Discontinuation of Long-Acting Reversible Contraceptive methods and associated factors among women in health facilities of Bahir Dar city, Northwest Ethiopia: A Cross-sectional study

Yilkal Dagnaw Melesse (  dagnaw.yilkal@gmail.com ) Assosa University Ambaye Minayehu Zegeye Assosa University Getahun Deguale Kebede Assosa University Yaregal Dessalew Tarik Assosa University Selamawit Lake Fenta Bahir Dar University Alemwork Abie Getu Bahir Dar University Toyiba Hiyaru Wassie Bahir Dar University Tewodros Worku Bogale Injibara University Magarsa Lami Dabalo Haramaya University Paulos Jaleta Assosa University


Background
Discontinuation of Long-acting reversible contraceptives is starting use and cessation or switching to other methods for any reason before completion of duration (1). Intrauterine contraceptive devices (IUCD) and implants are known as long-acting reversible contraceptives (LARCs). They represent highly effective methods of modern birth control and their duration of action is long ranging from 3 to 10 years (2). One of the fundamental practices of women's health care is the provision of effective contraception like LARCs; the annual pregnancy rates for IUCDs are less than 1 per 100 women and the clinical failure rate of Implant is less than 1%( 3).
Decreasing discontinuation of LARCs is an excellent strategy to avoid or at least reduce unwanted pregnancy (UP), which affects millions of women in the world (4). Discontinuation of LARCs contributes to unplanned pregnancies which expose a risk to the health of women, affects the family size, and contributes to higher fertility rate, social, economic, and physical health problems (5). Although there is an improvement in long-acting reversible contraception usage, discontinuation is becoming a major problem. In low-income countries, 20% of women discontinue LARCs within one year of insertion (6).
There was a good achievement in reduction of fertility rate and increasing modern contraceptive prevalence rate observed in the Millennium Development Goals era. Despite this, the government of Ethiopia is again running to address Sustainable Development Goal 3(SDG-3) which proposed that roughly 20% of married reproductive age women will use LARCs by 2030 (7).
Discontinuation of LARCs was high in Ethiopia in which 45% of IUCD and 61% of implants were discontinued by three years of utilization. The high LARCs discontinuation is coupled with low uptake; is a big challenge to achieve the targeted prevalence rate in the country and Among the associated factors of discontinuation, didn't receive pre-insertion counseling, experienced side effects, and desire for pregnancy were mentioned (8). Long-acting reversible contraceptives can prevent thousands of cases of maternal mortality by preventing annually occurring 87 million unintended pregnancies (9).
Discontinuation of LARCs is higher in third world countries than in the developed world; in Myanmar, LARCs discontinuation is 7% for intrauterine contraceptive devices, 0.2% for implants, and 28% for women shifted to other modern contraceptives (10).
Discontinuation of LARCs is high in developing countries and most of these occur among women who are at risk of unplanned pregnancy (11). World health organizations and national governments had started promoting LARCs as a strategy to implement family planning programs (12). Despite increases in availability, effectiveness, and knowledge on LARCs in Ethiopia, the overall utilization of LARCs is very low and there is high discontinuation including the Amhara region (13).
There is no study conducted on the proportion and associated factors of all the three components of LARCs (IUCD, Jaddles, and Implanon) in Bahir Dar city. Therefore, this study was forwarded to have the contribution in searching of the associated factors and help to decrease the bad outcomes of discontinuation of LARCs in Bahir Dar city health facilities.
Findings from this study will direct health professionals to improve counseling service provision, the ability of planning and decision-making to look at possible solutions regarding LARCs. The ndings will also be an input for the Amhara regional health bureau, health planners, and other concerning organizations working in the area of family planning and maternal health. Moreover, this study will also be used as baseline data for researchers who want to study in this area of research. Finally, the investigator believes that the ndings from this study will help to the achievement of sustainable development goal 3 of family planning at the country level.

