Globally, nearly 34.0 million people are living with HIV of which almost 50% are women [1]. HIV is also the major cause of mortality among women of reproductive age [2]. Worldwide every year, about 2.5 million women who become pregnant are HIV positive and an estimated 3.34 million children are living with HIV [3,4].

Sub-Saharan Africa bears the largest share of the global burden of HIV where 59% of adults living with HIV are women. Women aged 15 to 24 in the region are between 2 to 6 times as likely to be infected as young men of the same age [3]. Mother-To-Child Transmission (MTCT) of HIV is an important contributor to HIV transmission. In 2012 an estimated 260 000 children were newly infected with HIV, and an estimated 3.3 million children were living with HIV globally [5]. Over 900 children get infected with HIV each day and most of them get the infection from their mothers either during pregnancy, childbirth or breastfeeding [6]. In 2011, nearly 390,000 women over the age 15 and 180,000 children under 15 were living with HIV in Ethiopia [7].

To prevent the transmission of HIV from mother to baby, the World Health Organization (WHO) promotes a four pronged approach: Primary prevention of HIV infection among women of childbearing age; preventing unintended pregnancies among women living with HIV; preventing HIV transmission from a woman living with HIV to her infant; and providing appropriate treatment, care and support to mothers living with HIV and their children and families [8].

HIV counseling and testing during pregnancy and the provision of antiretroviral drugs during delivery and following birth are the two primary approaches to Prevent Mother To Child Transmission of HIV (PMTCT). Study done in the President’s Emergency Plan for AIDS Relief (PEPFAR) focus countries in 2008 has shown that, contraception as an HIV prevention strategy is cost effective compared with other PMTCT approaches. Hence, a contraceptive strategy prevents 28.6% more HIV-positive births than nevirapine [9].

The occurrence of unintended pregnancy among HIV positive women was seen to be significant in different setups. In the United States in 2014, of 382 pregnancies that occurred among HIV positive women, 326 (85.3%) of them reported ≥ 1 unplanned pregnancy [10]. A study done in Pune, India, has discovered that HIV-positive women are nearly four times more likely to not have planned their most recent pregnancy as compared to HIV negative women. As would be expected, HIV-positive women were four times more concerned about this pregnancy (Due to the risk of HIV transmission) than HIV negative pregnant women. Repeat pregnancies were more likely to occur for women who did not disclose their HIV status to their spouse. Thus, the majority of the repeat pregnancies for HIV-positive women were both unplanned and unwanted [11].

In 2008, in Cross River State, Nigeria, FP was integrated into Antiretroviral (ART) services in five local government areas composed of two groups: the basic model and the enhanced model. The basic model consist of capacity assessment of FP service providing facilities, FP training for ART providers and provision of M&E tool to track FP referrals and FP service provision in three of these areas. The enhanced model, consist of training of FP providers from ART, VCT and PMTCT service outlets, facility level regular supportive supervision and mentorship and community mobilization in the other two facilities. Baseline interviews were conducted in 2008 and follow-up interviews were also made 12-14 months later with 274 female ART clients aged 18-45 across the five areas. Unmet need for contraception was high at baseline (28-35 percent) but modern contraceptive use rose in the enhanced and basic groups. Most of the increase was in consistent condom use. Despite an increase in FP counseling by ART providers, referrals to family planning services for non-condom methods were low. The conclusion reached was to provide an alternative strategy for FP/HIV integration in settings where large families and low contraceptive use are normative phenomena [6].

With an objective of measuring changes in service utilization of a model integrating family planning with HCT, ART, and PMTCT in Nigerian public health facilities, a retrospective survey of attendance and family planning commodity uptake in 71 health facilities analyzed pre-integration and post-integration periods between March 2007 and January 2009. Mean attendance at family planning clinics increased significantly from 67.6 clients per day pre-integration to 87.0 clients per day post-integration. The mean Couple Years Protected (CYP) increased significantly from 32.3 pre-integration to 38.2 postintegration. The service ratio of referrals from each of the HIV clinics was low but increased in the post-integration period by 4, 34 and 42 per 1000 clients from HCT, ART and PMTCT clinics, respectively. Service ratios were higher in primary healthcare settings than in secondary or tertiary hospitals. Attendance by men at family planning clinics was significantly higher among clients referred from HIV clinics. The conclusion made was that FP/HIV integration improved family planning service utilization by clients accessing HIV services, but further improvement was possible. Male utilization of family planning services also improved [6].

