AIDS and Behavior, Vol. 4, No. 1, 2000
Correlates of HIV Test Results Seeking and Utilization of Partner Counseling Services in a Cohort of Male Factory Workers in Zimbabwe Rhoderick Machekano,1 William McFarland,2,5 Esther S. Hudes,2 Mary T. Bassett,1 Michael T. Mbizvo,3 and David Katzenstein4 Received Dec. 7, 1998; revised May 10, 1999; accepted June 21, 1999
Correlates of individual HIV test results-seeking and utilization of partner counseling services were identified among male factory workers who participated in a longitudinal HIV prevention intervention study in Harare, Zimbabwe. Men working at participating factories were offered HIV voluntary counseling and testing (VCT) for themselves and their partners. While risk assessment counseling was offered in the workplace, result disclosure and partner counseling occurred at an off-workplace location. Of the 3,383 men undergoing risk assessment and testing, 1,903 (56%) chose to receive their test results and 230 (7%) brought their partners for VCT. Factors associated with receiving test results were history of STD and lower salary. Factors associated with bringing a partner for VCT were history of STD, being married, being employed at a factory with a peer educator, lower salary, and no prostitute contact. Incorporating VCT into STD treatment services is likely to reach a large number of men and their partners at highest risk for both types of infections. Because men are often the main decision-makers in sexual and reproductive matters, VCT must be easily accessible to urban, working African men. KEY WORDS: HIV counseling and testing; Zimbabwe; factory workers; men; Africa.
INTRODUCTION
oping countries where the vast majority of HIV transmission occurs. Many of the purported benefits of VCT would also apply to sub-Saharan Africa. VCT serves an important role in primary prevention by assisting persons in reducing their risk for acquiring HIV infection (Sangiwa et al., 1998; Sweat et al., 1998). Counseling of HIV-infected individuals may also reduce their risk of transmitting infection to others. With the advent of more effective antiretroviral therapies, the earliest possible detection of HIV infection may reduce morbidity and prolong survival. Although antiretroviral therapies are generally not available in sub-Saharan Africa, VCT can help HIVpositive individuals formulate coping strategies. Additionally, more affordable, shorter course antiretroviral regimens for the prevention of mother-tochild transmission have been found to be effective in trials in Thailand and Coˆte d’Ivoire (Shaffer et al., 1999; Wiktor et al., 1999). Access to VCT will
Voluntary HIV counseling and testing (VCT) has been the cornerstone of HIV prevention in developed countries, garnering a considerable proportion of prevention resources (Holtgrave et al., 1993). However, VCT has not been as widely available in devel1
Zimbabwe AIDS Prevention Project, Department of Community Medicine, University of Zimbabwe Medical School, Harare, Zimbabwe. 2 Center for AIDS Prevention Studies, University of California, San Francisco, San Francisco, California. 3 Department of Obstetrics and Gynaecology, University of Zimbabwe Medical School, Harare, Zimbabwe. 4 Center for AIDS Research, Department of Infectious Diseases, Stanford University School of Medicine, Stanford, California. 5 Correspondence, including requests for reprints, should be addressed to William McFarland, M.D., Ph.D., Director of HIV Seroepidemiology; 25 Van Ness Avenue, San Francisco, California 94102-6033; e-mail:
[email protected]
63 1090-7165/00/0300-0063$18.00/0 2000 Plenum Publishing Corporation
64 therefore be needed in order to implement motherto-child HIV transmission prevention programs in sub-Saharan Africa. These potential benefits indicate that Zimbabwe, a country with very high prevalence and incidence of HIV, may benefit immensely from wider availability of VCT. Studies conducted in sub-Saharan Africa point to the importance of VCT in reducing risky behavior, promoting condom use, and lowering HIV incidence in serodiscordant couples and pregnant women (Allen et al., 1992a,b; Heyward et al., 1993). Preliminary results from a randomized study in Tanzania, Kenya, and Trinidad indicate that subjects who receive VCT are more likely to report reducing unprotected sex with nonprimary partners and prostitutes compared to subjects receiving health information only (Sangiwa et al., 1998). Studies in Rwanda find that condom use among couples is more common if the male partner had previously tested for HIV (Allen et al., 1992a,b). Moreover, HIV-negative women with untested male partners are unlikely to discuss or attempt to negotiate condom use (van der Straten et al., 1995). These studies recommend involvement of male partners in VCT since they often control sexual decisionmaking. To date, experience with the delivery of VCT directed primarily toward men in sub-Saharan Africa remains limited. In view of the fact that men in sub-Saharan Africa are often the primary decision-makers in issues related to sex and reproduction (Mbizvo and