three-fourth of study participants were 18 and 19 years of age (75.5%) and the overall mean age was 18 years.Majority of respondents, 60.8% had no formal education and 45.3% of them were orthodox Christians.Around 60.7% of participants were unemployed and 57.6% of women were from poor household wealth status.Most, 89.8% of women were married while only 43.5% of them had media exposure.Regarding household size, 75.7% of the respondents were from a household size of two to ve.Majority of the women were rural residents (91.1%) and from large central regions (92.2%).Distance to health facility was a big problem for 71.2% of respondents and 55.6% of them not involved in decision for health care (Table 1).Reproductive history of the study participants Among 2544 total study respondents, about two-third (67.6%) of women had one live child, 28.2% had two or more children while 4.2% had no live birth child.About two-third (66.0%) of respondents were fteen and younger years old during their rst sex.About 38.5% of respondents had at least one Antennal care (ANC) visit history, and only 13.0% of study participants delivered their child in a health facility.Most, 89.8% of woman knew at least one family planning method while13.4% of them correctly knew that a women is most likely to conceive half way between two periods and only 23.1% of them had ever used contraceptive.Among 2544 total study respondents, 1677 (66.0%) had wanted then (planned) pregnancies, 657(25.9%) of them had wanted later (miss-timed) pregnancies while 206(8.1%)had wanted not at all pregnancies.While, 132(5.2%) of respondents had abortion (Table 2).

Prevalence of abortion among adolescents in Ethiopia
Over the study period (2000-2016), the overall prevalence of abortion was 5.2% (95%CI: 3.9-6.7%).Women who had no live child had the highest prevalence of abortion (10.2%), followed by women who resided in the metropolis regions (8.2%) and those women who were urban residents (7.4%).The lowest prevalence of abortion was observed among women from female headed households (1.4%), followed by women from middle wealth households(2.7%)(Table 5).