A cross sectional study that was conducted between December 2010 and February 2012 in six government hospitals of Addis Ababa, Ethiopia assessed unmet reproductive health care needs and occurrence of unintended pregnancy among HIV positive women enrolled in the ART units (n=548). The results showed that unmet need for contraception was 31%: 25% for spacing and 6% for limiting. Generally, HIV positive women who had a higher chance of unmet need for contraception were those with a sero-discordant partner [AOR (95% CI) =2.4 (1.04-5.64)] and women who had faced unintended pregnancy after being HIV positive [AOR (95%CI) = 10.1 (4.6-22.3 [7].

Between November 2006 and February 2008, Pathfinder International-Ethiopia initiated integration of family planning services into HIV services through sensitization of administrative staff and health care providers on the importance of integrating FP with HIV services in semi-urban hospitals and health centers of Oromia region. A five-day long training was given based on a curriculum that included basic information on the benefits of FP, contraceptive options, sideeffects, counseling messages and protocols developed specifically for VCT clients. Pre (n=4019 clients) and post (n=4027) family planning intervention cross sectional surveys on client’s contraceptive practices and needs were conducted in 2006 and 18 months later, respectively. It was found that among married and other sexually active women, 70% were using contraceptives. Of women in current sexual unions, 17% had unmet need for contraceptive. The authors concluded that despite the relatively low need for family planning services among study clients, there was an impressive increase in the provision of family planning information in VCT [12].

Integrating Family Planning services into HIV prevention, treatment, and care services provides an opportunity to increase access to contraception among clients of HIV services who do not want to become pregnant, or to ensure a safe and healthy pregnancy and birth for those who wish to have a child. However, data regarding the outcome and effectiveness of integration strategies in addressing the contraceptive needs of women living with HIV are limited. This study therefore tries to assess the effect of FP/HIV service integration, focusing on the integration of FP with ART within the Integrated Family Health Program (IFHP) support areas, which is a five years program supported by USAID.

Study area, study design and study population

With the purpose of assessing the effect of FP/ART/HIV service integration, facility-based comparative cross-sectional study was conducted in IFHP supported institutions in four regions of Ethiopia namely; Oromia, Amhara, Tigray and Southern Nations Nationalities and People’s Region (SNNPR). Data on women living with HIV and attending ART clinics was collected from April 23 to May 2, 2012 from two types of IFHP supported facilities that were to be compared: facilities that were identified as well performing in terms of integration of FP/ART/HIV services and facilities where FP/ART/HIV has not yet been integrated.

Sample size and sampling technique

The sample size was determined using a formula for comparison of two population proportions. A key outcome of interest was unmet need among ART clients. The sample size calculation was based on a national figure of 25% unmet need for family planning among married women [13], 95% confidence level, margin of error of 5%, and power of 80%. It was assumed that there would be a 6% difference in unmet need between health centers with and without FP/HIV integration. Accordingly, the total sample size, including a non-response rate of 10% was calculated to be 822, each for the FP/HIV integration and non-integration sites.

IFHP supported facilities that were identified as well-performing in terms of integration of FP within HIV and AIDS services in the selected regions of Ethiopia were purposefully selected as integration sites. Equivalent number of facilities where the integration did not take place, but with similar characteristics as the integration sites (tier level, type of service, client flow, accessibility and health facility’s capacity), were purposively selected as non-integration sites. A total of 40 facilities; 20 facilities from the integration sites and 20 facilities from the non-integration sites were selected.

Eligible women clients in child bearing age were interviewed until the calculated sample size was reached using exit interview technique from both integration and non-integration sites.

Data management and analysis

A structured questionnaire was used to assess the level of contraceptive use, types of FP methods accessed and level of unmet need. The data collection instrument was developed using a tool that was applied for a similar survey in Uganda. The tool was translated in to the local languages spoken in the study areas. A pre-test was carried out during the training of data collectors before the tool was finalized. The pre-test was carried out in nearby facilities that were not part of the study but had similar characteristics with the identified facilities. In the pre-test, questions with ambiguity were recorded and amended.

Forty (40) data collectors, 40 supervisors and 4 regional research coordinators were recruited and a 2-day training was provided on the data collection technique, quality control and related issues. The research team developed a training manual that addressed the objectives of the study, data collection tools, data collection procedures, and data handling mechanisms. The team led the process in conducting training, supervising the overall data collection procedures and handling administrative issues.

Review and editing of the questionnaires was carried out on a daily basis by the supervisors in order to identify errors, omissions and inconsistencies. Data were checked for completeness, accuracy and clarity by the survey core team of researchers and supervisors.

Data were entered and analyzed using SPSS version 17 (2008. Chicago: SPSS Inc.). Data cleaning and outlier checks were done. Descriptive statistics including frequencies, graphical presentations; and associations using appropriate statistical tests were computed. Mean differences were tested using two sample t-tests. Odds ratios and 95% CI were used to measure the strength of associations between the independent and outcome variables.