Bassett, 1996; van der Straten et al., 1995), it is imperative to characterize the delivery, acceptability, and impact of VCT in this population. Men working in the industrial sector in Africa are at increased risk for HIV because they often live separately from their partners and may have disposable income to spend on prostitutes. At the same time, working men in Africa are difficult to reach where their primary partners reside because they often migrate from rural to urban areas or internationally to find work. Because young men are concentrated in the workplace in urban Africa, the workplace may offer a suitable venue for offering HIV VCT. From 1993 to 1997, an HIV prevention intervention was conducted at factories in Harare, Zimbabwe to evaluate the impact of a peer education program on HIV seroconversion (Bassett et al., 1996; Mbizvo et al., 1996). The study provides an opportunity to assess the uptake of VCT targeted to a large population of urban African men at high risk. VCT was offered to factory workers and their partners, yet not all subjects chose to receive their results or make use
Machekano et al. of partner counseling services. The objectives of the present analysis are to describe the correlates of HIV test results-seeking behavior and the use of partner counseling and testing services among study participants.
METHODS Overview The Zimbabwe AIDS Prevention Project (ZAPP) recruited an open cohort of male workers from 40 factories in Harare, Zimbabwe to participate in the evaluation of a peer education intervention program (Bassett et al., 1996; Mbizvo et al., 1996). Recruitment occurred from March 1993 to December 1997. Follow-up of subjects occurred at approximately 6-month intervals beginning in September 1993 and ending in June 1997. All employees at the factories, an estimated workforce of 10,000, were eligible. A total of 3,383 male factory workers volunteered. Enrollment and follow-up entailed HIV serological testing with pretest counseling at the factory and posttest counseling at the project clinic. Subjects were encouraged to bring their sexual partners to the project clinic for VCT. In July 1994, the 40 factories were randomly assigned to intervention or control arms. Coworker peer education programs were implemented at intervention factories; offering of HIV counseling and testing continued in both control and intervention factories. Minimum goals of the peer education program were to have at least one active peer educator for every 100 workers, to maintain a continuous supply of free condoms at the worksite, to organize an AIDS prevention drama, and to arrange two presentations by persons with AIDS (one male, one female). Peer educators also led group and one-on-one discussions, distributed education materials, put up posters, and arranged video and slide show presentations. Study protocols were reviewed and monitored by committees for human research in Zimbabwe and the United States.
HIV Counseling HIV counseling included pretest (risk assessment and awareness counseling) and posttest (results disclosure counseling) sessions. Pretest counseling involved an individual risk assessment and the formula-
HIV Counseling and Testing in Zimbabwe tion of a plan of action for risk reduction. This oneto-one session took place in a private location at the workplace with a research nurse trained in HIV counseling techniques by the National AIDS Coordination Program and a private counseling service organization in Harare. Because the workplace was not considered an appropriate location to disclose HIV test results, participants were instructed and encouraged to come to the project clinic for disclosure counseling. Participants were told that their results would be available after 2 weeks, the clinic was open all working days as well as Saturdays, and the costs of transportation to and from the clinic would be reimbursed. Participants were also encouraged to come to the project clinic for STD treatment any time should they suspect that they had an infection. Posttest counseling took place at the project clinic centrally located in downtown Harare. Most factories are within 10 km of the project clinic. The posttest session involved disclosure of the most recent test results, a review of what the results meant, and discussion of HIV transmission modes and behaviors likely to result in transmission. Means of preventing transmission were also discussed and the individual plan on reducing risk behavior was reinforced. Subjects were again offered partner counseling and testing at the project clinic. Sexual partners of men enrolled in the study were pre- and posttest counseled at the study clinic after obtaining a separate informed consent. All counseling sessions were held in the language best understood by the subject (Shona, Ndebele, or English). After study end, it was no longer possible to encourage subjects to obtain their test results in person. Subjects seen after study end were counseled on their most recent results and, when appropriate, referred to VCT available outside the study for retesting.