Conceptual framework
The detailed factors or variables for discontinuation of LARCs are mentioned in the section of the Conceptual framework ( Figure 1)

Study setting and Design
An institution-based cross-sectional study design was conducted in Bahir Dar city health facilities which provide Family Planning services including LARCs from April 1 to May 30 /2021. Bahir Dar is the capital city of Amhara National Regional State in the Federal Democratic Republic of Ethiopia located in the Northwest direction at 565 km away from Addis Ababa, the capital city of Ethiopia (14). The population of Bahir Dar City was estimated to be 389,177 in the year 2020 based on the 2007 census (15).
The Bahir Dar city has one specialized referral, one comprehensive referral, and one primary public hospital (Tibebe Ghion, Felege Hiwot, and Addis Alem respectively), 6 health centers (Abay health center, Han health center, Shimbet health center, Menelik health center, Bahir Dar health center, and Shum-abo health center), 10 health posts, two specialized clinics (family guidance association and Maire stope), 4 private general hospitals and 35 medium private clinics (16). Family guidance association and Maire stop specialized clinics are non-governmental facilities that provide consistent family planning services including implants and Intrauterine contraceptive devices. Therefore, these were included in this study.

Source population
All reproductive-age women who were using long-acting reversible contraceptive methods in health facilities of Bahir Dar city. Study population All reproductive-age women who were using long-acting reversible contraceptive methods came to selected health facilities of Bahir Dar city with contraceptive-related issues during the study period. Inclusion criteria and Exclusion criteria Inclusion criteria All reproductive age women who were using LARCs came to the selected health facilities for any issue concerning the method before completion of duration; like removal, side effects, and follow-up during the actual data collection period.

Exclusion criteria
All women who were using LARCs came to the selected health facilities for the removal of LARCs who got the service outside Bahir Dar city and who came to change Implanon to Jaddles, Jaddles to Implanon, and implants to IUCD or vice versa.

Sample size determination
Accordingly, for the rst and second speci c objectives, the sample size was calculated separately, and the larger sample size was taken to be used for this study. Sample size determination for objective one was calculated by using single population formula by considering the following assumptions: 95% CI, the prevalence (P) of LARC discontinuation taken from a study conducted in Hawassa city which was 57% with a non-response rate of 10% (1).
Sample size determination for objective two was calculated by using double population formula with Epiinfo version 7.2.1 Stat Calc by considering the following assumptions: Two-sided signi cant level (1alpha): 95% CI, power (1 -beta, % chance of detecting):80%, the ratio of Exposed: Unexposed =1, and nonresponse rate of 10%. Exposed in this case means exposed to factors for removal (experiencing side effects, lack of counseling, and desire for pregnancy). These factors were taken from a study conducted in Hawassa city (1) ( Finally, the calculated sample size for the second objective with a 10% nonresponse rate is less than that of the rst objective. Therefore, the sample size of the rst objective was the sample size of this study which is 415.

Sampling technique
All public health facilities of Bahir Dar city which has LARC services were included in the study (3 hospitals and 6 health centers). Two Nongovernmental specialty clinics (Family Guidance Association and Maire stops) were also included since they provide LARCs service. The other private health facilities were not included since they have no full LARC services (insertion and removal). Hence a total of eleven (11) health facilities were included in this study.
The study subjects were identi ed based on the information obtained from the Family planning registration book. The information obtained from two Months report from the family planning registration book of one year preceding this study period in similar months observed that a total of 830 women were booked or used LARCs in health facilities of Bahir Dar City. The average two-month client ows for health facilities were taken.
The proportionally allocated sample size for each facility was calculated as ni=Ni*nf/N; where ni=sample size in each facility, Ni=number of women who were using LARCS in each facility within two months, nf =total sample size of this study (415), and N=all women in Bahir Dar City health facilities who were using LARCs (830).
Finally, the total sample size was met by using the systematic sampling technique of daily caseload. K was calculated by dividing the total two-month caseload in Bahir Dar City for the total proportional allocated cases as 830/415=2. Therefore, every other woman was included in the sample until the total sample size for this study was obtained. Based on the client's card order information was collected from participants after completion of the service. The procedure was continued throughout the data collection period until the required sample size was achieved ( Figure 2

Operational De nition
Long-acting reversible contraceptive methods (LARCs): contraceptive methods which serve as 3-10 years but can be removed at any time (not permanent); only implants and IUCD (17).
Discontinuation of LARCs: starting using and cessation of long-acting reversible contraceptives before completion of duration due to any issue concerning with the method (1).
Misconception: without scienti c evidence, the woman perceived LARCS can cause infertility, cancer, shifting to other body sites, and reduce sexual activity (18).
Follow-up: is schedule or appointment given for the mother to come to health facilities on regular basis or when she has concerns with the method not only appointed by the due date.
Side effect: when the women develop at least one of the following after LARC insertion: menstrual disruption, insertion site pain, di cultly to work, headache, acne, and weight change (14,17).

