Share this article on:

Enrollment and Retention of HIV Discordant Couples in Lusaka, Zambia

Kempf, Mirjam-Colette MPH, PhD*†; Allen, Susan MD, MPH, DTM&H*‡; Zulu, Isaac MD, MPH*§; Kancheya, Nzali MD, MPH*§; Stephenson, Rob PhD‡; Brill, Ilene MPH*†; Tichacek, Amanda MPH*‡; Haworth, Alan MD*§∥¶; Chomba, Elwyn MD*¶; the Rwanda Zambia HIV Research Group

JAIDS Journal of Acquired Immune Deficiency Syndromes: 1 January 2008 - Volume 47 - Issue 1 - pp 116-125
doi: 10.1097/QAI.0b013e31815d2f3f
Epidemiology and Social Science

Background: Biased enrollment and attrition compromise the power of clinical trials and limit generalizability of findings. We identify predictors of enrollment and retention for HIV-discordant couples enrolled in prospective studies in Zambia.

Principal Findings: A total of 1995 discordant couples were invited to enroll. Predictors of nonenrollment, loss to follow-up, and missed appointments were evaluated using multivariate models. M+F couples were more likely to be eligible and to enroll and less likely to be lost to follow-up than F+M couples. Substantial losses to follow-up occurred between testing and enrollment (21.3% of M+F and 28.1% of F+M) and between enrollment and the first follow-up visit (24.9% of M+F and 30.5% of F+M). Among M+F and F+M couples, residence far from the clinic, younger age, and women's age at first intercourse ≤17 years were predictive of attrition. No income, ≤2 lifetime sex partners, no history of sexually transmitted infection in women, and recent extramarital contact in their male partners predicted attrition in F+M couples.

Conclusions: Discordant couples are critical to observational studies and clinical trials to prevent male-to-female and female-to-male transmission. Retention biases must be taken into account during analysis. Run-in designs that delay randomization may improve retention in clinical trials.

From the *Zambia-Emory HIV Research Project, Lusaka, Zambia; †Department of Epidemiology, Ryals School of Public Health, University of Alabama at Birmingham, Birmingham, AL; ‡Hubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, GA; §School of Medicine, University of Zambia, Lusaka, Zambia; ∥University Teaching Hospital, Lusaka, Zambia; and ¶Counseling Services Unit, Ministry of Health, Lusaka, Zambia.

Received for publication April 6, 2007; accepted October 4, 2007.

Supported by funding from the Fogarty AIDS International Training and Research Program FIC 2D43 TW001042; the Social and Behavioral Core of the Emory Center for AIDS Research P30 AI050409; National Institutes of Health grants AI23980, AI 40951, HD 40125, MH 66767; and the International AIDS Vaccine Initiative.

This work has previously been presented at the following meetings: Second International AIDS Society Conference on HIV Pathogenesis and Treatment, Paris, France, July 13-17, 2003 (oral presentation and abstract 159), and AIDS Vaccine 2003 Conference, New York, NY, September 18-21, 2003 (oral presentation).

Correspondence to: Susan Allen, MD, MPH, DTM&H, Hubert Department of Global Health, Rollins School of Public Health, Emory University, 1520 Clifton Road NE, Suite 234, Atlanta, GA 30322 (e-mail:,

The HIV epidemic in East and Southern Africa has evolved to a generalized epidemic characterized by heterosexual transmission accompanied by mother-to-child transmission. Voluntary counseling and testing (VCT) for HIV has been shown to be an efficacious and cost-effective prevention strategy in some risk groups.1-5 Most new HIV infections occur in cohabiting couples,6-11 (Allen et al, submitted for publication) and the beneficial effect of VCT is further enhanced by counseling and testing couples together.12-20

A variety of socioeconomic and biologic factors play a role in the magnitude of the epidemic in sub-Saharan Africa: high rates of sexually transmitted infections (STIs) other than HIV, low levels of education, high rates of poverty, gender inequity, and high population mobility.8,12,21-23 Although transmission in couples can be decreased with appropriate counseling, discordant couples remain at relatively high risk, with transmission associated with low literacy rates, extramarital sex partners, high viral load, and genital ulcers.24-29 Couples' voluntary counseling and testing (CVCT) should be linked to prevention research in discordant couples, including observational studies and clinical trials.30

The aim of this study is to provide data to inform future HIV research efforts. In observational studies and clinical trials, enrollment and follow-up of study participants into research studies present a significant challenge. It is critical to identify factors that contribute to nonenrollment and loss to follow-up, particularly among discordant couples, who are uniquely suited to studies of heterosexual transmission. CVCT has been conducted by our research group in Lusaka, Zambia since 1994 as a recruitment mechanism for studies of HIV transmission and the natural history of HIV disease. We present here the sequence of events from screening through enrollment and follow-up of the largest single-site prospective study of discordant couples published to date. Cross-sectional analyses are used to compare couples who enroll and return for follow-up with those who do not. The implications of bias in recruitment and retention are discussed, and strategies to improve study participation are presented.

