The decision whether or not to have children is often complex and influenced by many factors. These include the number of prior children, the socio-economic situation of the family, perceptions and opinions of partners/family members about childbearing, family social structures and the availability of contraceptive services and supplies [1–4]. HIV-positive individuals have additional considerations, largely negative, to take into account when deciding whether or not to have children. These include the possibility of passing HIV from mother to child and the likelihood that one or both parents could die prior to the child reaching adulthood . Furthermore, HIV-positive women may be less fertile and may be more likely to experience adverse maternal outcomes [6,7].
Regardless of these concerns, many people still elect to have children after receiving an HIV diagnosis for various personal, cultural and economic reasons . In most places in Africa, a common expectation of marriage is that the couple will have children. This is an especially important expectation in Uganda as children become members of the paternal clan [9,10]. Women are often valued for their ability to bear children and a high social good is placed on fertility. Because of this value and these expectations, women and couples may continue childbearing even when they are HIV-infected or destitute in order to avoid social stigma and isolation.
Though there is evidence that HIV/AIDS reduces women's fertility, the relative contributions of social, biological and behavioral factors have not been fully elucidated [11,12]. Overall evidence from Africa as to whether or not HIV influences reproductive decision-making is mixed. Some studies have shown that an HIV diagnosis causes people to choose to have fewer children; however, changes in fertility are generally small and the desire to limit childbearing often only exists for those who already have children [1,9,13]. Other research has shown that HIV/AIDS does not have a marked impact on fertility decisions, particularly for those who do not show signs or symptoms of disease [2,8,14]. As of yet, there have been few studies, which have specifically studied the effect of an HIV diagnosis on reproductive decision-making in Uganda. Uganda may differ from other sub-Saharan African countries given its extremely high fertility rate and its aggressive efforts to reduce HIV transmission .
We conducted a literature review of the PubMed database, using the keywords ‘HIV’, ‘family planning’, ‘desire for children’ and ‘childbearing’. We found five articles directly related to HIV infection and desire for children in African countries: two from Uganda [16,17], one from Malawi  and two from South Africa [19,20]. Overall, the results of our literature review revealed the shortage of published material in regard to HIV infection and childbearing decisions. To further investigate the association between HIV status and desire for future children, we undertook a study on this topic in the Kabarole and Kamwenge districts in western Uganda in 2006. The purpose of the study was to examine whether and how a negative and positive HIV diagnosis generally influences reproductive desires, behaviors and the decision-making processes in individuals living in rural districts in western Uganda; to investigate in further detail the effect of a positive HIV diagnosis on individuals' desire to stop childbearing; and to describe the contraceptive behavior and the types of contraceptives used by HIV-positive individuals in this geographical area.
The study used a mixed design that combined a cross-sectional, quantitative component with a qualitative component. In this paper, we report the findings from the quantitative survey, which used a structured questionnaire administered through interviews to gather data from participants.
Participants were recruited through three health centers located in Rwimi and Kibiito sub-counties in the Kabarole district and the Bigodi sub-county in the Kamwenge district. The health centers were located along two major roads and are government-run. They offer clinical and public health services, as well as voluntary counseling and testing (VCT) and prevention counseling for mother-to-child transmission of HIV/AIDS. The Rwimi and Kibiito Health Centers also offer free combination antiretroviral therapy (cART) for HIV patients.
Recruitment of participants
The study inclusion criteria were age 18–44 years, married or cohabitating with a partner, having an HIV test result and a known village address. Persons who were bedridden were excluded from the study. HIV-positive and HIV-negative persons were selected from the VCT registries from two health centers (Rwimi and Kibiito) using systematic sampling, whereby all consecutive persons who had been HIV tested were selected. In order to increase the sample size, all HIV-positive individuals of an HIV patient support group in the Bigodi sub-county were also included in the study. Spouses of HIV-positive participants who met the eligibility criteria were invited to participate. Individuals were initially contacted by healthcare workers to seek their participation in the study. Participants had the option of being interviewed either at their home or at their local health center. We had determined that a sample size of 198 participants in each of the HIV-positive and HIV-negative group would enable us to detect a difference of 10% in the main outcome variable between the two groups if the point estimate of the main outcome variable was low (10%) and a minimum difference of 15% if the point estimate was high (50%), with α = 0.05 (two-tailed) and β = 0.20. In order to allow for nonresponses, we enrolled a total of 421 participants in the study.