Data Collection Instruments, Collectors, and Procedures
The data for this study were collected by using a pre-tested structured interview-based questionnaire which was adapted from previous literature (1) with some modi cations. There are four main parts to this questionnaire. These are socio-demographic characteristics, obstetric-related factors, contraceptive method and health facility-related factors, and others factors including myths and misperceptions about LARCS. To check for its consistency, the questionnaire was rst developed in English and translated into Amharic and then nally back to English.
Data collectors were ve Nurses (3 diplomas and 2 degrees) and six Midwives (2 diplomas and 4 degrees) who were not working in health facilities. Data were collected from study subjects through a face-to-face interview-guided questionnaire. Data were collected and the questionnaire was lled by data collectors after obtaining written consent from the study participants.
Data collectors have gathered information from study subjects around the family planning room after completion of the service provided by health care providers. Three supervisors with a quali cation of degree and skilled in data collection supervised the data collection. The principal investigators supervised and provided all necessary items for data collection, checked completeness and logical consistency, and solve problems during the time of data collection daily.

Data Quality Assurance
To assure the quality of data, properly designed data collection tools were used. Data collectors and supervisors were trained about research objectives, data collection tools, procedures, and interview techniques for one day. Before the actual data collection, the questionnaire was tested on 5% of the total sample size (21 women) who were used LARC methods at Zenzelima Health Center (11 women) and Meshenti Health Center (10 women) to check the contextually of data.
The questionnaires were revised and get amended as necessary after the pretest was done. The principal investigator, together with three supervisors supervised the technique of data collection and completeness of tools on the daily basis, and accordingly, appropriate feedback was forwarded. To resolve differences and assure their quality, data double entry was performed.

Data Processing, Analysis, and presentation
After data collection, the questionnaire was checked for completeness, data entered, coded, checked, and cleared by using EPI DATA V-3.1 and then exported to Statistical Package for Social Science [SPSS] V-25 computer software. Then analysis (Descriptive statistic to summarize data) and Bi-variate logistic regression were performed at 95% con dence level with LARCs discontinuation for each factor. From the bivariate regression analysis, variables at p < 0.25 were entered into multivariate logistic regression analysis with a 95% con dence level, multicollinearity test was done to check the correlation between each independent variable, and the model goodness of t was tested using Hosmer and Lemeshow test.
Variables with a p-value less than 0.05 in the multivariate logistic regression analysis were considered as statistically signi cant factors for LARCs discontinuation. Finally, the descriptive statics results were presented using text, tables, and graphs based on the type of data.

Ethical Consideration
The study gets research ethics approval from Bahir Dar University, College of Medicine and Health Sciences Institutional Review Board (IRB). The permission and agreement consent was obtained from the Bahir Dar city Health Department and the selected health facilities before data collection.
Voluntary, informed, written, and signed consent were obtained from heads of selected health facilities in Bahir Dar city and participants before data collection started. The participants were also assured that their responses could not result in any harm and offered full rights not to participate, the name did not write on the tool, and con dentiality was maintained.     More than half of the participants discontinued after one and before three years of utilization which accounts for 157(57.1%), 54(19.6%) discontinued between 3 and 4 years, and 15(5.45%) were discontinued after 4 years of utilization. The detailsares shown below (Figure 3).