Back to Top | Article Outline


All procedures were approved by US and Zambian ethical committees registered with the Office for Human Research Protections of the US Public Health Service. Joint written informed consent was obtained jointly from both partners at the time of CVCT and again at the time of enrollment into the prospective study.

Back to Top | Article Outline

Couples' Voluntary Counseling and Testing

The CVCT model used was developed and refined in a similar population in Rwanda, where it showed a beneficial impact on condom use, HIV/STI rates, and pregnancy.15,31,32 These findings were based on observational studies and extrapolation,24 because ethical mandates preclude the inclusion of a control group of discordant couples who are unaware of their HIV test results. A freestanding couples voluntary testing and counseling center was established in the northwest of Lusaka in 1994. The center provided same-day HIV testing and counseling services for cohabiting couples only.33 Community outreach workers and radio announcements were used to promote the services.34

The same-day CVCT session started with a group videotape and discussion session, followed by pretest counseling with a trained counselor. Couples who gave written informed consent for joint testing had their blood drawn and were given a code number that was linked to their test results. Nyanja, the language used in Lusaka,35 was employed for educational messages and data collection instruments. Rapid HIV tests and rapid plasma reagin (RPR) serology tests were performed on-site.33 Posttest counseling was provided to both partners jointly. Those with positive RPR serologic results were given free treatment on site. Most couples had concordant HIV test results: 51% had 2 HIV-negative partners and 26% had 2 HIV-positive partners.34 After posttest counseling, concordant couples were invited to return anytime for repeat tests or follow-up counseling. Couples with 1 HIV-seronegative partner and 1 HIV-seropositive partner (M+F [11% of couples tested] and F+M [12% of couples tested]) were invited to join a research study investigating predictors of HIV transmission and the natural history of HIV disease. All couples, whether they elected to be tested or not, were reimbursed for transportation costs, and on-site child care was provided. Further details of our CVCT procedures have been published previously.33

Back to Top | Article Outline

Enrollment and Follow-Up

Discordant couples who returned for enrollment completed social and medical questionnaires, followed by blood draws. Reimbursement of transportation and opportunity costs ($6 per person per visit) and provision of childcare were provided at enrollment and follow-up visits. Participants received medical benefits equivalent to standard care at the research clinic and government health insurance membership. The research clinic pharmacy was stocked with medications included on the World Health Organization (WHO) essential drugs list, including broad-spectrum antibiotics, antimalarials and anthelminthics, and anti-inflammatories/analgesics. Project physicians included internists and gynecologists from the University Teaching Hospital. The government health insurance allowed access to government clinics and hospitals to ensure off-hours outpatient care and access to specialty and inpatient care.

Recruitment and follow-up into the open cohort began in 1994, with CVCT offered 6 days a week through January 1997 and 3 days a week thereafter until November 1998. Enrollment and follow-up procedures remained constant over time. Couples were scheduled to come together when possible for quarterly follow-up visits; 85% (7650 of 9008) of visits included both partners seen on the same day; 10% (942) included partners seen on different days; and in 5% (416), 1 partner came for follow-up and the other missed that quarterly appointment. The mean duration of follow-up was 15 months (median = 12 months, range: 3 to 42 months). Study benefits continued for as long as the cohabiting sexual partners with discordant HIV results adhered with quarterly appointments. Seroconverters and their spouses remained in the study, although postseroconversion visits are censored here. In the event of a death, the surviving partner was provided with 6 months of care at the clinic and 1 year of government insurance.

Back to Top | Article Outline

Follow-Up and Retention Strategies

Study membership cards with the project logo, photographs, and appointment dates were provided.36 Couples who missed their appointment date received a reminder invitation from a community worker. Reminders were limited to 2 per missed visit for 2 consecutive visits. Those who did not come in, whether they verbally stated that they did not intend to enroll or simply did not come, were defined as “declined” (Fig. 1). Because of the high mobility of the urban Zambian population, the difficulty in locating residences in high-density residential areas (many of which are unplanned or squatter communities), and the possibility that some couples may have knowingly given false or incomplete locator information, couples who could not be found using available locator information were defined as “unable to contact” or “no contact.” Couples who separated were released immediately from the study. In the event that 1 or both members of a couple died, the spouse or caretaker was invited to the project to complete end of follow-up protocols.