Measures In this paper, we evaluate HIV test result-seeking behavior and bringing partners to VCT as primary outcome measures. After informed consent documentation, an interviewer administered a face-toface questionnaire to each subject, collecting demographic, sexual behavior, STD history, and condom use information. The individual risk reduction plan was recorded for a consecutive subsample of participants (N ⫽ 791). Approximately every 6 months, we revisited participants at their workplace, administered follow-up questionnaires on sexual behavior,
65 and collected blood specimens for repeat HIV antibody testing. Workers were also asked if they had received their previous HIV test results, and if not, the reason. Participants were again encouraged to visit the clinic for their new test results. At every visit to the project clinic, the reason for the participant’s visit was recorded.
Laboratory Methods Serum samples were tested for HIV-specific antibodies using a third-generation ELISA (Abbott HIV-1/HIV-2, Abbott Park, IL). If the sample was reactive or indeterminate, the sample was re-tested using another third-generation ELISA (Enzygnost Anti-HIV 1/2; Plus, Behring, Marburg, Germany). Samples were considered HIV-positive if the two ELISA tests were positive. We resolved inconclusive or discrepant results by Western blot (HIV 2.2, Diagnostic Biotechnology, Singapore) or polymerase chain reaction.
Statistical Methods Analyses were conducted using Stata software (Stata 5.0, College Station, TX). Demographic characteristics, reported behavior, and HIV serostatus were compared between subjects who chose to receive their HIV test results against those who did not, using the chi-square test for categorical variables or Wilcoxon rank sum test for continuous variables. Comparisons were also made between men who brought their sexual partners for HIV VCT and men who did not. We used stepwise logistic regression analysis to identify characteristics independently associated with HIV test result-seeking and partner counseling. For the multivariate analyses, the salary variable was log-transformed in order to decrease skewness. We used Stata’s ‘‘cluster’’ option with logistic regression in order to account for the statistical dependence of employees within the same factory.
RESULTS At baseline, 666 of 3,383 male factory worker participants were HIV-positive (HIV prevalence 20%, 95% confidence interval [CI], 19%–22%). By June 1997, 78% of subjects had at least one followup visit. During the follow-up period, 148 seroconver-
66 sions were observed in 6,984 person-years, giving an overall HIV incidence of 2.12 per 100 person years (95% CI, 1.79–2.49). Although HIV VCT was an integral part of study participation and all subjects enrolled consented to HIV testing, not all subjects chose to receive their HIV test results and disclosure counseling at the project clinic. Nearly all of the 3,383 participants (98%) said they wanted their test results, but only 1,903 (56%) ever visited the project clinic to obtain them. Among those who obtained their results, the median interval between the first time their blood was draw to the first time they were counseled on their results was 3.5 months (interquartile range: 3.5 weeks to 13.9 months). When a subsample of participants who did not return for results were interviewed on follow-up (N ⫽ 661), the most common reason offered for not obtaining test results was the inability to get time off from work (51%) followed by ‘‘missed the appointment’’ (26%). Being afraid to know their HIV serostatus was reported by 6%. Although all subjects who visited the project clinic were counseled on their HIV test results, only 44% of the visits were specifically motivated by the desire to obtain them. Of note, 11% of visits were motivated primarily by STD evaluation and 29% by other medical reasons. Table I summarizes bivariate comparisons between participants who chose to receive their HIV test results at any point in time and those who never