Determinants of long acting reversible contraception utilization in Northwest Ethiopia: An institution-based case control study

Background Though long-acting reversible contraceptives (LARCs) are highly effective, have minimal side effects, require minimal follow-up, and are low cost, only 10% of contraceptives used in Ethiopia are LARCs. The reason for this low uptake is not understood at the country or regional level. Therefore, this study identified determinants of LARC utilization in Northwest Ethiopia. Methods A facility-based unmatched case control study, using LARC users as cases and short- acting reversible contraception (SARC) users as controls, had been conducted at fourteen public health institutions in Northwest Ethiopia. A systematic random sampling technique was used to select participants with a 1:2 case to control ratio (n = 1167). Binary logistic regression analysis was used to identify determinants of LARC utilization among family planning service users. Results Wealth status [AOR:1.87, 95%CI (1.08, 3.24)], history of abortion [AOR:2.69, 95%CI (1.41, 5.12)], limiting family size [AOR: 2.38, 95%CI (1.01, 5.62)], good knowledge about LARCs [AOR: 2.52, 95%CI (1.17, 5.41)], method convenience [AOR: 0.23, 95%CI (0.16, 0.34)], good availability of method [AOR:0.10 (0.05, 0.19)], less frequent visits to health facility [AOR:2.95, 95% CI(1.89, 4.62)], health care providers advice [AOR:10.69, 95%CI (3.27, 34.87)], partner approval [AOR:0.66, 95%CI (0.45, 0.97)], and favorable attitude towards LARCs [AOR:13.0, 95%CI (8.60, 19.72)] were significantly associated with LARC utilization. Conclusion Professional support, favorable attitude towards LARC use, high economic status, history of abortion, advantage of less frequent visits, having good knowledge towards LARC and interest of limiting births were significantly associated with LARC Utilization. On the other hand, perceived method convenience, and contraception availability were inversely associated with it. Family planning education about the benefits of LARC should be done by health providers and media. Male involvement in the counselling and decision making about the advantage of using LARC may improve the negative influence of partners on LARC utilization. It is also recommended that, future qualitative research further explore perceptions of LARC use.


Conclusion
Professional support, favorable attitude towards LARC use, high economic status, history of abortion, advantage of less frequent visits, having good knowledge towards LARC and interest of limiting births were significantly associated with LARC Utilization. On the other hand, perceived method convenience, and contraception availability were inversely associated with it. Family planning education about the benefits of LARC should be done by health providers and media. Male involvement in the counselling and decision making about the advantage of using LARC may improve the negative influence of partners on LARC utilization. It is also recommended that, future qualitative research further explore perceptions of LARC use.

Background
In today's modern world, the voluntary control of fertility is especially important for couples to accomplish their individual goals and aspirations in addition to bearing children. The rapid growth of the human population in the world is another threat for survival for all, unless comparable economic growth of nations is attained. Hence birth control is of paramount importance for sustainable development and better living conditions [1][2][3][4][5].
Among modern birth control methods, long-acting reversible contraceptives (implants and Intra Uterine Contraceptive Device (IUCD)) are highly effective birth control technologies, determined to be one of the top tier methods by the World Health Organization [6]. Compared to short-acting reversible contraceptives (SARCs), long-acting reversible contraceptives (LARCs) have excellent effectiveness in avoiding unwanted pregnancy and its consequences [7][8][9]. They are also safe, with good continuation rates and low-cost. For these reasons, multiple studies have shown that they are the first-line contraceptive choices for all groups of clients including adolescents and women with chronic illness including HIV/AIDS [10][11][12][13][14]. Once placed properly, LARC users do not need to visit health care providers repeatedly, and LARCs are therefore considered to be "forgettable" for the women who use them. Studies also indicate that LARCs have failure rates of less than one percent [12], comparable to failure rates for tubal sterilization and vasectomy [15,16]. For this reason, LARCs are considered a means of chemical sterilization with an excellent potential to replace surgical sterilization techniques. Since LARCs are easily reversible, they are also good options to avoid post-sterilization regret [17]. In contrast to the very low failure rates of LARCs, the failure rate for SARC use among ideal users is 0.2% to 18% and the typical user failure rate is 6% to 28%, which may lead to higher rates of unwanted pregnancy and its consequences as compared to LARCs [11,18,19].
However, SARCs may be a better option for women who want to become pregnant sooner, and those who could benefit from the non-contraceptive benefits of SARCs such as decreasing incidence of ovarian and endometrial cancer and improvement of anemia [20].
Over all, there is ample evidence regarding the benefits of LARC use over SARC in terms of method effectiveness, side effects, safety and cost [18,[21][22][23].
Use of such highly effective birth control methods is very important for communities suffering from high maternal and perinatal mortality and morbidity exacerbated by high fertility rates as well as large populations and unbalanced economic growth like Ethiopia [2]. In addition, LARCs have the potential to decrease the very high maternal and neonatal mortality of developing countries like Ethiopia [24].
Globally, LARC utilization among sexually active reproductive-age individuals is lower than SARC utilization [10,25]. Despite the overall increase in contraceptive prevalence in Ethiopia over the last two decades, the current proportion of SARC utilization over LARCs is 3.6-fold. In the Amhara region of Ethiopia, the modern contraception prevalence rate is 47%, and LARC use accounts for just 15% of this [26].
Reasons affecting LARC utilization are multidimensional including socio-demographic factors such as older age, higher level of education, better socioeconomic status and place of residence [27]. In addition, user related factors such as higher parity, history of unwanted pregnancy, history of smoking, misconception towards LARCs, lack of knowledge, negative attitude and fear of side effects affect LARC utilization [28,29]. Health facility related factors such as level of the health facility, availability of the method, partner influence are the other points affecting LARC use. Professional related factors such as level of training, competency of providers, and attitudes of providers may affect LARC utilization as well [30][31][32]. However, the reasons for low uptake of LARC are not well known in Ethiopia, particularly in our study site of Northwest Ethiopia. Therefore, the aim of this facility-based case control study was to identify the determinant factors affecting LARC utilization in Northwest Ethiopia, information which can then be used by health officials and policy makers seeking to increase LARC utilization in Ethiopia.