Factors Associated with Discontinuation of Reversible Long-Acting Contraceptives
Findings from bivariate logistic regression analysis of this study resulted that maternal religion, maternal occupation, maternal education, desire to be pregnant soon, type of LARC used, past contraceptive utilization, counseling about the bene t of LARCs, side effects, and follow-up were entered into multivariate regression analysis. However, after adjusting other variables in a multivariable logistic regression analysis only maternal education, pregnancy desire in near future, no past contraceptive utilization, not counseled for bene t of LARCs, and experiencing side effects were found signi cantly associated with LARCs discontinuation.
According to this nding, the odds of discontinuation of LARCs among women who have no formal education were 51% less likely than college and above (AOR=0.49; 95% CI: 0.30-0.82). The odds of discontinuation of women who attended a primary level of education were reduced by 61% than college and above (AOR=0.39; 95% CI: 0.18-0.81) This study also revealed that the odds of discontinuing LARCs among women who desire to be pregnant shortly were 2.6-fold higher than their counterparts (AOR=2.57; 95% CI: 1.64-4.02). Women who didn't use contraceptives before current LARC utilization were about 2 times more likely to discontinue than their counterparts (AOR=2.01; 95% CI: 1.19-3.38).
The odds of discontinuing LARCs among women who experienced side effects were 2 times higher than those who didn't experience side effects (AOR=1.95; 95 CI:1.21-3.16). This study also showed that the odds of discontinuation of long-acting reversible contraceptives among women who had not received counseling about the bene t of LARCs was 1.7 times higher compared to their counterparts (AOR=1.68 95% CI:1.08-2.62) ( Table 6).

Discussion
In this study, the overall proportion of LARCs discontinuation among women was 66.3% (95% CI: 61.42-69.14). The discontinuation of Implants was 55.3% of which Implanon was 34.3% and the discontinuation of IUCD was 11%. Discontinuation of LARCS in this study is low compared with the secondary data analyzed from EDHS 2016, where 61% of implants and about 45% of IUCDs users were discontinued at the end of three years (8). This might be due to the discrepancy of population structure of the national pattern and variation in the duration of utilization before discontinuation.
This study is also lower than a study done in Colorado, Western United States that is 81.7% (19). The possible explanation might be the discontinuation in Colorado was within 30 months of utilization. The other explanation might be most of the health professionals in our setup might didn't easily accept complaints of removal since the proportion of utilization of LARCs is low (18.4%) (20).
However, discontinuation of LARCs in this study is higher than studies conducted in the Sidama region, southern Ethiopia, 10.3% (21), and Hawassa City, southern Ethiopia, 56.6% (1). This difference might be due to the duration of utilization in the Sidama region being within one year and the presence of the religious varieties in the Sidama region and Hawassa city as most of the study participants in Bahir Dar were Orthodox Christian. On the other hand, recently in Ethiopia customers visiting health facilities for removal is also increasing since in earlier times a signi cant number of users think impossible to discontinue the LARC before the due date after once inserted due to the awareness gap (7). Also due to the pandemic of COVID-19 health care providers might not expense much time for detailed counseling.
The current nding is also higher than the study conducted in Uganda, 6.8% (22), Nigeria, 58.1% (23), and Maryland, 33.2% (24). This difference might be due to culture and religion as more than 90% of this study participants were orthodox Christian. This difference might also be due to inadequate pre-insertion counseling as evidenced by 62.7% of the participants in this study didn't receive counseling about side effects and didn't receive counseling on the bene t of LARCs was one of the factors for LARCS discontinuation.
In this study, the odds of discontinuation of LARCs were reduced by 51% and 61% among women who have no formal education and attended primary school respectively compared with women who attended college and above. This is in line with related studies conducted in Bahir Dar, northwest Ethiopia (14), and Debre Markose town, northeast Ethiopia (25). The possible explanation might be women who had higher education levels might seek healthcare sooner than their counterparts when they have complaints and any disturbance from normal physiology (14).
This nding is not supported by a study done in the Sidama region, Southern Ethiopia (21), and Hawassa City, Southern Ethiopia (1). This is incoherent to my expectation and the possible reasons could be the differences in sociocultural of the population.
This study revealed that the odds of discontinuation of LARCs among women who desire to be pregnant soon were 2.6 times higher than their counterparts. This is in agreement with studies done in Ambo town, Central Ethiopia (26), Hawassa city (1), analyzed of EDHS 2016 data (8), Kenya (27), Nigeria (28), and Zambia (29). The possible reason might be the improper choice of the method as the nature of the method is long-acting. Evidence shows that LARCs must be removed when the clients have a desire to conceive (30). The other possible reason might be women with less or who have no children might intend to have children as evidenced by 19% and 55% of the study participants in this study had no alive children and have 1-2 children respectively.
The odds of discontinuation of LARCs among women who have not used any modern contraceptive before the current LARC were about 2.0 times high likely than their counterparts. This nding is not supported by studies conducted in the Sidama region (21), Hawassa city (1), and worldwide (4). This could be because experienced mothers acquired the necessary knowledge and attitude towards the LARCs, while others could be in uenced by false beliefs, myths, and misconceptions. The other possible reasons could be fear of delays in pregnancy and disagreement on the method used with a partner.
The odds of discontinuation of LARCs among women who had not received pre-insertion counseling services about the bene ts of the LARCs was 1.7 times higher than from their counterparts. This nding is not agreed with a study conducted in Myanmar (10). The suggested explanation for this might be due to the low socio-economic status, poor quality of services, unfavorable participants' attitudes towards LARC in our set up than Myanmar. This is evidenced by only 30% of all women using modern contraceptives being informed at the time they started the current method use about the method's side effects, what to do if they experience side effects, and other possible solutions (30).
The odds of discontinuing LARCs among women who experienced side effects were about 2.0 times higher than those who did not experience side effects. This is consistent with the studies conducted in Hawassa (1), analysis of EDHS 2016 data (8, 31), Nigeria (32), South Africa (33), Bangladesh (34), and North America (24). The most possible reason for this might be once women developed side effects, they may not tolerate the counseling and management services delivered by health professionals for their complaints due to fear of happening different complications on their health in the future (14).
On the other view, it may be due to discomfort; not well addressed pre-insertion counseling about possible side effects or development of side effects may lead to other physiological or psychological effects.
Unexpected changes in menstrual bleeding due to the nature of the method by itself might enforce women to seek removal (26). Also, when there is the development of side effects women might be in uenced by their relatives and partners (35).