did. Age, marital status, and years in school did not differ significantly between the two groups. Lower salary was associated with obtaining test results (Wilcoxon rank sum test, p ⬍ .001). Men who obtained their test results were more likely to have seroconverted to HIV-positive by study end (odds ratio [OR] 1.90, 95% CI 1.32–2.72), to report history of STD (OR 2.85, 95% CI 2.35–3.46), to have paid for sex (OR 1.27, 95% CI 1.09–1.49), and to have had multiple sex partners (OR 1.31, 95% CI 1.14–1.50). The presence of a peer educator in the workplace was not associated with individual test results-seeking. Only 230 (7%) of men brought their sex partners to the project clinic for HIV VCT during the study period (Table II). In bivariate analysis, men who brought in their partners for VCT were more likely to be married (OR 1.48, 95% CI 1.10–1.99), to have a lower salary (Wilcoxon rank sum test, p ⫽ .004), to have seroconverted (OR 1.98, 95% CI 1.19–3.31), to have had an STD (OR 3.41, 95% CI 2.59–4.50), to have had multiple partners (OR 1.46, 95% CI 1.11–
Machekano et al. 1.92), and to be employed at a factory with a peer educator (OR 1.37, 95% CI 1.04–1.79) compared to men who did not bring their partners for VCT. In multivariate analysis (Table III), reporting an STD (OR 2.78, 95% CI 2.25–3.43) and lower monthly salary (OR 0.85 per log[Z$100] increase, 95% CI 0.72–1.00) were independent predictors of individual test result-seeking. Independent predictors of bringing a partner for VCT were ever having an STD (OR 3.67, 95% CI 2.90–4.63), being married (OR 1.74, 95% CI 1.15–2.65), working at a factory with a peer educator (OR 1.39, 95% CI 0.94–2.07), lower monthly salary (OR 0.66 per log[Z$100] increase), and not paying for sex (OR 0.66, 95% CI 0.46–0.90). Table IV displays the risk reduction plans chosen by a consecutive subsample of 791 participants. The vast majority of men (72%) chose sticking to one partner as their risk reduction plan. Always using condoms, the next most common risk reduction plan, was chosen far less frequently (16% or less depending on partner type). Of note, only 50 men (6%) said they were going to get tested with their partner as part of a risk reduction plan.
DISCUSSION Despite bringing VCT to the workplace, reimbursement for travel to obtain results, multiple testing opportunities, frequent encouragement, and the fact that 98% of participants indicated that they wanted to know their serostatus, only 56% actually did obtain their HIV test results. Although time off from work was the most frequently cited reason for not obtaining results, offering results on Saturdays did not reduce this obstacle. In Zimbabwe, men often maintain rural and urban residences and travel between the two over weekends. The 2-week turnaround time for results is likely to be another substantial barrier. A study in Zambia showed that clients who had to wait 10 days or more for their HIV results expressed more fear over getting results and over 60% of participants indicated a preference to get their results the same day (McKenna et al., 1997). A recent study in Uganda found that the rapid HIV test resulted in a 27% increase in the proportion of clients who learned their serostatus and received posttest counseling (Kassler et al., 1998). Very few participants (7%) brought their female partners for VCT. Many men may have been unable to bring their wives to the project clinic if they resided in rural areas; unmarried men were less likely to bring
HIV Counseling and Testing in Zimbabwe
67
Table I. Individual Counseling. Demographic and HIV Risk-Related Characteristics of Men Who Chose to Receive Versus Chose Not to Receive Their HIV Test Results, Male Factory Worker Cohort, Harare, Zimbabwe, 1993–1997 Variables Total Want HIV test results No Yes Age group (years) 17–24 25–29 30–39 40–75 Marital status Single Married Years in school ⬍7 8–11 ⬎11 Monthly salary (Zimbabwe $) ⬍500 501–749 ⱖ750 HIV serostatus Negative Positive Seroconvertor Ever had an STD No Yes Ever paid for sex No Yes Ever reported multiple sex partners No Yes Presence of a peer educator at the workplace No Yes