Data collection and analysis
A questionnaire to collect socio-demographic characteristics and information on reproductive decision-making, HIV testing, HIV status and fertility desires, contraceptive use and methods of use and attitude toward childbearing of HIV-infected women/couples was developed in consultation with local experts. Most questions used were derived from published sources and were already tested for its reliability and validity, an example being the Demographic and Health Survey in Uganda . The final questionnaire was translated into the local language, Rutooro, and back translated into English for linguistic reliability and pretested in the study area on 15 persons who were not part of the study. The instruments' reliability was assessed through a test–retest exercise of 26 randomly selected participants 7 days after the questionnaire was first administered. The overall agreement obtained in the retest was 92.4% for all questions. For the most important question referring to the main study outcome variable ‘Do you want more children’, agreement was 96.2%. (For those participants who reported to be pregnant at the time of the interview, this question was phrased as ‘Do you want more children in addition to the current pregnancy’). The study questionnaire was administered by trained interviewers in the local language. The interviews lasted around 40 min.
Interview data were entered into Microsoft Access and then transferred into STATA version 9.0 for statistical analysis . A P value of less than 0.05 was considered to be statistically significant. Data from open-ended survey questions were coded by the researcher and analyzed using descriptive statistics. A χ2 and independent samples t-tests were used for bivariate data analysis. Logistic regression was used to model the variable ‘desire to stop childbearing’ with a binary outcome (yes/no) and the HIV serostatus as the main covariate of interest. Independent variables included demographic and socio-economic characteristics as well as various HIV-related factors such as the serostatus of the respondent's partner, experience of any AIDS-related symptoms or illness and whether the respondent was on ART. All the independent variables significant at P value less than 0.2 in bivariate analyses and confounded variables (e.g. sex) were selected and fit into a multivariate model. Variables found to be statistically significant in the multivariate model (P < 0.05) were kept in the final model. The model was adjusted by forcing economic status (as estimated by the quality of the residence), education and being pregnant at the time of the survey into the final parsimonious model. Two multivariate sub-analyses were done, one with men and women separately and one with pregnant (at the time of the interview) vs. nonpregnant individuals or spouses separately. All logistic regression models were checked for interactions.
Study approval and ethical considerations
Ethics approval was provided by the University of Alberta's Health Research Ethics Board Panel B. In Uganda, approval for the study was obtained from the Uganda National Council of Science and Technology, Kampala. The study was also approved by the District Officer of Health, Kabarole and Kamwenge districts, and the political representatives of the study areas. Each participant was informed about the study with an information letter read and handed out. All participants signed a consent form.
Demographic and socio-economic characteristics of study participants
The survey included interviews with 421 participants: 199 (47%) were HIV-positive and 222 (53%) were HIV-negative. The participation rate was 92%. Approximately, two thirds (64%) of the study participants were women. The majority of the study respondents were subsistence farmers with a primary level of education living in metal-roofed mud huts. The participants had a wide range of tribal and religious affiliations. HIV-positive participants were on average older and had higher parity than their HIV-negative counterparts. About half (46%) of the respondents cared for nonbiological children. Ten percent of HIV-positive women reported to be pregnant at the time of the survey compared with 66% of the HIV-negative women (P < 0.001). Similarly, 11% of HIV-positive male participants reported that their female partners were pregnant at the time of the survey compared with 34% of HIV-negative male participants who said that their female partners were pregnant at the time of the survey (P < 0.05) (see Table 1).
Overall, it was found that 35% (n = 145) of all participants wanted to have more children. HIV-positive individuals were much less likely to want more children (14% HIV+ vs. 53% HIV−, P < 0.001). Those who had two or more living children were also less likely to want to continue childbearing than those who had one or no living children (21 vs. 75%, P < 0.001). Of the HIV-positive participants, those who were on cART (n = 77) were more likely to want more children compared with those not on cART (18 vs. 11%); however, this difference was not statistically significant (P = 0.131). Those who wanted to continue childbearing wanted an average of 2.4 additional children (2.6 for HIV-negative vs. 1.6 for HIV-positive participants). The average number of children desired (completed family size) was 4.0 for HIV-positive participants compared with 4.6 children for HIV-negative participants (P = 0.003). Within 25 couples, each partner was independently asked how many more children he/she wanted and how many more children his/her partner wanted. Only 32% of respondents could correctly identify how many more children their partners wanted; 32% overestimated their partners' fertility desire and 36% underestimated it.