Study design and setting
A facility-based unmatched case control study was conducted after getting ethical approval from Institutional Review Board of the University of Gondar. The study aimed to assess the determinants of LARC utilization among modern family planning service users at health institutions in Northwest Ethiopia. The study was conducted by comparing women who use LARCs and SARCs among modern contraceptive users in Northwest Ethiopia's public and non-governmental organizations (NGO) health facilities. Unlike descriptive cross-sectional studies, the case control design used in this study has the power to identify independent risk factors strongly associated with the specific outcome of interest; which, here was LARC use. This study was conducted from July 1st, 2016 to September 30th, 2016 in fourteen public and NGO health facilities located in Northwest Ethiopia. Northwest Ethiopia is part of the Amhara region, which includes Bahirdar, the current capital city of the Amhara region, and the city of Gondar. The topography of the area includes highlands and lowlands with different cultural, religious and ethnic groups. Currently, the study area has an estimated population of 9,011,940 of which approximately 50% are females [33].
Out of 41 health institutions (based on case load and geographic distribution) in the study area, we selected a total of 14 (34%) health facilities using simple random sampling method. The facilities included were five government hospitals (Dabark, Gondar, Metema, Debretabor and Felegehiwot); seven government health centers, (Debretabor, Kokit, Dabark, Gondar, Maraki, Bahirdar and Abaymado) and Gondar and Bahirdar Family Guidance Association (FGA) clinics.
The calculated sample size was proportionally allocated to each selected health facilities based on the previous consecutive three months average daily client flow of the units which was obtained from family planning registration log books. Every second person of the cases and controls were selected by using systematic random sampling procedure from family planning service users who were visiting the selected health institutions during the data collection period. The first client in each health facility was selected by lottery method. In all of the selected health facilities, family planning services are provided upon the request of clients. When clients in the family planning clinics seek contraceptives health care professionals provide them with the appropriate information about the available methods and offer them their method of choice.

Participants and recruitment
Study population. The study population included all females using modern contraceptives who were attending family planning clinics in the selected fourteen health facilities during the study period.
Inclusion and exclusion criteria. Cases were all women using LARCs (IUCD, or implants) during the study period. Controls were women, using SARCs (oral contraceptive pills, condoms, or injectables) during the study period. Those who were seriously ill and/or unable to communicate for any reason were excluded.
Sample size calculation. The sample sizes for cases and controls were calculated based on the following assumptions: a 95% confidence level, power of 90%; and a 1:2 case to control. The required sample size was calculated using Epi Info version 7, and four variables from three previous studies conducted in similar settings were considered [34][35][36]. Variables such as residence, parity, working status, and number of children were taken from the three studies. Based on the calculation, the largest sample size out of the four variables was obtained from the variable "number of children," and it was 1,061 [34]. With the consideration of a 10% nonresponse rate, the final sample size was determined to be 1,167 women.