Back to Top | Article Outline

Data Collection

During CVCT, a short questionnaire was completed with the couple, collecting basic demographic data, including location of residence, age, duration of the couple's union, number of deliveries, use of contraception, and pregnancy.

At enrollment, data were collected separately for men and women to ensure confidentiality and accurate reporting. Information concerning children, marriage, income, education, language, alcohol consumption, and sexual history was recorded. Detailed locator information, including directions from the nearest bus stop and names of neighbors, was recorded,36,37 because most participants did not have telephones or mailing addresses.

Back to Top | Article Outline

Data Analysis

Data were entered on-site and analyzed using the SAS software package (version 8.2; SAS Institute, Cary, NC). Data were evaluated using frequency distributions or univariate descriptive statistics as appropriate. Categoric and continuous variables were dichotomized, and cross-tabulations were generated comparing couples who enrolled with those who did not; couples who enrolled and had follow-up with those who enrolled but did not have follow-up; couples with missing appointments for the HIV negative partner in the first 12 months with those couples with regular attendance for the HIV negative partner during the same time period; and, finally, couples lost to follow-up with those remaining in the study. Cochran-Mantel Haenszel statistics and Fisher exact tests were performed to test for associations between each outcome and the dichotomous predictor variables. Multivariate logistic regression modeling was used to adjust all risk estimates for possible covariates in cross-sectional comparisons. Variables showing significance levels of P ≤ 0.05 were considered for entry in multivariate logistic regression models. Ages of men and women were included in all models. Risk estimates are given as odds ratios (ORs) with 95% confidence intervals (CIs).

Back to Top | Article Outline


Eligibility Screening, Enrollment, and Retention

Of 1995 discordant couples identified at CVCT, 928 (47%) were M+F and 1067 (53%) were F+M (see Fig. 1). At baseline, 7.5% of M+F and 14.4% of F+M were ineligible for enrollment into the study. The most common reason for ineligibility was not having cohabited for at least 12 months. Of the 1771 discordant couples eligible for study enrollment, 858 (48%) were M+F and 913 (52%) were F+M.

Among M+F couples, 183 (21.3%) eligible couples did not enroll, of whom 148 (81%) declined to participate and 9 (5%) separated. More than a quarter of eligible F+M couples (n = 257 [28.1%]) did not enroll in the prospective study. Of those, most (n = 217 [84%]) declined to participate and 6 couples (2%) separated.

Among couples who completed the baseline enrollment visit, 168 (24.9%) M+F and 200 (30.5%) F+M did not return for follow-up. Of the M+F couples who did not return for a follow-up visit, 88 (52%) declined to continue and 18 (11%) separated. In comparison, of the F+M couples who did not return after the enrollment visit, 91 (46%) declined to continue and 34 (17%) separated. M+F couples were more likely to be eligible, more likely to enroll if eligible, and more likely to return for follow-up than F+M couples. Overall, 507 (55%) of 928 M+F couples and 456 (43%) of 1067 F+M couples were enrolled with follow-up (see Fig. 1).

Back to Top | Article Outline

Demographic Characteristics of Eligible Discordant Couples

M+F couples were slightly older than F+M couples (mean = 26.6 years for women vs. 25.8 years; P = 0.014 and mean = 33.9 years for men vs. 32.6 years; P = 0.002), with longer durations of union (7.1 vs. 5.1 years; P = 0.0001) and more deliveries (2.8 vs. 2.1; P = 0.0001). At the time of CVCT, 15% of women in M+F couples and 13% of women in F+M couples were pregnant (P = not significant [NS]). Few couples reported current use of any sort of contraception (11%); 8% reported oral contraception and 1% injectable hormonal contraception, 1% reported that they were sterile, and 2 couples reported the use of an intrauterine device (IUD) (Table 1). Not surprisingly, given that eligibility data were recorded on the day the couples were jointly tested and counseled, only 2% reported using condoms. Contraceptive use was more frequent among M+F couples than among F+M couples (14% vs. 8%; P = 0.001). Most couples (62%) lived in the northwest section of Lusaka (near the research site), with 34% living in other parts of town and only 4% residing outside the city limits.

Back to Top | Article Outline

Predictors of Enrollment

Younger men and women in M+F and F+M couples were less likely to enroll (see Table 1; first set of columns in Tables 2A, B). Duration of union ≤3 years and ≤2 deliveries were also associated with not enrolling, but both correlated with age and were not independently associated with enrollment. Current pregnancy was not significantly associated with enrollment. Among M+F couples, the few who used contraception were significantly more likely to enroll, and this was the only variable other than age that remained significant in the multivariate analysis (see Table 2A). Residence far from the research clinic was independently predictive of nonenrollment among F+M couples, and only this and men's age ≤30 years remained predictive in the multivariate model (see Table 2B).