Chose to receive HIV test results N (%)
Chose not to receive test HIV results N (%)
Odds ratioa (95% confidence interval)
P valuea
1,903 (56)
1,480 (44)
NA
—
27 (1) 1,862 (99)
39 (3) 1,421 (97)
Referent 1.89 (1.16–3.09) NA
— .010 .183
Referent 0.96 (0.83–1.10) NA
— .534 .801
NA
⬍.001
597 429 498 377
(31) (23) (26) (20)
450 332 367 331
(30) (23) (25) (22)
678 (36) 1,224 (64)
512 (35) 967 (65)
377 (20) 1,378 (72) 144 (8)
313 (21) 1,029 (70) 137 (9)
792 (42) 562 (30) 540 (29)
564 (38) 410 (28) 493 (34)
1,367 (73) 390 (21) 103 (6)
1,109 (78) 276 (19) 44 (3)
Referent 1.15 (0.96–1.36) 1.90 (1.32–2.72)
— .122 ⬍.001
1,412 (74) 491 (26)
1,319 (89) 161 (11)
Referent 2.85 (2.35–3.46)
— ⬍.001
1,378 (73) 513 (27)
1,135 (77) 332 (23)
Referent 1.27 (1.09–1.49)
— .003
863 (45) 1,040 (55)
771 (52) 709 (48)
Referent 1.31 (1.14–1.50)
— ⬍.001
967 (51) 936 (49)
770 (52) 710 (48)
Referent 1.05 (0.92–1.20)
— .484
a
Chi-square test except for age, years in school, and salary compared by Wilcoxon rank sum test as continuous variables. Categories do not always add up to total due to missing data.
b
their partners at all. Offering of VCT to men in the workplace did not appear to be an effective means of reaching female partners or counseling couples together, a finding that has implications for preventing HIV transmission between discordant couples (Allen et al., 1992b). Moreover, as men are often the primary decision-maker for having children as well as sex and condom use, programs to prevent mother-to-child HIV transmission will have to explore other ways to involve both partners with counseling on test results. The association of history of STD with obtaining
individual results and with partner counseling has implications for maximizing the benefit of HIV VCT in sub-Saharan Africa. Many men obtained their HIV test results after primarily visiting the clinic for STD evaluation. Offering VCT together with STD screening and treatment services therefore serves to attract persons at highest risk for HIV who may benefit most from risk reduction counseling. At the same time, effective STD treatment may be an important prevention strategy in some locations (Grosskurth et al., 1995; Wasserheit, 1992; Wawer et al., 1999). We also found that persons with STD are more likely to bring
68
Machekano et al. Table II. Partner Counseling. Demographic and HIV Risk-Related Characteristics of Men Who Brought Their Partners for HIV Counseling and Testing, Male Factory Worker Cohort, Harare, Zimbabwe, 1993–1997 Variables Totalb Want HIV test results No Yes Age group (years) 17–24 25–29 30–39 40–75 Marital status Single Married Years in school ⬍7 8–11 ⬎11 Monthly salary (Zimbabwe $) ⬍500 501–750 ⱖ750 HIV serostatus Negative Positive Seroconvertor Ever had an STD No Yes Ever paid for sex No Yes Ever reported multiple sex partners No Yes Presence of a peer educator at the workplace No Yes
Partner counseled N (%)
Partner not counseled N (%)
Odds ratioa (95% confidence interval)
230 (7)
3,151 (93)
NA
—
3 (1) 223 (99)
63 (2) 3060 (98)
Referent 1.53 (0.50–4.64) NA
— .471 .953
Referent 1.48 (1.10–1.99) NA
— .010 .959
NA
.004
67 58 63 42
(29) (25) (28) (18)
980 703 802 666
P valuea
(31) (22) (26) (21)
63 (27) 167 (73)
1,127 (36) 2,024 (64)
40 (17) 173 (75) 17 (7)
650 (21) 2,234 (71) 264 (8)
105 (46) 73 (32) 51 (22)
1,251 (34) 899 (29) 982 (31)
163 (72) 45 (20) 18 (8)
2,313 (76) 621 (20) 129 (4)
Referent 1.03 (0.73–1.45) 1.98 (1.19–3.31)
— .873 .009
133 (58) 97 (42)
2,598 (82) 555 (18)
Referent 3.41 (2.59–4.50)
— ⬍.001
171 (75) 58 (25)
2,342 (75) 787 (25)
Referent 1.00 (0.74–1.37)
— .953
91 (40) 139 (60)
1,543 (51) 1,610 (49)
Referent 1.46 (1.11–1.92)
— .006
102 (44) 128 (56)
1,653 (52) 1,518 (48)
Referent 1.37 (1.04–1.79)
— .028
a
Chi-square test except for age, years in school, and salary compared by Wilcoxon rank sum test as continuous variables. Categories do not always add up to total due to missing data.