Limitation
Since this study was conducted on reversible long-acting contraceptive discontinuation as a whole unlike most other studies which were done on a single contraceptive, it was somewhat challenging to obtain literature in line with the title (especially in Ethiopia).

Conclusion And Recommendations
The overall discontinuation of reversible long-acting contraceptives among women who were using LARCs was found to be low compared to the analysis of EDHS 2016. The associated factors for discontinuation of LARCs among user women of reproductive age were educational level; the desire to pregnant soon did not use modern contraceptives before reversible long-acting contraceptives, not received counseling about the bene t of LARCs, and experienced side effects. Therefore, health professionals better give counseling, implement early management of side effects and reassure clients of any concerns. Before data collection respondents were informed about the objective and purpose of the study to get written informed consent from each respondent. At the time of data collection written informed consent was obtained from each study participant for those ages greater than 18 years and from parents/guardians for those ages less than 18 years. All respondents were assured that the data would not have any negative consequences on many aspects of their life. The participants were also assured that their responses could not result in any harm and offered a full right to participate or decline from participating in the study and the study participants assured for an attainment of con dentiality of the information obtained from them. Their name did not written on the tool. Their privacy was guaranteed during the whole period of the study by anonymous completion of the questionnaires. For the respondents, it was noti ed that they have the right to refuse or terminate at any point of data collection.

Competing interests
The authors declare that they have no competing interests.

Consent to Publish
Not applicable

Availability of Data and Materials
The data supporting the ndings of this study are available at the hand of co-responding author but some restrictions may apply to the availability of these data as there are some sensitive issues. However, data are available from the corresponding author (Yilkal Dagnaw) upon reasonable request.

Funding
Assosa University as a requirement for postgraduate studies supports this research nancially. The university has no role in the design of the study, collection, analysis, and interpretation of the data, and in writing the manuscript.
Author contributions YM developed the protocol and was involved in the design, selection of study, and developing the initial drafts of the manuscript. SF, AG, and TW are involved in quality assessment. YM, AZ, GK, YT, PW, TB, and MD prepared and revised subsequent drafts as well as prepared the nal draft of the manuscript. All authors read and approved the nal draft of the manuscript.

Figure 1
Conceptual framework for factors associated with LARCs adapted from pieces of literature, BahirDar, 2021(1) Figure 2 Schematic presentation of sampling the procedure for discontinuation of reversible long-acting contraceptives and associated factors, Bahir Dar city, 2021