Bivariate and multivariate logistic regression analysis results are shown in Table 2. Results from the multivariate logistic regression analysis indicated that the odds of wanting to stop childbearing were 6.5 times greater for HIV-positive than for HIV-negative individuals after controlling for age, sex, education, ethnic group, economic status as estimated by the quality of the residence, current pregnancy and the number of living children. In addition to a positive HIV status, older age [odds ratio (OR) 1.13, P < 0.001], female sex (OR 2.42, P = 0.027), Mutooro ethnic group (OR 3.20, P = 0.006) and a greater number of living children (OR 1.62, P < 0.001) were significant predictors for wanting to stop childbearing. Other important but not statistically significant variables included in the parsimonious model of desire to stop childbearing were economic status, education and being pregnant at the time of the survey.
As there was a large discrepancy between the number of pregnancies by HIV serostatus, a multivariate sub-analysis was conducted comparing individuals (or partners of individuals) who were pregnant vs. nonpregnant (Table 3). Both those respondents (or their partners) who reported to be pregnant and those who reported not to be pregnant during the interview showed a positive association between the desire to stop childbearing and a positive HIV status (i.e. for both groups, the OR pointed in the same direction). However, as the sample size of those respondents who reported that they were pregnant was smaller, this association did not reach statistical significance (3.24, P = 0.200), whereas the association in those who reported not to be pregnant was statistically significant (8.14, P < 0.001).
Similarly, a sub-analysis was conducted to compare the fertility desires of men and women separately as HIV status could have a different impact on the fertility intentions of each sex given the differences in the roles that men and women play in childbearing (Table 3). When male and female respondents were analyzed separately, the main association between the desire to stop childbearing and a positive HIV serostatus was confirmed in both groups (OR for men 10.3 and for women 2.9), confirming the main overall result in this paper that a positive HIV serostatus is positively associated with the desire to stop childbearing. As the sample size in the sub-analysis was much smaller, this association did not reach statistical significance for female respondents, though it did reach statistical significance for the male respondents (see Table 3).
Sixty-seven percent of all HIV-positive individuals (n = 199) who wished to stop childbearing mentioned their positive HIV status as one of the main reasons for this decision. Thirty-three percent of HIV-positive respondents reported that they would have had more children if their test result had been negative instead of positive. HIV-negative respondents were also asked a hypothetical question as to whether or not their fertility desires would change if their test result had been positive rather than negative. Fifty-five percent of those individuals stated that a positive HIV diagnosis would change their fertility desires completely, whereas only 6% said that it would not. Thirty-nine percent stated that it would not influence their fertility desires because they had already achieved their desired family size. These differences between HIV-positive and HIV-negative individuals were not statistically significant.
Almost all participants (93%) stated that HIV-infected couples should not have children. There was no difference in the responses by HIV-positive and HIV-negative participants in their negative attitude toward childbearing by HIV-positive couples or individuals.
Use of family planning methods
In total, 42% (n = 176) of respondents were currently using contraception. HIV-positive participants (74%, n = 148) were much more likely (P < 0.001) to be current users of contraception than HIV-negative participants (74 vs. 13%, P < 0.001), and men were more likely to be current users than women (53 vs. 36%, P = 0.001). However, most HIV-positive individuals reported that they were only using condoms to prevent HIV infection/reinfection.
An analysis of only those who said that they did not want more children (and were at risk of becoming pregnant) showed that 81% were using some method of family planning, but only 22% were using a highly effective method. Thus, 78% of all participants had an unmet need for using a highly effective method of contraception. The unmet need for using a highly effective contraceptive method was higher in HIV-positive participants (90%).
Those currently using contraception were asked which methods(s) they were using (see Fig. 1).
The most common method used overall was the male condom (70%), followed by injectable (13%) and oral contraceptives (7%). There were only eight individuals (5% of current users) who reported using dual protection in the form of male condoms combined with either injectables, oral contraceptives or tubal ligation. Seven of these individuals were HIV-positive.
Current choice of methods used differed for HIV-positive and HIV-negative individuals. As Fig. 1 shows, HIV-positive respondents tended to favor the male condom, whereas HIV-negative respondents most commonly used injectable. Current users of contraception, both HIV-positive and HIV-negative, were asked why they chose to use their method of choice over others. The most commonly cited reason was to avoid HIV infection and/or HIV reinfection (35%), followed by convenience and ease of use (18%). Those not currently using contraception were asked their reasons why they were not using contraception. The most commonly cited reasons they gave were being pregnant or breastfeeding (n = 117) or that they wanted more children (n = 47). Additional reasons were fear of adverse effects of hormonal contraceptives (n = 31), stigma/misconceptions (n = 16) and partner opposition (n = 13).