Study variables
The dependent variable was LARC utilization. The independent variables were socioeconomic and demographic factors (age, marital status, client's level of education, partner's educational status, occupation, residence, wealth index, religion, and ethnicity), user-related factors (fear of side effects, fear of needles/pain, knowledge, previous use of LARC, attitude towards LARC, method effectiveness, parity, history of unintended pregnancy, and smoking) and health facility related factors (type of health institution, location of health institution, availability of contraceptives (LARC or SARC). In addition, independent variables included partner influence (partner's knowledge and attitude about contraceptives, women's autonomy and decision making, lack of discussion with partners), and provider related factors (type of profession, experience and training was collected through document review; and, competency of method provision, provider beliefs and attitude, and provider counseling skills) were assessed using reported client perception. In addition, religion was included due to variation in guidance on family planning approaches across religions.

Operational definitions
Long acting reversible contraceptive methods (LARCs). Modern contraceptive methods are used for more than a year once the method is provided. In this study, intra-uterine contraceptive devices (IUCDs) and implants were the only LARC methods included.
Short acting reversible contraceptive methods (SARCs). Modern contraceptive methods that are used for less than or equal to three months once the method is provided, such as oral contraceptive pills, injectables and condoms.

Knowledge of LARC.
Women were considered to be knowledgeable if they were familiar with at least one of the LARCS (IUCD and/or implant).
Parity-women were classified as: Nulliparous: women without a history of delivering a child Primipara: women having delivered one child Multiparous: women having delivered more than one child Grand: multiparous-women delivered greater than or equal to five children Gravidity: women were classified as: Primigravida: women having a history of one pregnancy Multi-gravida: women having more than one pregnancy Abortion: loss of a fetus/unborn baby either intentionally or unintentionally before 28 weeks of gestation.

Data collection procedures
The data was collected using a structured questionnaire in an exit interview of clients that came to the health facilities for family planning services. Participants provided written, informed consent before data was collected. The questionnaire was designed to obtain information on the socio-demographic characteristics of contraceptive users in public health facilities, as well as ascertain their reproductive history, utilization of modern contraceptive family planning use and factors affecting LARC methods. The questionnaire was prepared in English and translated into Amharic by a language professional. Seven interviewers and four supervisors were involved in the data collection process. The data collectors and supervisors took part in a two-day training covering topics including how to conduct an interview, rights of the study subjects, and ethical issues such as confidentiality. After the training, a pretest was administered to 60 women (5% of the sample size) in health facilities, which were not selected for this study. The questionnaire was modified after the pre-test was administered. On average, a single interview took around 30 minutes.

Data processing and analysis
The data were entered into Epi info version 3.5.3, cleaned, and then transferred into STATA version 14.0 software for analysis. To capture wealth differences, Principal Component Analysis (PCA) scores were generated. The participants were asked about all possible household properties they owned. The common factor scores were summed up and ranked from lowest to highest.
Participants were said to be knowledgeable about LARC if they knew at least one of the LARCs. There were six Likert-scaled questions, which were used to assess the attitude of participants towards LARC. Those who scored above the mean (equivalent to 50 percent and above) were classified as having a favorable attitude and those who scored less than the mean (equivalent to less than 50 percent) were classified as having an unfavorable attitude. Binary logistic regression was used to identify socio-demographic determinants of LARC utilization. We incorporated variables, which had a p-value of less than 0.2 in the bivariable analysis in the final multivariable model. The Hosmer-Lemeshow goodness-of-fit test was used to test the overall goodness-of-fit. Adjusted odds ratios (AOR) with 95% confidence intervals (CI) were used to report the strength of the associations between LARC use and its explanatory variables.

Ethical consideration
Ethical clearance was obtained from the University of Gondar Institutional Review Board. A formal letter was given to the regional health bureau, zonal and city administration health bureaus, hospitals, health centers, and FGA clinics. We obtained informed verbal consent from each woman enrolled in the study. The ethical review committees approved the procedure outlined in our study proposal. To maintain confidentiality, data containing personal identifiers of subjects were not collected and the data was kept locked.

Current contraceptive methods
As shown in Fig 2,

Factors associated with LARC utilization
All variables with a p-value of less than or equal to 0.2 in the bivariable model was fitted to the multivariable model. Variables included in the multivariable model were wealth status, history of abortion, desire to limit child bearing, knowledge about LARC, convenience of the method of choice, perceived ease of reversibility of the method, easy availability of the method of choice, less frequent visits, advise by health professionals, husband/partner approval, and favorable attitude towards LARC, as each was significantly associated with LARC utilization [ Table 5].  The odds of using LARC were 1.9 times higher [AOR: 1.9, 95% CI (1.1, 3. 2)] among women who were in the fourth quintile wealth status than those in the poorest quintile. Women who had a history of abortion were 2.7 times more likely to use LARC than those who had never had an abortion [AOR: 2.7, 95% CI (1.4, 5.1)]. Women who reported wanting to limit their births were 2.4 times more likely to use LARCs than those who reported wanting to have a child soon [AOR: 2.4, 95% CI (1.01, 5.6)]. Women who had good knowledge about