Back to Top | Article Outline

Demographic Characteristics and Risk Factors Among Enrolled Couples

A total of 1331 eligible discordant couples enrolled in the prospective study after receiving CVCT services. Age, duration of union, number of deliveries, and contraceptive use were similar to those reported previously for eligible couples as a whole. The average monthly income for the male partner was 54,265 Kwacha (US $40.6) and 10,934 (US $8.2) Kwacha for the female partner (the per capita poverty line in Zambia during this time was $30.24 per month38). Self-reported literacy in English was more common in men compared with women and more common in HIV-positive men and women compared with their HIV-negative counterparts of the same gender (Table 3; P < 0.05 for each comparison). Most couples were recruited from the area around the project site and reported living an average of 3.1 years at their current address (not shown). F+M couples were comparatively mobile, having lived at their current location for an average of 2.6 years versus 3.5 years for M+F couples (P < 0.0001).

Men reported high alcohol intake more often than women (Table 4). Women's median age at first intercourse was reported to be 16 years versus 17 years for the men, but no significant difference was observed between HIV-positive and HIV-negative women. The average number of lifetime sex partners was 3.0 for women and 9.9 for men. HIV-negative women reported an average of 2.7 lifetime sex partners versus 3.5 lifetime sex partners for HIV-positive women (P < 0.0001). Few (3%) of the female partners reported having outside sex partners during the past 3 months, compared with 14% of the male partners (see Table 4). A history of STIs during the past 5 years was more often reported by HIV-positive men and HIV-positive women (P < 0.0001 in each case). Two thirds of HIV-positive partners were classified as stage 1 or 2 using the Kigali staging system (see Table 4).39

Back to Top | Article Outline

Predictors of No Follow-Up After Enrollment

None of the dichotomized demographic variables or risk factors presented remained significant predictors of having no follow-up after the enrollment visit for M+F couples in univariate analysis (see Table 3; see second set of columns in Table 2A). For F+M couples, however, women's age ≤25 years, men's age ≤30 years, low or no alcohol intake for the woman, and the male partner reporting an outside sex partner during the past 3 months were predictive of not having follow-up in univariate analyses, and the latter 3 variables remained independently predictive in the multivariate analysis (see Table 3; see second set of columns in Table 2B).

Back to Top | Article Outline

Predictors of Missing Visits

Among M+F couples, men reporting ≤6 lifetime sex partners was the only significant predictor of missing visits (see third set of columns in Table 2A). Three variables were predictive of missing visits among F+M couples in univariate and multivariate analyses. All were characteristics of the woman: poor English literacy, high alcohol intake, and no history of STIs in the past 5 years (see third set of columns in Table 2B).

Back to Top | Article Outline

Predictors of Loss to Follow-Up

Among couples with follow-up after the enrollment visit, women's age <17 years at first intercourse was predictive of subsequent loss to follow-up in M+F and F+M couples (see right-hand columns in Tables 2A, B). In M+F couples, women's age ≤25 years and not residing close to the research site were independently predictive of loss to follow-up in multivariate analysis. No income for women and comparatively good health for men (Kigali HIV disease stage 1 or 2) were significant in univariate analysis but did not persist in multivariate analysis (see Table 2A). Among F+M couples, the female partner reporting no income, ≤2 lifetime partners, and not having an STI in the past 5 years were significant predictors of loss to follow-up and remained so in multivariate analysis (see Table 2B).

Back to Top | Article Outline


Heterosexual couples are the largest HIV risk group in Africa.9 In Zambia, although couples benefit from CVCT,2 incidence rates in discordant couples remain high relative to other risk groups and most new infections are acquired from spouses.40 This presents an ideal opportunity to study the differences between female-to-male and male-to-female transmission.26 Discordant couples are also uniquely suited to pathogenesis studies of contagion in the HIV-positive partner and susceptibility in the HIV-negative spouse and allow clinical trials of interventions that target 1 or both partners. In this study of the largest single-site discordant couples' study published to date, approximately equal numbers of M+F and F+M couples were identified in our CVCT centers, similar to the composition of a community-based discordant couples' cohort in Uganda25 and to results found by serosurveys and testing centers in Rwanda, Tanzania, and Kenya.15,41,42

Less than half of discordant couples identified in study-sponsored CVCT sites enrolled and returned for follow-up. Among couples with HIV-positive men, there were few differences between the 55% who enrolled with follow-up and the 45% who were ineligible, did not enroll, or had no follow-up beyond enrollment. In contrast, among couples with HIV-positive women, there was biased retention with differences in demographic and risk factor profiles between the 43% with follow-up and the 57% without.