b
their partners for HIV VCT, often in conjunction with partner STD treatment. VCT combined with STD services may therefore provide multiple opportunities for behavioral and biological HIV prevention interventions. The efficiency gained by such a strategy is particularly appealing in the resource-limited settings of developing countries. Other findings point to the relative uptake of individual and partner HIV VCT. Lower salary was associated with increased likelihood of individual test results-seeking and bringing in a partner for VCT. Offering STD treatment, the opportunity to consult
with a physician on other health problems, and condoms free of charge as well as reimbursement for transportation is likely to have created a relatively stronger incentive for persons with lower income than for those with higher income to visit the project clinic. Not surprisingly, partner counseling was higher among married men. Married men had more opportunities to bring their wives to the project clinic or were more able to initiate discussions of HIV with their wives compared to unmarried men with steady or casual partners. Although not directly measured, the borderline association between partner counsel-
HIV Counseling and Testing in Zimbabwe
69
Table III. Independent Predictors of HIV Test Result-Seeking Behavior and Partner Counseling, Male Factory Worker Cohort, Harare, Zimbabwe, 1993–1997 Variable
Adjusted odds ratioa (95% confidence intervalb)
P valueb
Model I: individual test result-seeking (N ⫽ 3,361) Ever had an STD Monthly salary (per ln Z$100 increase)
2.78 (2.25–3.43) 0.85 (0.72–1.00)
⬍.001 .048
Model II: partner counseling (N ⫽ 3,334) Ever had an STD Married Peer education factory Monthly salary (per ln Z$100 increase) Ever paid for sex
3.67 1.74 1.39 0.66 0.64
⬍.001 .009 .101 .003 .009
(2.90–4.63) (1.15–2.65) (0.94–2.07) (0.50–0.87) (0.46–0.90)
a
Odds ratios adjusted for other variables in the model. Adjusted for factory clustering.
b
ing and the presence of a workplace peer education program suggests that peer educators are reinforcing the importance of partner counseling to their coworkers. The association of no prostitute contact with increased likelihood of bringing a partner for VCT in the multivariate analysis (in contrast to the bivariate analysis) resulted from confounding with STD history. We recognize limitations of our data. Our study was not designed to assess the impact of VCT on HIV seroconversion and behavior change— outcomes which future studies will need to address to assess the prevention impact of VCT in the region. Nor was the study intended to directly assess the impact of peer education on HIV test result-seeking behavior and utilization of partner counseling services. We report observations in the context of a longitudinal cohort study using result-seeking behav-
Table IV. Risk Reduction Plans at Enrollment, Male Factory Worker Cohort, Harare, Zimbabwe, 1993–1997 (N ⫽ 791) Risk reduction plan Stick to one partner Always use condoms Get married No sex with prostitutes Always use condoms with casual partners Always use condoms with regular partners Other plan Abstain from sex Get tested with partner Avoid beer halls Always use condoms with commercial sex workers
Number of times chosen (%) 573 130 123 123 81
(72) (16) (16) (16) (10)
69 (9) 67 52 50 30 8
(8) (7) (6) (4) (1)