Ethical considerations

Before the commencement of the study, ethical clearance was obtained from the Institutional Review Board (IRB) of the College of Health Sciences of Addis Ababa University. During training of supervisors and data collectors, ethics of data collection was addressed in detail emphasizing voluntary participation, privacy and confidentiality. All interviews were conducted in places where the privacy of the study participants was maintained. Each participant was briefed about the purpose and importance of the study before the initiation of interviews. The respondents were told that their participation was entirely voluntary and that whether they choose to participate or not, all the services they would normally receive will continue. Verbal consent was obtained from each study participant.

total of 1,515 women (822 in the integration and 693 in the nonintegration sites) were participated in the study. The overall response rate was 99.9%, 100% in the integration and 99.7% in non-integration sites.

Socio-demographic characteristics

The mean age of respondents (mean ± SD) was 31.8 ± 7.2 and 31.74 ± 7.3 years in the integration and non-integration sites, respectively. Nearly half of the women in both groups were not married and had no education. Higher proportion of women in the non-integration sites had secondary or higher level of education than in the integration sites. The majority of the respondents in both groups were orthodox religion followers. The percentage of Protestants and Muslims in the integration sites was higher than the non-integration sites. There mean age between the two sites was almost the same (Table 1).

Table 1: Socio-demographic characteristics of women in IFHP integration and nonintegration sites, July 2012.

Knowledge and current use of FP by clients

Knowledge of clients about the safe use of FP was high in both groups with 1,318 (94.5%) women reporting that women with HIV can safely use FP. Seven hundred forty one (95.7%) and 577 (93.1%) the women in the integration and non-integration sites, respectively, reported that women with HIV can safely use FP. Adjusted for other factors the difference was not statistically significant (P>0.05).

Nearly half (48%) of all women were using modern FP method. The Contraceptive Prevalence Rate (CPR) in the integration and nonintegration groups was 52.6 % and 42.9%, respectively. This difference in CPR was statistically significant [(AOR (95%CI)=1.56 (1.23,1.920)]. Out of 334 contraceptive users ( with the exclusion of condom use or dual protection) in the integration sites, nearly 66.7% of women were using dual methods unlike 55% of women in the non-integration sites, indicating higher proportion of women in the integration sites using dual methods. This difference was statistically significant [AOR 95% CI 1.50 (1.01, 2.2)] (Table 2).

Table 2: Contraceptive use at the integration and non-integration sites, July, 2012.

In both sites, the majority of women (74%) in the age group of 15- 34; women who were married and in union and those who had primary or higher education reported that they had discussed the use of family planning with their providers when receiving HIV care and support services. On the other hand, except women with no education in the integration sites, 83% and higher proportion of women in the different age groups, marital status and level of education in both integration and non-integration sites had discussed about special consideration for using a contraceptive method (Table 3).

Table 3: Percentage of women who discussed FP with health service providers by background characteristics at integration and non-integration areas, July, 2012.

Comparing the types of family planning methods used in the two study groups, 249 (55.8%) women in the integration and 162 (52.7%) in the non-integration sites were using injectables followed by male condoms, 93 (20.9%) in the integration and 58 (18.2%) in the nonintegration group (Figure 1).

Figure 1: Percentage distribution of current use of family planning methods in integration and non-integration sites. July 2012.

Most of married women (70.8%) in the non-integration and most of women in the integration sites who reported to be in union (79%) were using some type of contraceptive methods. Older women (35- 49) tend to use permanent and long acting methods compared to the younger age women (15-24) in both categories (Table 4).

Table 4: Contraceptive use, by method and by background characteristics in the integration and non-integration sites, July, 2012.

Fertility desire and unmet need for FP

One hundred twenty three (8.1%) women said they would want to wait for two years or more before having a birth and 775 (50.7%) said they would not want to have other children. Of these 55 (7.1%) were using long acting or permanent methods (sterilization, IUD Implants).

The total unmet need was 16.2 percent, 15.7 percent in integration and 16.9 percent in the non-integration sites. Total unmet need was not statistically different between the two groups [(OR (95%CI)=0.94 (0.63, 1.390)] (Table 5).

Table 5: Unmet need, met need and total demand by integration and non-integration group, July 2012.

The met need was 73 percent in the integration sites and 67.6 percent in the non-integration sites. The total unmet need for spacing and limiting in the study population was 3.7 percent and 10.3 percent, respectively. In the integration sites, unmet need for spacing was 4.8 percent whereas for limiting was 8.8 percent. In the non-integration sites, unmet need for spacing and limiting were 2.3 percent and 12.4 percent, respectively. There is also a high total demand for family planning among women living with HIV in both the integration (88.4%) and non-integration (83.2%) sites (Table 5).