We conducted a cross-sectional study on fertility desires in HIV-positive and HIV-negative individuals in the reproductive age group between 18 and 44 years in western Uganda. We included participants who had been more recently tested for HIV infection. The results of our study contribute to the body of new knowledge about HIV infection and fertility intentions in Ugandan and in sub-Saharan African populations. Our findings support the notion in the literature that HIV-infected persons/couples are less likely to want more children in future compared with non-HIV-infected persons. One strength of our study is that we included both HIV-positive and HIV-negative men and women, and thus we are able to make comparisons between respondents of HIV-positive and negative serostatus. The majority of studies reviewed included only HIV-positive participants.
The most important finding of our study was that the HIV status of participants was highly predictive of the desire for future childbearing with many HIV-positive participants saying that they want to stop childbearing (OR 6.5, P < 0.001) compared with HIV-negative participants. This association was confirmed in bivariate and multivariate analysis as well as in two sub-analyses, one with male and female participants and one with pregnant and nonpregnant participants separately (see Table 3). As the ORs from these sub-analyses are in the same direction as the OR in the main dataset, we believe that the association between a positive HIV serostatus and the desire to stop childbearing, as reported in this paper, is quite robust. Additionally, participants have ‘confirmed’ the validity of their responses regarding HIV status and fertility intentions through their actual behavior, as a very high number of HIV-negative participants (or their partners) reported to be pregnant during the survey, whereas a very small number of HIV-positive respondents (or their partners) reported to be pregnant during the interview, thus validating the interview data.
Our findings support those from other studies, which reported a similar association between a positive HIV status and lower fertility desires [13,14,23], though some studies have found the opposite [2,8,14]. One study from Uganda of HIV-infected women by Homsy et al.  revealed that 7% of study participants wanted more children and 16.9% became pregnant during the follow-up period of the study. As this was a longitudinal study, the authors were able to verify responses at baseline with actual behavior over a 2-year period. This observed difference reported vs. actual may indicate that either participants have changed their mind after the baseline interviews, underreported their true desires for more children or experienced an unintended pregnancy. The participants in the study by Homsy et al.  were clients of a home-based care research program in eastern Uganda. As all were on cART, one could have expected a higher proportion of participants saying that they wanted more children, as cART reduces dramatically the risk of vertical transmission of HIV. As this did not happen, one could guess that participants in the study by Homsy et al. may not have been fully aware of the huge cART benefits on vertical transmission of HIV to prevent HIV infection to their unborn/newborn children.
A second study from eastern Uganda by Nakayiwa et al.  showed that 16% of HIV-infected men and women desired more children in the future. This finding is in line with our result, in which 14% of HIV-infected respondents wanted to have more children in future. Interestingly, as Nakayiwa et al. interviewed clients of ‘The AIDS Support Organization’ (TASO) who resided in the urban and semi-urban areas in eastern Uganda, we would have expected some differences as educated persons usually have a better understanding of how one's health can be improved. The difference between the study by Nakayiwa et al. and our study was that we interviewed only rural dwellers with relatively low education (no education 18%, secondary education 14%), whereas TASO clients had a much higher educational level (no education 11%, secondary education 38%). Predictive factors of HIV-positive persons for wanting more children in the future were similar in a study by Nakayiwa et al. and in our study. The studies by Homsy et al. and Nakayiwa et al. did not include HIV-negative persons, and therefore, the association between HIV serostatus and the desire for more children could not be established in their studies.
A study in Malawi found that the desire for more children in HIV-positive women was 15.1% , whereas a study of HIV-positive women receiving cART in South Africa found the desire to have more children to be 29%. This is much higher than that observed in our study or the study by Homsy et al. . A second study has emerged from rural Malawi, which examined fertility intentions after HIV testing using longitudinal data from an ongoing data collection project . In this study, persons who received a positive HIV test result had less than half the odds (OR 0.40) of wanting more children compared with those who received a negative HIV test result. Other predictors for the desire to have more children in those who received an HIV-positive diagnosis were male sex, younger age, lower educational level and fewer living children, which is somewhat similar to our findings. This is also consistent with results of the study by Yeatman , except that in a study by Yeatman, HIV-positive respondents with a higher educational level were less likely to want more children, a finding we could not confirm.