A main barrier to HIV prevention and treatment in the African setting is traditional gender roles,21,43,44 which may also influence a woman's decision to participate in research. It is plausible that the couples enrolled here represent a subset of couples in which women have greater decision-making autonomy. Enrolled couples were older and had longer durations of union and more children than nonenrolled couples; these may also be characteristics that are associated with greater female autonomy (if autonomy develops over time as a woman's position in the family is secured through childbearing).

Among HIV-positive women, having no income and having a partner who reported recent contact with other partners was associated with loss to follow-up, and being illiterate in English was predictive of irregular attendance. The lower socioeconomic and educational status of these women,45 and the threat posed by the man's other partners,43 may have reduced their ability to negotiate participation in a research study. It is hard to establish whether the CVCT activities attract couples with greater gender equity; however, to address this, outreach activities should focus on the young, illiterate, and unemployed who were underrepresented here.

On average, women in our study had initiated sex at the age of 16 years and reported 2 premarital partners in the 4 to 5 years before entering their current union. Early sexual debut of African women is associated with early infection with HIV.46-49 Women's age <17 years at first intercourse was independently predictive of loss to follow-up in HIV-positive and HIV-negative women. Other measures related to retention and adherence included high alcohol intake in HIV-positive women, which was rare (6%) and associated with a paradoxically increased likelihood of enrollment but also with a greater likelihood of missing appointments. High alcohol intake was much more common among men50 but was not associated with enrollment or follow-up. Early sexual debut and alcohol consumption, both nonnormative behaviors for women in the cohort, are clearly associated with enrollment and retention, although in differing ways. Although these may be reflections of the socioeconomic status of the women,51-53 they may also reflect variations in familial background or psychological functioning. Further research is warranted to elucidate clearly the reasons behind these relations and to ensure that services are tailored toward the unique needs of these women.

HIV-positive women who reported a history of STI were significantly more likely to remain in follow-up, in all likelihood because they valued the STI screening and treatment provided to study participants. Free health care is much appreciated as a benefit of research participation,54 and it ensures good documentation of manifestations of HIV disease and potential cofactors for HIV transmission. The increased enrollment among modern contraceptive users suggests that family planning services were also appreciated.31,32,55 Discordant couples are advised to use condoms to prevent HIV transmission and an additional user-independent method to ensure added protection against unplanned pregnancy.56,57 Younger couples were less likely to enroll and remain in the study, possibly because they had not achieved their desired family size and were reluctant to participate in a study that placed a heavy emphasis on condom use. Fertility intentions should be discussed during counseling, and couples should be reassured that researchers are not negatively judging study participants.

Enrolled couples were reimbursed for transport, and childcare services were provided at the research site. Although this helped to reduce logistic obstacles,16,36,37,58 residing far from the clinic remained a significant predictor of attrition. Participants could walk to the clinic if they lived nearby, but bus fares were expensive. Couples needed enough money to travel, often with small children, to the clinic. Mechanisms that facilitate travel, such as vouchers or satellite clinics, should be considered in settings where distance inhibits study participation. Few participants had telephones or mailing addresses, and reminders were delivered to the home. With the increasing availability of low-cost mobile phones, telephone cards may be a useful addition to retention efforts in similar areas.36,37

The research activities described here have demonstrated the feasibility of enrolling and retaining large cohorts of serodiscordant couples; however, there were a number of issues unique to couples that must be considered. First, couples' testing was not widely available; thus, there were limited sites through which to identify discordant couples.34,59 The success of the enrollment and retention activities described here rested on the presence of a couple-focused clinical service (CVCT) from which to recruit. With the availability of more and simpler HIV rapid tests, novel strategies such as home-based HIV testing may be explored as a mechanism to expand access HIV testing.60

To remain active in a discordant couples' cohort, both partners had to be cohabiting in a sexual union. Consent procedures required individual consent from both partners, and enrollment was not limited by a need for the male partner to consent on behalf of the couple61 (although a woman would not be able to enroll if her partner refused his own participation, a man's participation would be similarly limited by his partner's refusal to consent to enrollment). Retention was also a challenge; studies of high-risk groups, particularly women, in HIV vaccine preparedness and other clinical trial programs have yielded retention rates as low as 44% to 66%.62-65 Focusing on couples adds new challenges, because the loss of either partner as a result of death or seroconversion or dissolution of the union results in loss of the couple. Separation of couples was uncommon, reported in only 4% of eligible discordant couples, a similar rate to that found in a clinical trial of VCT that took place in Kenya, Tanzania, and Trinidad and Tobago.42 In our experience, as in other reports,37 most couples were lost between screening (CVCT) and enrollment and immediately after the baseline visit. In such circumstances, clinical trials may benefit from a “run-in” design that begins with several visits before randomization. Additionally, couples-focused clinical trials may have additional exclusion criteria, such as pregnancy or highly active antiretroviral therapy (HAART) use, that further reduce the screen-to-enroll ratio. Any couples-focused research must recognize the competing social, economic, and reproductive forces that may influence continued participation. Interventions that allow couples to discuss and manage these factors are likely to encourage greater rates of enrollment and retention.