Factors associated with unintended pregnancy among adolescents in Ethiopia
In the null model, about 35.6% of the total variation on unintended pregnancy was occurred at the community level which is accounted to the community-level factors (ICC=0.356).The variation of unintended pregnancy between clusters suggested to conduct multilevel analyses.The model tness comparison was carriedout using information criteria (IC) and the model with the lowest IC value (Model 4) was the best-tted model (Suplemetary Table 1).In the multivariable multilevel logistic regression analysis, age, marital status, parity, and regions were factors signi cantly associated with unintended pregnancy among adolescents aged from 15-19 years old in Ethiopia.Adolescents aged above 17 years old had lower odds of having unintended pregnancy compared to those aged 15-17 years old (AOR=0.30;95% CI: 0.17-0.51).Married women had signi cantly lower odds of having untended pregnancy compared to those who were unmarried (AOR=0.14;95% CI: 0.07-0.28).Women who had two and more live children had signi cantly higher odds of unintended pregnancy compared with their counter parts (AOR= 1.68; 95%CI: 1.02-2.75).In addition, women who resided in large central (AOR= 3.32; 95%CI: 1.86-5.63),and metropolis (AOR=3.12;95%CI: 1.20-8.12)regions had signi cantly higher rate of having unintended pregnancy compared to those who resided in the small peripheral region (Table 6).[24].Factors associated with a lower likelihood of unintended pregnancy among adolescent girls in Ethiopia included being aged above 17 years old and being married.While, higher parity, resided in large central and metropolis regions were signi cantly associated with higher odds of untended pregnancy.
The prevalence of abortion also decreased signi cantly from 8.3% in 2000 to 4.1% in 2016.Belonging to middle and rich wealth households, having unintended pregnancy, female head households and being rural residents were factors signi cantly associated with lower likelihood of abortion among adolescents in Ethiopia.These reduction may be attributable to the relatively high maternal health service utilization like modern contraceptive utilization (35%) ) in 2016 [24] compared to (6%) in 2000 [30].These ndings re ected that reproductive health interventions targeting the reduction of unintended pregnancies and induced abortion among adolescents should be strengthen more in order to reduce unintended pregnancies, abortion related mortality and morbidity, improve adolescent health and permitting adolescents to attain their economic potential [5,[31][32][33][34].
This study showed that adolescent girls aged below 18 years old had higher odds of having unintended pregnancy.
Similarly, studies conducted in Africa indicated that the risk of unplanned pregnancies decreases with age that older mothers had lower odds of unintended pregnancy in developing countries [35][36][37][38][39][40].This might be due to older women had relatively better knowledge on contraceptive methods to prevent unintended pregnancy and lower contraceptive failure rate [35,41].Moreover, as they are getting older, women might also become more literate about the importance and accessibility of reproductive or maternal health services.In addition, this could be also as the results of older women are less likely to engage in risky sexual behaviors such as unprotected sexual intercourse and sex under the in uence of drinking alcohol [42,43].However, other literatures revealed a negative relationship between age and unintended pregnancy [37,44,45].This nding might be related to the fact that adult women might already have the desirable number of children and considered any additional pregnancy as mistimed or unwanted.
In this study, education was no signi cantly associated with untended pregnancy.There was no agreement across researches regarding the impact of the level of education on the risk of having unintended pregnancies.A systematic review on prevalence and determinants of unintended pregnancy in Sub-Saharan Africa revealed that the outcomes ranged from reporting no association with level of education and odds of having unintended pregnancies in Nigeria[46], South Africa [47], Ethiopia[48], and Kenya [39], through an increased risk of reporting the last pregnancy as unintended as the level of education increases in the Democratic Republic of Congo and Ethiopia [49,50], to an increase in the level of education being protective against unintended pregnancies in Nigeria [51], in Tanzania [52] and in Ethiopia [53].Researches that analyzed demographic health survey data also found discordant ndings in Ethiopia [23]; in the study by Tebekaw, education was protective against unintended pregnancies [23], while the other study found no association between education and unintended pregnancy.A study done in Nairobi also found that education was protective against having an unintended pregnancy, with a 10% decrease in the risk of having an unintended pregnancy [53].
In this study, married women had signi cantly lower odds of untended pregnancy compared to those who were unmarried.
This nd is consistent with studies conducted in Ethiopia [6, 29, 54-56], in Kenya [39] and in South Africa [57]; reported that unmarried mothers were at higher odds of unintended pregnancy.The potential reason is that unmarried adolescent girls may unintentionally participate in sexual activity and this is most likely unwanted if the pregnancy is occurred.Additionally, our culture and community might cause a signi cant in uence on unmarried women not to use contraceptive because of sex is not recommended before the women became married.In addition, this could be due to the fact that married mothers are less likely to engage in risky sexual behaviors such as unprotected sexual intercourse and sex under the in uence of drinking alcohol [42][43][44].On the other hand, unmarred adolescent girls could engage in sexual activity for pleasure.Therefore, if pregnancy occurs it is more likely to be unintended.Furthermore, they are less likely to use contraceptive methods [56,58].Other reasons could be di culty in accessing contraception, not having adequate nancial and social support to provide for their unborn child [58].
The study also indicated that adolescent mothers who had relatively higher parity experienced more risk of having unintended pregnancy.This nding is supported by researches done in Ethiopia [31,56,59], in Ghana [45] and in Pakistan [35], that reported multiparous and grand multiparous mothers had higher odds of unintended pregnancy compared with primiparou women.Similarly, studies from Brazil and USA have reported that mothers who have more alive children are more likely to experience an unintended pregnancy [60].The possible explanation might be high parity woman might already have adequate children with a decreasing intention for the next pregnancy.This might be also due to mothers with higher parity might be busy in caring their children and family and this affects their getting of information, accessing and utilization of maternal health services such as contraceptive methods which in turn end up with unintended pregnancy.In addition, it might imply the gaps in provision of contraceptives counseling and education [4].
Evidence has shown that prevalence of unintended pregnancy differs across socioeconomic levels at national and subnational levels [4,28,29].This study also found signi cant variations of unintended pregnancy rate across regions in Ethiopia that adolescent girls from large central and metropolitan regions had higher likelihood of unintended pregnancy.
This regional difference of unintended pregnancy prevalence is congruent with research ndings from Ethiopia [23,27], Kenya [39], and Ghana [45].This could be due to pregnancies might be appreciated and accepted in small peripheral regions and mothers in this region did not think about unintended pregnancy.Nevertheless, women in metropolitan and large central regions may be busy because of their intention to advance their socio-economic status and mostly their pregnancies are more likely to be unintended [28,29].These variations a cross regions re ect the signi cance of disaggregated data for evidence-based policymaking and program design and context-speci c interventions are required in order to reduce unintended pregnancy and associated abortion among adolescents in Ethiopia.
Adolescents from middle and rich wealth households had signi cantly lower rate of abortion compared to those from poor wealth households.This nding is supported by previous researches evidence that women with lower income had higher rate of induced abortion than those with higher income [61,62].Finding from the EDHS Surveys have shown that modern contraception use increases sharply with wealth, ranging from 20% in the lowest wealth index to 47% in the highest wealth index quintile [24].Hence, the low contraceptive utilization among adolescents with lower economic status may account for the higher odds of abortion [63].It might also due to nancially not prepared to raise their unborn child.Therefore, women in poor wealth households may get an induced abortion in order to limit births due to economic, demographic and social reasons [64].Abortion was also less likely to occur among women whose husband had occupation [45,67].This might be because employed husbands have a positive attitude in avoiding pregnancy termination by increasing their economic status and women might get frequent health institution visits and health information about consequences of pregnancy termination.Adolescents from poor income families may be also more risky to sexual coercion and rape that can lead to unintended pregnancy and induced abortion [59] .
Furthermore, adolescents from female headed households experienced signi cantly lower rate of abortion compared to those from male headed households.Likewise, previous studies also documented that female-headed households were less likely to experience unwanted adolescent pregnancy [4,51].This might be due to adolescent girls from female headed households would have access to information where to get the services and perhaps may be they had their own income to raise their child, as a result women might get frequent health institution contact for antenatal care, delivery and health information about consequences of pregnancy termination.And it could be also due to relatively better level of reproductive health knowledge and access to family planning services, freedom of decision making on reproductive matters among female head households compared with male headed household counterparts [45,66].
In this study, the likelihood of abortion among women who had unintended pregnancy was signi cantly lower than women who had intended pregnancy.A study conducted in West Arsi Zone, Ethiopia also showed a negative relationship between abortion history and unintended pregnancy [56].Possible explanation may be even though women who had unintended pregnancy tend to avoid unwanted pregnancies by inducing their current pregnancy, they might still lack the self-rule to make decisions about their pregnancy intention.As it is the case with many other areas of reproductive health, husbands appear to be the primary decision makers with consider to pregnancy [67].Moreover, studies also indicated that previous unintended pregnancies and previous abortions were reported to be risk factors for subsequent unintended pregnancies [68].This could be an indication of inadequate sex education, contraception education, or sub-optimal post-abortion counseling among women with previous unintended pregnancies Place of residence was also a signi cant factor associated with abortion.Adolescents from rural areas also had lower rate of abortion compared to urban counterparts.This nding was supported by studies conducted in Ethiopia[69].This might be because adolescents from urban areas are exposed to different factors that make them risky to sexual behaviors that will end in unwanted pregnancy and induced abortion.Urban girls have a lot of access to addiction of stimulants and alcohol than rural girls as well as better access to abortion services.Moreover, adolescents from rural areas might also fear to induce abortion believing that terminating pregnancy as a sin due to their religion believes [24].