Our finding of a strong association between a positive HIV status and the desire to stop childbearing is somewhat surprising, as anecdotal information from the study area has indicated that many HIV-positive couples have children and that the desire for more children is very strong. These results are even more remarkable, given that the study districts have traditionally been areas of higher fertility compared with the national average . Cultural factors in the study area also strongly dictate the importance of childbearing, which makes it difficult for couples to remain childless or with fewer children. Perhaps, many HIV-positive persons and couples have recognized the high risk of vertical transmission of HIV to their unborn or infant child, still reflecting the time when the prevention of mother-to-child transmission of HIV with antiretrovirals was not readily available and not well known to the population.
In our study, we found a high unmet need for contraception in both HIV-positive and HIV-negative participants who did not want more children, but who did not use highly effective family planning methods. Dual protection against both prevention of HIV/sexually transmitted infection (STI) infection and unwanted pregnancies was very low (5%). This is consistent with the literature as other studies on dual protection concluded that knowledge and use of dual protection against both HIV/STIs and pregnancy is generally very low [25–28]. Homsy et al.  reported from eastern Uganda that only 5% of contraceptive users applied dual protection, which is similar to what we found in our study. The importance of dual protection is considered to be crucial for achieving good reproductive health [28,29].
Our study had some limitations. First, the ORs were based on cross-sectional data, which precludes assessing the causality of the associations described. Second, social desirability bias cannot be excluded as the information collected was sensitive. However, we used highly trained and experienced interviewers to minimize this bias. Third, we may have experienced a selection bias, as we were not able to perform a true random sampling procedure. This led to imbalances in the two groups and may limit the generalizability of our study findings. The most important difference between the groups was the high pregnancy rate in HIV-negative respondents (66%) compared with that in HIV-positive respondents (10%). The sub-analysis of pregnant and nonpregnant groups in separate logistic regression models revealed that the positive association between a HIV-positive serostatus and the desire to stop childbearing was also confirmed in these two sub-groups, which validates our main study finding that HIV-positive respondents are more likely to want to stop childbearing. This is the same result as in the main multivariate model (see Table 2), in which pregnancy status was not a predictor of fertility intentions. The fourth study limitation was that as our sample had a high proportion of pregnant HIV-negative participants, there remained only a small number of HIV-negative female participants at risk of pregnancy, which reduced our ability to examine differences in current contraceptive practices based on HIV serostatus.
In order to greatly assist those HIV-positive couples who wish to better plan their childbearing, we have three recommendations for district health teams in western Uganda. The first is to extend the delivery of both family planning services and HIV/STI prevention and care, including the prevention of vertical transmission of HIV to all couples in all areas of the districts. The second is to explore options for accelerated integration of HIV/AIDS/STIs and family planning programs and to ensure that counseling about both the prevention of STI/HIV and unwanted pregnancy is the norm. Emphasis should also be put on dual protection of preventing both HIV/STI infections and unwanted pregnancies. The third is to vigorously inform the sexually active population of the very beneficial effects of ART on mother-to-child transmission of HIV, so that HIV-affected couples can make a better choice on a greater range of available options if they want to have more children. This ‘harm reduction’ counseling service should be provided as a supportive component within fertility counseling in all HIV treatment, care and prevention programs.
We thank Jean Kipp for her very useful comments on the first drafts of the manuscript. The study was financed by a research grant from the Canadian Institutes of Health Research (CIHR), grant no. MOP-74586, and the Fund for Support of International Development Activities, University of Alberta.
Conflicts of interest: None.
1. Allen S, Serufilira A, Gruber V, Kegeles S, Van de Perre P, Carael M, Coates TJ. Pregnancy and contraception use among urban Rwandan women after HIV testing and counseling. Am J Public Health 1993; 83:705–710.
2. Baylies C. The impact of HIV on family size preference in Zambia. Reprod Health Matters 2000; 8:77–86.
3. Moyo W, Mbizvo MT. Desire for a future pregnancy among women in Zimbabwe in relation to their self-perceived risk of HIV infection, child mortality, and spontaneous abortion. AIDS Behav 2004; 8:9–15.
4. Delvaux T, Nostlinger C. Reproductive choice for women and men living with HIV: contraception, abortion and fertility. Reprod Health Matters 2007; 15:46–66.
5. Newell ML, Brahmbhatt H, Ghys PD. Child mortality and HIV infection in Africa: a review. AIDS 2004; 18:S27–S34.
6. French R, Brocklehurst P. The effect of pregnancy on survival in women infected with HIV: a systematic review of the literature and meta-analysis. Br J Obstet Gynaecol 1998; 105:827–835.