PLOS ONE
Determinants of long acting reversible contraception utilization Women who chose LARC were 13 times more likely to have favorable attitude towards LARCs than SARC users [AOR: 13.0, 95% CI (8.6, 19.7)] and LARC users were more than 10 times more likely to choose their method due to the advice of health care professionals than SARC users [AOR: 10.7, 95% CI (3.3, 34.9)]. In addition, LARC users were 1.5 times more likely to choose their method due to its ease of reversibility than SARC users [AOR: 1.5, 95% CI (1.0, 2.2)] and it's also 3.0 times more likely to choose their method due to less frequent visits than SARC users [AOR: 3.0, 95% CI (1.9, 4.6)]. On the other hand LARC users were 90% less likely to choose their method due to its ease of availability than SARC users [AOR: 0.10 (0.05, 0.19)] and it's also 77% less likely to choose their method due to its convenience than SARC users [AOR: 0.23, 95% CI (0.16, 0.34)]. Approval of their partner/husband to choose their method need were 34% less likely to occur among LARC than SARC users the [AOR: 0.66, 95% CI (0.45, 0.97)].

Discussion
The purpose of this study was to identify determinants of LARC utilization among women accessing facility-based medical care at public health institutions in Northwest Ethiopia. In this study, higher economic status, history of spontaneous or induced abortion, desire to limit family size, good knowledge of LARCs, less frequent medical visits, advise of health professionals, and a positive attitude of clients toward LARCs were associated with LARC utilization. In contrast, perceived method convenience, easy availability of the method and partner influence, were negatively associated with using LARC.
In this study, clients with the highest income level were found to use LARCs more often than those who earn the least amount of income. This finding is consistent with studies conducted in West Iran, Zambia and Ethiopia in which women who utilize LARC were wealthier than those who did not use LARC [36,37]. In the context of Ethiopia, many wealthy people are relatively educated, live in urban areas, and have easier access to media and information, which could lead to well-informed understanding of the available methods and thus, the use of LARCs. In contrast, a systematic review conducted in France showed that women in difficult financial situations were positive towards IUCD use [31]. One qualitative study exploring obstacles to the use of LARCs among women in Seattle, USA, revealed that expense and billing were major barriers [38]. Taken together, these studies support the notion that both wealthy and poor women in the developed and developing world use LARCs. Differences in patterns of LARC use among these studies can be explained by differences in the study population and settings. From this point, it would be of paramount importance for policy makers and health planners to be focused on strategies that are more feasible and faster to provide LARCs at a subsidized rate while working to empower women on economic activities.
In this study, we found that participants who are currently using modern contraceptives and have history of abortion were 2.69 times more likely to use LARCs than their counterparts who did not have a history of abortion. This finding is consistent with studies set in France and Ethiopia (Adigrat town) in which history of unintended pregnancy was positively associated with LARC utilization [31,39,40]. Patients who had history of abortion may be highly motivated to utilize a very effective method of contraception. This was also evidenced in a study conducted in the United States in which women who were offered immediate post-abortion contraception were more likely to choose an IUCD or an implant than women without a history of a recent abortion [41]. It would be good for health managers to work in creation of public awareness about the effectiveness of LARCs in preventing unintended pregnancy. In this study, women who wanted to limit their birth were 2.38 times more likely to use LARC than those who wanted to have a child more quickly. This finding is consistent with the results of a study that took place in 14 European countries, France and Ethiopia in which LARCs were mainly used by women who had children and did not wish to have any more [30,31,42]. One of the most important benefits of LARCs is their long duration of use once they are placed. For this reason, more women who want to limit their family size may tend to utilize LARCs which are long term, effective and reversible. As such, LARCs may be used as a replacement for surgical sterilization, and because of their ease of reversibility, they may help women to avoid post-sterilization regret.
In the present analysis, lack of knowledge and information about long-acting family planning was found to affect LARC use. Women who had good knowledge about LARCs preferred to use LARCs. This is similar to studies done elsewhere which showed that better/good knowledge led clients to LARCs utilization while lack of knowledge about LARCs was negatively associated with their use [35,40,[43][44][45][46][47][48][49][50][51][52][53].
Several cross-sectional studies conducted in Europe and Africa, as well as a qualitative study in Australia found a similarly negative effect of lack of knowledge on LARC utilization [40,43,46,54,55]. LARCs are highly effective in preventing unwanted pregnancy with great convenience, fewer side effects and less cost than SARCs [22,23,56]. Clients with good knowledge will understand that LARCs are highly effective and safe choices, and therefore utilize them. Activities aimed at increasing the public's knowledge about the benefits of LARCs should be a focus of the health sector.