Strengths And Limitations Of The Study
The research is based on nationally representative data with large sample size so that the observed ndings more likely to show the status of unintended pregnancy and abortion among adolescents in Ethiopia.The study has the potential to provide evidence for policy-makers and program planners to design appropriate intervention strategies both at national and regional levels because the estimates are based on the national survey data.Nevertheless, the study had limitations because of the Ethiopian demographic and health surveys data are mostly based on respondents' self-report and could have the possibility of recall bias.In addition, it is di cult to establish a cause-effect relationship between the outcomes and independent variables due to the research is based on cross-sectional data.

Conclusion
The result showed that the prevalence of unintended pregnancy among adolescents in Ethiopia signi cantly decreased from 41.4% in 2000 to 25.1% in 2016.Despite this reduction, approximately one-fourth of pregnancies were unintended among adolescents in Ethiopia, indicating still the magnitude of the problem is high.This nding suggested that unintended pregnancy is still one of the .majorreproductive health challenges faced by adolescents in Ethiopia.Factors associated with a lower likelihood of unintended pregnancy among adolescents included being aged above 17 years old and being married.While, higher parity, resided in large central and metropolis regions were signi cantly associated with higher odds of untended pregnancy.Similarly, the prevalence of abortion was also decreased signi cantly from 8.3% in 2000 to 4.1% in 2016.However, abortion is still one of the common reproductive health problems among Ethiopian adolescent girls.Belonging to middle and rich wealth households, having unintended pregnancy, being rural resident and female head households were factors signi cantly associated with lower likelihood of abortion among adolescents in Ethiopia.Thus, providing maternal and child services along with adequate sex education, contraception education, and health information about consequences of pregnancy termination would be vital.Moreover, those identi ed high-risk groups deserve special attention in terms of improving adolescents' reproductive health knowledge and access to FP services, enhancing decision making autonomy on reproductive matters including increasing accessibility and availability of quality maternal health services.
years old) who gave birth in the 5 years preceding the surveys or pregnant during the surveys were included.Trends in the prevalence of unintended pregnancy and abortion were investigated based on a series of the Ethiopia Demographic and Health Surveys (EDHS) data for the years 2000 (n=741), 2005 (n=708), 2011 (n= 616), and 2016 (n=479).