Back to Top | Article Outline


The authors thank the study participants, staff, interns, and Project Management Group members of the Zambia-Emory HIV Research Project in Lusaka, Zambia. They also thank Martha Conkling, PhD, for her editorial review and comments.

Back to Top | Article Outline


Dr. Mirjam-Colette Kempf participated in the preparation of the manuscript and coordinated the analysis, Robert Stephenson coordinated the revision in response to reviewer comments, and Ilene Brill served as the statistician for the paper. Susan Allen is the Principal Investigator for the Rwanda Zambia HIV Research Group and was responsible for securing funding, supervising data collection, and finalizing the revised manuscript. Isaac Zulu and Nzali Kancheya oversaw all recruitment and retention efforts in the field and were an active part of the paper revision process while completing their Master of Public Health degrees in the United States. Elwyn Chomba is the Senior Co-Investigator of the Zambia Emory HIV Research Project. Amanda Tichacek participated in all aspects of the study from study design to creation of instruments and data analysis and reviewed the manuscript. Alan Haworth assisted with manuscript preparation and revision.

Back to Top | Article Outline


1. Sweat M, Gregorich S, Sangiwa G, et al. Cost-effectiveness of voluntary HIV-1 counselling and testing in reducing sexual transmission of HIV-1 in Kenya and Tanzania. Lancet. 2000;356:113-121.
2. Allen S, Karita E, Ng'andu N, et al. The evolution of voluntary testing and counseling as an HIV prevention strategy. In: Gibney L, DiClemente RJ, Vermund SH, eds. Preventing HIV in Developing Countries: Biomedical and Behavioral Approaches. New York: Plenum Press; 1999:87-108.
3. Colebunders R, Ndumbe P. Priorities for HIV testing in developing countries? Lancet. 1993;342:601-602.
4. Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Efficacy of voluntary HIV-1 counselling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: a randomised trial. Lancet. 2000;356:103-112.
5. Wolff B, Nyanzi B, Katongole G, et al. Evaluation of a home-based voluntary counselling and testing intervention in rural Uganda. Health Policy Plan. 2005;20:109-116.
6. Allen S, Lindan C, Serufilira A, et al. Human immunodeficiency virus infection in urban Rwanda. Demographic and behavioral correlates in a representative sample of childbearing women. JAMA. 1991;266:1657-1663.
7. Hudson CP. The importance of transmission of HIV-1 in discordant couples in the population dynamics of HIV-1 infection in rural Africa. Int J STD AIDS. 1996;7:302-304.
8. Piot P, Bartos M, Ghys PD, et al. The global impact of HIV/AIDS. Nature. 2001;410:968-973.
9. Bagala A. Uganda: married couples top HIV infection rates. The Daily Monitor. December 4, 2006. Available at: Accessed December 6, 2006.
10. Allen S, Karita E, Chomba E, et al. Promotion of couples' voluntary counselling and testing for HIV through influential networks in two African capital cities. Biomed Central. 2007; in press.
11. Malamba SS, Mermin JH, Bunnell R, et al. Couples at risk: HIV-1 concordance and discordance among sexual partners receiving voluntary counseling and testing in Uganda. J Acquir Immune Defic Syndr. 2005;39:576-580.
12. Painter TM. Voluntary counseling and testing for couples: a high-leverage intervention for HIV/AIDS prevention in sub-Saharan Africa. Soc Sci Med. 2001;53:1397-1411.
13. Allen S, Meinzen-Derr J, Kautzman M, et al. Sexual behavior of HIV discordant couples after HIV counseling and testing. AIDS. 2003;17:733-740.
14. Kamenga M, Ryder RW, Jingu M, et al. Evidence of marked sexual behavior change associated with low HIV-1 seroconversion in 149 married couples with discordant HIV-1 serostatus: experience at an HIV counselling center in Zaire. AIDS. 1991;5:61-67.
15. Allen S, Tice J, Van de Perre P, et al. Effect of serotesting with counselling on condom use and seroconversion among HIV discordant couples in Africa. BMJ. 1992;304:1605-1609.
16. Antelman G, Smith Fawzi MC, Kaaya S, et al. Predictors of HIV-1 serostatus disclosure: a prospective study among HIV-infected pregnant women in Dar es Salaam, Tanzania. AIDS. 2001;15:1865-1874.
