This study has documented the level of knowledge of, and factors associated with family planning use among a PLHIV population in the resource-poor, post-conflict region of Northern Uganda. We found a very low level of current family planning use despite a high level of knowledge about contraceptive methods. Factors associated with using family planning methods in this PLHIV population included having ever gone to school, discussion of family planning with a health worker or with one's spouse, not attending the Catholic-based clinic and spouse's non-desire for children. Discussion with a spouse have also been found to be associated with use of hormonal contraceptives in Rakai, Uganda . Religion also has an impact on the uptake of contraception , through its influence at both the individual level and the institutional level, where faith-based health facilities may not directly provide family planning services to clients, thus limiting the access by PLHIV to these services.
Fear of side effects, reduction in pleasure, misinformation, negative perceptions, and gender-inequality have also been identified in other studies as barriers to adopting family planning [14–16]. As found in other studies [5, 17], male sterilization was not used: Strong aversion to vasectomy has been linked to fear of male impotence in some societies [18, 19], and/or reluctance to terminate males' reproductive career . Our study also showed low use of dual methods of contraception among PLHIV. Use of a barrier method in combination with other contraceptives maximizes contraceptive efficiency and reduces the risk of HIV transmission to sexual partners .
PLHIV in our study who did not desire to have more children were often unable to access the family planning services they needed. The lack of association between desire to have children with use of family planning methods in this PLHIV population could be explained by the structural barriers that exist in Northern Uganda as a consequence of the long period of conflict in the region, which led to the outmigration of skilled health workers, the limited number of existing family planning clinics, and lack of provision of family planning services within the HIV clinics. The generally low level of contraception use may be explained by the high level of desire for children in this population which may arise from esteem associated with large families , and low levels of female autonomy and literacy.
The strong desire to have children in this population may be further influenced by the prolonged civil conflict and high levels of infant and child mortality. Families, including couples living with HIV, which have lost their children during the conflict to either disease or violence, may have a strong desire to have more children. In societies with low literacy, endemic poverty, high child mortality and lack of social welfare and security programs, children are considered as a form of insurance to provide support in old age. Furthermore, having children in Uganda increases a person's social status  and this also applies to couples living with HIV.
Family planning programs and health workers mainly target women for family planning, but it is apparent that this approach did not result in discussion with their spouses or uptake of family planning services. Whether or not condoms were used was very much determined by the male spouse, particularly when the relationship was unstable. Our study showed that proportionally more females than males had discussed family planning with health workers. However, females generally reported not having discussed family planning with their spouse, whereas males reported high levels of spousal discussion on family planning, suggesting the focus of such discussions may have a different perspective for males and females. Fewer women than men reported using any method. Considering that men are the reproductive decision-makers in most traditional Ugandan homes , it is essential that reproductive health services also target men, educate them, and involve them in reproductive educational programs.
The ecological framework, as applied in this study, views the use of contraception among PLHIV as the outcome of interaction of factors at several levels: individual, interpersonal, and structural. At the individual level factors include demographic factors such as education status, sex, as well as personal attitudes and experiences of contraception. At interpersonal level, discussions and interactions with health workers, and spouses impact on the use of contraception. At the structural level, limited provision of family planning services in the general population and lack of integration of these services within HIV clinics inhibited the use of contraception among PLHIV. The usefulness of this framework is that it allows development of multi-level strategies to address the issue. Understanding the interdependency of factors at each level allows a holistic, and more effective approach to improving access while taking into account broader public health considerations.
Integration of family planning services with HIV services utilising a multi-level approach to improve the uptake is urgently needed in this region. Family planning programs should cater to PLHIV who wish to limit their family size, and also to those who wish to continue to have more children with a goal of achieving better health outcomes for the PLHIV through birth spacing and use of effective and safe contraception. Such integration has potential not only to improve reproductive health outcomes [21–24], but to ultimately reduce paediatric HIV infections , and hence reduce the amount of antiretroviral therapy needed. This is particularly important in countries such as Uganda where MTCT at 18% of new infections is a major route of HIV transmission .