Almost all invited respondents participated in the study. There was no statistically significant difference in age distributions between the integration and the non-integration sites. A higher proportion of women in the integration group were married. Illiteracy was high in both sites. However, lower proportion of women in the integration sites were illiterate whereas a higher proportion of women in the non-integration sites had secondary education and above. The fact that half of the women were illiterate could be because this study was done in rural areas where women had limited access to education [13]. Similarly, the high proportion of married women could be because majority of the women included in this study were above 25 years in addition to the early marriage practices that is common in the study areas [13].

Knowledge of women about the safe use of FP was considerably high with 1,318 (95%) of the women reporting that women with HIV can safely use FP. This is encouraging and indicates that given comprehensive information and access to quality family planning services, contraceptive coverage and unmet need can improve a lot in both populations. This finding is similar to the results of a study conducted in Uganda [9]. Overall contraceptive prevalence rate (CPR) in the study was found to be 48.2% which is higher than the national average [13]. The CPR in the integration and non-integration sites was 52.6 % and 42.9% respectively. Adjusted for other factors, the difference in CPR was statistically significant indicating women in the integration sites were more likely to use contraceptives than women in the nonintegration sites. Thus, integration of services is highly likely to increase contraceptive use. This is in accordance with the result of a systematic review of peer-reviewed articles which indicated that FP integration with HIV interventions always resulted in positiveor mixed results for key outcomes; but no negative results [14].

Percentage of women who had discussion with the provider on family planning both in the integration and non-integration sites is high and with variation by socio demographic characteristics. This could be because of service providers bias in that they might advise HIV positive women not to become pregnant using contraceptives lest vertical transmission of the virus and infecting their partners. For the same reason, HIV positive women may seek counseling on family planning to avoid pregnancy.

Two hundred forty nine (55.6%) women in the intervention and 162 (54.0%) in the comparison group were using injectables followed by male condom 93 (20.8%) in integration and 58 (19.3%) in the nonintegration sites. This finding showed a lower condom use by PLHIV when compared to a facility based study done in Kenya, where male condom was the most popular (81.5%) contraceptive method used [15].

Dual method of contraception was practiced by 13.5% of the respondents. This is lower than reported by another similar study conducted in India [16]. Where condoms were the most common contraceptive method, with the prevalence of 92%.The lower prevalence of condom use in the current study could be explained by the limited acceptability of condom as contraceptives among HIV positive women [17]. However, a higher proportion of women in the integration sites were using dual methods than women in the non-integration sites, which shows the valuable contribution of integration in promoting the recommended family planning methods for HIV positive women.

The total unmet need for FP in the study was 14.1% which is much lower than the national average [18]. The total unmet need in the integration and non-integration sites was 13.6% and 14.7% respectively. The study found that total unmet need was not statistically different between the study groups (p>0.05). Thus, the integration does not seem to make a difference in decreasing unmet need for family planning. This can be at least partly explained by the fact that unmet need may not be affected for certain period of time when more demand is created for FP in the community as a result of community mobilization efforts. The fact that about 51 % of women who were not using family planning methods were either not currently having sex or they were at the age of menopause or are infecund could also account for the low level of unmet need for FP in this study. This is similar to the findings of five-countries study of FP and HIV integrated services conducted in Kenya, Uganda, Ethiopia, South Africa and Rwanda where unmet need for FP was never above 18% which was due to the high proportion of women who reported no need of FP because they were not sexually active, as well as some wanted to become pregnant or desired their current pregnancy [11]. On the other hand, it is important to find out what factors (cultural, institutional) were responsible for lack of effect of integration on unmet need while it was found that integration had significant effect in improving contraceptive coverage. The study could not claim to have controlled for all possible confounders since this was not a randomized controlled trial.

Of the 775 women who said they didn’t want to have any children, only 55 (7.1%) were using long acting and permanent methods (LAPMs). The factors for this low utilization from the supply, demand or both sides need to be explored further.

As revealed in this study, integration has improved CPR, where higher rate of CPR was observed among women in integration sites compared to women in the non-integration sites. Thus, integration of family planning in to ART service outlets is important in improving the uptake of family planning methods. Similarly, with higher proportion of women in the integration sites reporting use of dual methods than women in the non-integration sites, integration has also contributed to increased use of the recommended family planning method for HIV positive women. Therefore, given the valuable contribution of integrating family planning services in to ART outlets, integration need to be scaled-up to other facilities.

It is premature to conclude now that integration of family planning with HIV doesn’t contribute significantly to reductions in unmet need. More time may be required to see its effect on unmet need in the integration sites. However, it has been noticed that unmet need for family planning among the study population (HIV positive women) is much lower than the national average. While this is an encouraging finding, the issue may need to be explored further.