17. Becker S. Couples and reproductive health: a review of couple studies. Stud Fam Plann. 1996;27:291-306.
18. Guthrie BL, de Bruyn G, Farquhar C. HIV-1-discordant couples in sub-Saharan Africa: explanations and implications for high rates of discordancy. Current HIV Research. 2007;5:416-429.
19. Hugonnet S, Mosha F, Todd J, et al. Incidence of HIV infection in stable sexual partnerships: a retrospective cohort study of 1802 couples in Mwanza Region, Tanzania. J Acquir Immune Defic Syndr. 2002;30:73-80.
20. Padian NS, O'Brien TR, Chang Y, et al. Prevention of heterosexual transmission of human immunodeficiency virus through couple counseling. J Acquir Immune Defic Syndr. 1993;6:1043-1048.
21. Kalipeni E. Health and disease in southern Africa: a comparative and vulnerability perspective. Soc Sci Med. 2000;50:965-983.
22. Serwadda D, Gray RH, Wawer MJ, et al. The social dynamics of HIV transmission as reflected through discordant couples in rural Uganda. AIDS. 1995;9:745-750.
23. Maman S, Mbwambo JK, Hogan NM, et al. HIV-positive women report more lifetime partner violence: findings from a voluntary counseling and testing clinic in Dar es Salaam, Tanzania. Am J Public Health. 2002;92:1331-1337.
24. Allen S, Serufilira A, Bogaerts J, et al. Confidential HIV testing and condom promotion in Africa. Impact on HIV and gonorrhea rates. JAMA. 1992;268:3338-3343.
25. Carpenter LM, Kamali A, Ruberantwari A, et al. Rates of HIV-1 transmission within marriage in rural Uganda in relation to the HIV sero-status of the partners. AIDS. 1999;13:1083-1089.
26. Fideli US, Allen SA, Musonda R, et al. Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus type 1 in Africa. AIDS Res Hum Retroviruses. 2001;17:901-910.
27. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet. 2001;357:1149-1153.
28. Wawer MJ, Gray RH, Sewankambo NK, et al. Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis. 2005;191:1403-1409.
29. Wawer MJ, Sewankambo NK, Serwadda D, et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: a randomised community trial. Rakai Project Study Group. Lancet. 1999;353:525-535.
30. Pequegnat W, Stover E. Behavioral prevention is today's AIDS vaccine! AIDS. 2000;14(Suppl 2):S1-S7.
31. Allen S, Serufilira A, Gruber V, et al. Pregnancy and contraception use among urban Rwandan women after HIV testing and counseling. Am J Public Health. 1993;83:705-710.
32. King R, Estey J, Allen S, et al. A family planning intervention to reduce vertical transmission of HIV in Rwanda. AIDS. 1995;9(Suppl 1):S45-S51.
33. McKenna SL, Muyinda GK, Roth D, et al. Rapid HIV testing and counseling for voluntary testing centers in Africa. AIDS. 1997;11(Suppl 1):S103-S110.
34. Chomba E, Allen S, Kanweka W, et al. Evolution of couples' voluntary counseling and testing for HIV in Lusaka, Zambia. J Acquir Immune Defic Syndr. 2008;47:108-115.
35. United Nations Network in Public Administration and Finance. Available at: Accessed September 10, 2007.
36. Aitken L, Gallagher R, Madronio C. Principles of recruitment and retention in clinical trials. Int J Nurs Pract. 2003;9:338-346.
37. Ribisl KM, Walton MA, Mowbray CT, et al. Minimizing participant attrition in panel studies through the use of effective retention and tracking strategies: review and recommendations. Eval Program Plann. 1996;19:1-25.
38. Independent Evaluation Group-World Bank. Annual review of development effectiveness 2006: getting results: Appendix D, 2006. Available at: Accessed September 14, 2007.
39. Lifson AR, Allen S, Wolf W, et al. Classification of HIV infection and disease in women from Rwanda. Evaluation of the World Health Organization HIV staging system and recommended modifications. Ann Intern Med. 1995;122:262-270.
40. Trask SA, Derdeyn CA, Fideli U, et al. Molecular epidemiology of human immunodeficiency virus type 1 transmission in a heterosexual cohort of discordant couples in Zambia. J Virol. 2002;76:397-405.
41. Bakari JP, McKenna S, Myrick A, et al. Rapid voluntary testing and counseling for HIV. Acceptability and feasibility in Zambian antenatal care clinics. Ann N Y Acad Sci. 2000;918:64-76.