Several levels of integration are possible. Family planning education should be provided within the HIV clinics and integrated into routinely provided general education programs with information on the effectiveness, safety, and possible side effects of all contraceptive methods. Doctors, nurses, and community workers attached to the HIV clinics could be trained in family planning counselling for PLHIV, and contraceptives could be provided free. Health workers can facilitate discussions of family planning with couples, either at health facilities or in the communities, and by doing so they can assist women in broaching the subject to their spouses and hence improve family planning use. HIV clinics have regular and prolonged contact with HIV-infected clients, and are ideally placed to meet their reproductive health needs over time . While there has been some success in integration at PMTCT clinics , this is a temporary contact with HIV-infected clients that lasts only for the duration of pregnancy. Women generally do not return for post-natal family planning counselling , and PMTCT clinics target only women, whereas HIV clinics can target both men and women.
Family planning services can also be provided at the facility level, where clients are referred to separate clinics within the same health facility. It is also possible to have an active district-wide referral and follow-up service so that clients are appropriately referred to facilities that provide the service. Faith-based health facilities that may not directly provide family planning counselling and services can become part of a referral network. Although no difference was seen in this study between respondents' family planning discussions with health workers by the clinic they attended, actual use of family planning methods were significantly different, suggesting a need for active referral systems. Surgical contraceptive services should be readily available, sustainably funded, and provided by locally-trained doctors who could also deliver services at more remote clinics on a rotational basis. Nursing staff, in collaboration with community village health workers, could counsel and prepare clients for operations that are available on a regular schedule. The suggested measures could be coordinated and implemented by the local district health departments in collaboration with health facilities, local community organizations, government agencies, and UN partners. Though possible constraints include lack of time due to large client numbers and commodity shortages, local government health departments could determine funding sources, training requirements and implementation strategies.
This is the first study on family planning use among a PLHIV population in a conflict/post conflict region and it adds to the literature on family planning use among male and younger PLHIV. The majority of previous studies have examined family planning use among women only. Information from females alone is insufficient, particularly in the context of a patrilineal and male-dominant society. By documenting use of family planning among males, their access to and perceptions of its use, a clearer and more holistic picture of why their spouses may or may not be using contraception is revealed. The sampling approach also ensured that the outcomes of interest (family planning use) could be assessed on adequate numbers of males and females in the different age groups as well as allowing statistical comparisons across sex and age groups. Additionally the combination of quantitative and qualitative methods has provided important information about the use of family planning methods. The quantitative findings provided us with information on the level of knowledge of and use of family planning among this PLHIV population and reveal the variables independently associated with the use of family planning. The qualitative data highlight gender inequality and limited access to and poor quality of available contraceptives as important contributing factors for the low use of family planning among PLHIV. The qualitative methods also allowed for exploration of additional concepts not captured in the survey questionnaire, such as covert use of contraceptives by women and targeting of women by family planning programs.
Limitations of this study include the cross-sectional design and, hence, causality cannot be determined. The non-random sampling and recruitment at the health facilities also result in a bias towards clients who are able to access health facilities, who are more urban-based or wealthier than those who had no access. The younger respondents aged 15-19 years and male respondents may have been more prone to positive health-seeking behaviours than their counterparts in the general population. Social desirability bias may have occurred when respondents were interviewed: PLHIV may feel that they have to indicate that they are using condoms to prevent further spread of the infection, especially if condom use has been previously promoted by health workers. While the ratio of males to females in this sample is similar to that in the general HIV population in Northern Uganda, caution needs to be exercised in generalizing findings to the general HIV population. Nevertheless, the findings provide important information about factors that are associated with use or non-use of family planning methods and, despite the unique complexities of this post-conflict region, may have implications for HIV populations elsewhere.
Future studies could consider comparison of HIV-infected with non-infected clients to determine the impact of HIV on access to family planning and its use. Research on the general PLHIV population is needed to measure unmet needs for family planning services among PLHIV. Interviewing couples separately to ascertain reported condom use is recommended for future research.