In this study, the family planning provider's advice on decision-making of clients is found to significantly impact the utilization of LARCs. Clients advised by health care professionals were more likely to use LARCs than those who chose by themselves. This is consistent with a study done in France which showed that professional training and experience was significantly associated with LARC use irrespective of specialty [31,39,57]. Another study conducted in West Ethiopia found that, discussions with health care providers about long acting and permanent contraceptive methods positively affected LARC utilization [42,58].
Moreover, a qualitative study in Australia showed that providers' lack of confidence and support for LARC insertion had a negative effect on LARC use [46]. Providers have a responsibility to clearly communicate and support their clients to choose the method which best fits their personal circumstance. It would be good to provide in-service training for providers on how to support their clients in explaining the effectiveness of LARCs during their counselling session.
In the present study, perceived convenience of the method was less likely to be reported among LARC users, a finding that is counter to two studies demonstrating the importance of convenience in the selection of LARCs as a contraceptive method [30,59]. This discrepancy may be due to rumors, myth and misconceptions, as seen in El Salvador negatively affecting IUCD utilization [60], and fear of procedure related pain and side effects [61] may additionally account for this discrepancy. Advocating for the convenience of LARCs and dispelling myths, misconceptions and rumors should be a strategic approach for increasing LARC uptake.
In this study, clients who were using LARCs didn't choose it due to its easy availability, which is contrary to Rwanda where availability of LARCs was seen to increase implant utilization in postpartum HIV-positive women [62]. Similarly, the issue of access was one of the limiting factors for LARC utilization in Australia [46]. This may be due to, LARCs services being free and easily accessed in Ethiopia, and therefore the issue of availability may not be a concern for clients. The second reason may be lack of awareness. Due to earlier introduction of SARCs, which are more well-known contraceptive methods within the community studied here, clients prefer to use SARCs and may not bother to inquire about the availability of LARCs [26].
Moreover, the influence of the husband/partner also played a role in decision-making which negatively affects LARC utilization in this study. This finding is consistent with studies conducted in the Tigray and Oromia regions of Ethiopia where no or limited support from male partners and lack of discussion between partners was an obstacle to LARC use [34,63]. In another study in the Tigray region of Ethiopia, women who selected their contraceptive method alone were more likely to use LARCs as compared to those who decided jointly with their partners [40]. Similarly, married women with partners who did not permit them to use LARCs, partner's poor knowledge of LARCs, and negative attitudes about LARCs were negatively associated with LARC use. However, this is in contrast to studies done in Uganda, Zambia, West Ethiopia and a case-control study from the Ethiopian EDHS 2011, in which joint decision making, women who live with their partners, and women's attitude that male partners' choice influences their contraceptive decisions were positively associated with current use of LARCs [34,36,37,44,58,63,64]. Some male partners may not understand the benefits of LARCs, and some may even have misconceptions that do not support their utilization.
In this study, women who had a positive attitude towards LARCs were more likely to use LARC than those who did not. This study is consistent with studies completed in the Wolayta zone, the town of Adigrat and the West Arsi zone of Ethiopia which found that women with positive attitudes towards LARCs were positively influenced to use them [40,55,65].

Strength and limitation of the study
This study tried to incorporate as many determinants of LARC utilization as possible. In the present study, selection bias has been minimized because both study cases and controls were recruited from the same health facilities using systematic random sampling.
Temporal relationships can be established since the study used incident cases, and recall bias was not a major problem since the majority of questions inquired about basic healthcare information and personal opinions. However, it is possible that recall bias affected some variables such as age at marriage, age at first birth, history of ANC and PNC. The findings of this study may not be generalizable to community members who do not attend health facilities.

Conclusion and recommendations
Professional support, favorable attitude towards LARC use, high economic status, history of abortion, advantage of less frequent visits, having good knowledge towards LARC and interest of limiting births were significantly associated with LARC Utilization. Whereas perceived method convenience, and contraception availability were inversely associated with it. Family planning education about the benefits of LARC should be done by health providers and media. Male involvement in the counselling and decision making about the advantage of using LARC may improve the negative influence of partners on LARC utilization. It is also recommended that, future researchers can try to explore the deep perception of LARC users using qualitative method.
Supporting information S1 Tool. Tool used for data collection. (ZIP) S1 Data. Data underlying the study findings. (DTA)