42. Grinstead OA, Gregorich SE, Choi KH, et al. Positive and negative life events after counselling and testing: the Voluntary HIV-1 Counselling and Testing Efficacy Study. AIDS. 2001;15:1045-1052.
43. Mbizvo MT. Gender dynamics and the challenges for HIV prevention. Cent Afr J Med. 1996;42:351-354.
44. Susser I, Stein Z. Culture, sexuality, and women's agency in the prevention of HIV/AIDS in southern Africa. Am J Public Health. 2000;90:1042-1048.
45. Stephenson R, Barker J, Cramer R, et al. The demographic profile of sero-discordant couples enrolled in clinical research in Rwanda and Zambia. AIDS Care. (In press).
46. Kwesigabo G, Killewo JZ, Sandstrom A. Sentinel surveillance and cross sectional survey on HIV infection prevalence: a comparative study. East Afr Med J. 1996;73:298-302.
47. Boerma JT, Urassa M, Senkoro K, et al. Spread of HIV infection in a rural area of Tanzania. AIDS. 1999;13:1233-1240.
48. Fylkesnes K, Ndhlovu Z, Kasumba K, et al. Studying dynamics of the HIV epidemic: population-based data compared with sentinel surveillance in Zambia. AIDS. 1998;12:1227-1234.
49. Fontanet AL, Messele T, Dejene A, et al. Age- and sex-specific HIV-1 prevalence in the urban community setting of Addis Ababa, Ethiopia. AIDS. 1998;12:315-322.
50. Coldiron M, Stephenson R, Chomba E, et al. The relationship between alcohol consumption and unprotected sex among known HIV-discordant couples in Rwanda and Zambia. AIDS Behav. 2007 Aug 18; [Epub ahead of print].
51. Pettifor A, van der Straten A, Dunbar MS, et al. Early age of first sex: a risk factor for HIV infection among women in Zimbabwe. AIDS. 2004;18:1435-1442.
52. Hargreaves JR, Glynn JR. Educational attainment and HIV-1 infection in developing countries: a systematic review. Trop Med Int Health. 2002;7:489-498.
53. Gupta N, Mahy M. Sexual initiation among adolescent girls and boys: trends and differentials in sub-Saharan Africa. Arch Sex Behav. 2003;32:41-54.
54. Stephenson R, Shutes E, Allen S, et al. The impact of project closure on HIV incidence and mortality in a cohort of couples in Lusaka, Zambia. AIDS Care. (In press).
55. Mark K, Meinzen-Derr J, Stephenson R, et al. Evaluating enhanced approaches to providing contraception among HIV concordant and discordant couples in Zambia: a randomized, controlled trial. J Womens Health. 2007;16:1187-1197.
56. Allen S. International data. J Acquir Immune Defic Syndr. 2005;38(Suppl 1):S7-S8.
57. Allen S. Why is fertility an issue for HIV-infected and at-risk women? J Acquir Immune Defic Syndr. 2005;38(Suppl 1):S1-S3.
58. Godfrey-Faussett P, Baggaley R, Mwinga A, et al. Recruitment to a trial of tuberculosis preventive therapy from a voluntary HIV testing centre in Lusaka: relevance to implementation. Trans R Soc Trop Med Hyg. 1995;89:354-358.
59. Deschamps MM, Johnson WD Jr, Pape JW. Feasibility and cohort development for HIV vaccine trials in Haiti. AIDS Res Hum Retroviruses. 1994;10(Suppl 2):S231-S233.
60. Fylkesnes K, Siziya S. A randomized trial on acceptability of voluntary HIV counselling and testing. Trop Med Int Health. 2004;9:566-572.
61. Homsy J, King R, Malamba SS, et al. The need for partner consent is a main reason for opting out of routine HIV testing for prevention of mother-to-child transmission in a rural Ugandan hospital. J Acquir Immune Defic Syndr. 2007;44:366-369.
62. Brown-Peterside P, Chiasson MA, Ren L, et al. Involving women in HIV vaccine efficacy trials: lessons learned from a vaccine preparedness study in New York City. J Urban Health. 2000;77:425-437.
63. El-Bassel N, Ivanoff A. Preventing HIV/AIDS in drug-abusing incarcerated women through skills building and social support. Soc Work Res. 1995;19:131-141.
64. Jenkins RA, Chinaworapong S, Morgan PA, et al. Motivation, recruitment, and screening of volunteers for a phase I/II HIV preventive vaccine trial in Thailand. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;18:171-177.
65. Kelly JA, Murphy DA, Washington CD, et al. The effects of HIV/AIDS intervention groups for high-risk women in urban clinics. Am J Public Health. 1994;84:1918-1922.

discordant couples; HIV; retention; sub-Saharan Africa; voluntary counseling and testing; women

© 2008 Lippincott Williams & Wilkins, Inc.