7. Lewis JJ, Ronsmans C, Ezeh A, Gregson S. The population impact of HIV on fertility in sub-Saharan Africa. AIDS 2004; 18:S35–S43.
8. Gregson S, Zhuwau T, Anderson RM, Chandiwana SK. Is there evidence for behaviour change in response to AIDS in rural Zimbabwe? Soc Sci Med 1998; 46:321–330.
9. Feldman R, Maposhere C. Safer sex and reproductive choice: findings from ‘positive women – voices and choices’ in Zimbabwe. Reprod Health Matters 2003; 11:162–173.
10. Lutalo T, Kidugavu M, Wawer MJ, Serwadda D, Zabin LS, Gray RH. Trends and determinants of contraceptive use in Rakai district, Uganda, 1995–98. Stud Fam Plann 2000; 31:217–227.
11. Fabiani M, Nattabi B, Ayella EO, Ogwang M, Declich S. Differences in fertility by HIV serostatus and adjusted HIV prevalence data from an antenatal clinic in northern Uganda. Trop Med Int Health 2006; 11:182–187.
12. Ross A, Van der Paal L, Lubega R, Mayanja BN, Shafer LA, Whitworth J. HIV-1 disease progression and fertility: the incidence of recognized pregnancy and pregnancy outcome in Uganda. AIDS 2004; 18:799–804.
13. Grieser M, Gittelsohn J, Shankar AV, Koppenhaver T, Legrand TK, Marindo R, et al. Reproductive decision making and the HIV/AIDS epidemic in Zimbabwe. J South Afr Stud 2001; 27:225–243.
14. Rutenberg N, Baek C. Review of field experiences: integration of family planning and PMTCT services. Population Council 2004. Washington, DC.
15. Wakabi W. Population growth continues to drive up poverty in Uganda. Lancet 2006; 367:558.
16. Nakayiwa S, Abang B, Packel L, Lifshay J, Purcell DW, King R, et al. Desire for children and pregnancy risk behavior among HIV-infected men and women in Uganda. AIDS Behav 2006; 10:S95–S104.
17. Homsy J, Bunnel R, Moore D, King R, Malamba S, Nakityo R, et al. Reproductive intentions and outcomes among women on antiretroviral therapy in rural Uganda: a prospective cohort study. PLoS One 2009; 4:e4149. Epub.
18. Hoffmann IF, Martinson FE, Powers KA, Chilongozi DA, Msiska ED, Kachipapa EI, et al. The year-long effect of HIV-positive test results on pregnancy intentions, contraceptive use, and pregnancy incidence among Malawian women. J Acquir Immune Defic Syndr 2008; 47:477–483.
19. Myer L, Morroni C, Rebe K. Prevalence and determinants of fertility intentions on HIV-infected women and men receiving antiretroviral therapy. AIDS Patient Care STDs 2007; 21:278–285.
20. Peltzer K, Chao LW, Dana P. Family Planning Among HIV Positive and Negative Prevention of Mother to Child Transmission (PMTCT) Clients in a Resource Poor Setting in South Africa. AIDS Behav 2008 [Epub ahead of print].
21. Uganda Bureau of Statistics, Macro International Inc. Uganda Demographic and Health Survey 2007; final report; Kampala, Uganda.
22. STATA Statistical Software: release 9. College Station, Texas, USA: STATA Corporation; 2005.
23. Baek C, Rutenberg N. Addressing the family planning needs of HIV-positive PMTCT clients: baseline findings from an operations research study. Population Council 2005; Washington, DC.
24. Yeatman S. The impact of HIV status and perceived status on fertility desires in rural Malawi. AIDS Behav 2009; 13:S12–S19.
25. Bull SC, Shlay JC. Promoting ‘dual protection’ from pregnancy and sexually transmitted diseases: a social ecological approach. Health Promot Pract 2005; 6:72–80.
26. Kleinschmidt I, Maggwa BN, Smit J, Beksinska ME, Rees H. Dual protection in sexually active women. S Afr Med J 2003; 93:854–857.
27. Maharaj P, Cleland J. Condom use within marital and cohabiting partnerships in KwaZulu-Natal. Stud Fam Plann 2004; 35:116–124.
28. Brady M. Preventing sexually transmitted infections and unintended pregnancy, and safeguarding fertility: triple protection needs of young women. Reprod Health Matters 2003; 11:134–141.
29. Berer M. HIV/AIDS, sexual and reproductive health: intimately related. Reprod Health Matters 2003; 11:6–11.
© 2009 Lippincott Williams & Wilkins, Inc.