Figures
Figures

Table 3
Prevalence trend of unintended pregnancy by socio demographic and economic variables among adolescents in Ethiopia, 8.2(1.8-14.6)29.5(16.9-41.9)12.8(2.3-23.2) 40(22.5)-19.9(-36.5, -3.3) 0.019 Note: n (%): weighted count and proportion for each variable, % (95%CI): point prevalence with its 95% con dence interval, aDiff: indicates the point percentage change in prevalence of unintended pregnancy between 2000 to 2016 (aDiff% 95% CI is the percentage point change with corresponding 95% CI of the outcome variable calculated by each of the study factors to examine the changes over the study period from 2000 to 2016), SNNPR: Southern Nations Nationalities and Peoples Region

Table 4
Prevalence trend of unintended pregnancy by reproductive history variables among adolescents inEthiopia, 2000Ethiopia,  -2016 Note: n (%): weighted count and proportion for each variable, % (95%CI): point prevalence with its 95% con dence interval, aDiff: indicates the point percentage change in prevalence of unintended pregnancy between 2000 to 2016((aDiff% 95% CI is the percentage point change with corresponding 95% CI of the outcome variable calculated by each of the study factors to examine the changes over the study period from 2000 to 2016), SNNPR: Southern Nations Nationalities and Peoples Region Note: n (%): weighted count and proportion for each variable, % (95%CI): point prevalence with its 95% con dence interval, aDiff: indicates the point percentage change in prevalence of unintended pregnancy between 2000 to 2016((aDiff% 95% CI is the percentage point change with corresponding 95% CI of the outcome variable calculated by each of the study factors to examine the changes over the study period from 2000 to 2016), SNNPR: Southern Nations Nationalities and Peoples Region

Table 5
Prevalence trend of abortion by variables among adolescents in Ethiopia, 2000-2016 Note: n (%): weighted count and proportion for each variable, % (95%CI): point prevalence with its 95% con dence interval, aDiff: indicates the point percentage change in prevalence of abortion between 2000 to 2016 (aDiff% 95% CI is the percentage point change with corresponding 95% CI of the outcome variable calculated by each of the study factors to examine the changes over the study period from 2000 to 2016), SNNPR: Southern Nations Nationalities and Peoples Region Note: n (%): weighted count and proportion for each variable, % (95%CI): point prevalence with its 95% con dence interval, aDiff: indicates the point percentage change in prevalence of abortion between 2000 to 2016 (aDiff% 95% CI is the percentage point change with corresponding 95% CI of the outcome variable calculated by each of the study factors to examine the changes over the study period from 2000 to 2016), SNNPR: Southern Nations Nationalities and Peoples Region

Table 6
Multilevel logistic regression analysis of individual and community-level factors associated with unintended pregnancy among adolescents in Ethiopia Note: Variables with p-value <0.25 in bivariable multilevel logistic regression analysis were included in multivariable analysis to control confounders.*= signi cant at p-value<0.05

Table 7
Multilevel logistic regression analysis of individual and community-level factors associated with abortion among adolescents in Ethiopia Discussion This study assessed trend of unintended pregnancy, abortion and associated factors among adolescents in Ethiopia from 2000 to 2016.The overall prevalence of unintended pregnancy was 34.0% among adolescents in Ethiopia.Studies done in sub-Saharan Africa countries including Ethiopia also showed that unintended pregnancy rate in sub-Saharan Africa remains high, especially among adolescents [4, 28, 29].The prevalence of unintended pregnancy among Ethiopian adolescents decreased signi cantly from 41.4% in 2000 to 25.1% in 2016, potentially re ecting the impacts of the national reproductive health interventions in Ethiopia Note: Variables with p-value <0.25 in bivariable multilevel logistic regression analysis were included in multivariable analysis to control confounders.*= signi cant at p-value<0.05