ior as a surrogate for HIV VCT-seeking behavior. Furthermore, we do not have information from workers who did not wish to participate in the cohort. As such, inference on VCT-seeking behavior for centers established in communities at large may be limited. Finally, many findings presented must be interpreted cautiously; the magnitude of differences between workers who obtained test results or used partner counseling services and those who did not may not be clinically relevant although they achieved statistical significance due to the large sample size. Of note, the largest effects observed in the present analysis were the associations between history of STD and individual results-seeking and history of STD and utilization of partner counseling services. Despite these limitations, the study represents the largest program delivering VCT in Zimbabwe to date and one of the few targeted specifically to men in sub-Saharan Africa. We believe that the lessons learned from this study are of importance to the expansion of publicly available VCT not only in Zimbabwe, but also to other countries in the region. In order to have the greatest impact, we recommend that HIV counseling and testing services be offered with STD treatment and should be easily accessible to working men and their partners. We concur with other researchers that the use of rapid HIV testing assays would reduce barriers due to time constraints, mounting fear of results, and the inconvenience of multiple trips to a testing center (Farnham et al., 1996; Kassler et al., 1998; Wilkinson et al., 1997). However, immediate, on-site disclosure of tests may not be appropriate for certain outreach or study settings such as the workplace. Researchers in Tanzania and Kenya have found that the cost of VCT (around $30 per client) and cost-effectiveness (around $250–$300
70 per infection averted) compare favorably with other prevention interventions (Sweat et al., 1998). The same study suggests that VCT is more cost-effective when targeted to HIV-positive persons, couples, and women. However, because men are often the main decision-makers for having sex and children, the impact of VCT on preventing sexual and mother-tochild HIV transmission in Africa will be severely blunted if the service is not easily accessible to urban working men.
ACKNOWLEDGMENTS We thank our Zimbabwe AIDS Prevention Project coinvestigators and team members, B. Brown, P. Chikukwa, L. Gwanzura, M. Jeche, S. Kagoro, A. Latif, C. Ley, M. Mandisodza, A. Mashingaidze, P. Mason, C. Maposhere, M. Matshaka, L. Moses, D. Mvere, V. Mzezewa, J. Parsonnet, S. Ray, L. Rogers, H. Smith, O. Tobaiwa, K. Vuraya, and S. Whytehead, for their dedicated input to this study.
REFERENCES Allen, S., Serufilira, A., Bogaerts, J., Van de Perre, P., Nsengumuremyi, F., Lindan, C., Carael, M., Wolf, W., Coates, T., and Hulley, S. (1992a). Confidential HIV testing and condom promotion in Africa: Impact on HIV and gonorrhea rates. Journal of the American Medical Association, 268, 3338–3343. Allen, S., Tice, J., Van de Perre, P., Serufilira, A., Hudes, E., Nsengumuremyi, F., Bogaerts, J., Lindan, C., and Hulley, S. (1992b). Effects of serotesting with counseling on condom use and seroconversion among HIV discordant couples in Africa. British Medical Journal, 304, 1605–1609. Bassett, M. T., McFarland, W. C., Ray, S., Mbizvo, M. T., Machekano, R., van de Wijgert, J. H., and Katzenstein, D. A. (1996). Risk factors for HIV infection at enrollment in an urban male factory cohort in Harare, Zimbabwe. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology, 13, 287–293. Farnham, P. G., Gorsky, R. D., Holtgrave, D. R., Jones, W. K., and Guinan, M. E. (1996). Counseling and testing for HIV prevention: Costs, effects, and cost-effectiveness of more rapid screening tests. Public Health Reports, 111, 44–54. Grosskurth, H., Mosha, F., Todd, J., Mwijarubi, E., Klokke, A., Senkoro, K., Mayaud, P., Changalucha, J., Nicoll, A., and Kagina, G. (1995). Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: Randomised controlled trial. Lancet, 346, 530–536. Heyward, W. L., Batter, V. L., Malulu, M., Mbuyi, N., Mbu, L., St Louis, M. E., Kamenga, M., and Ryder, R. W. (1993).
Machekano et al. Impact of HIV counseling and testing among child-bearing women in Kinshasa, Zaı¨re. AIDS, 7, 1633–1637. Holtgrave, D. R., Valdiserri, R. O., Gerber, A. R., and Hinman, A. R. (1993). Human immunodeficiency virus counseling, testing, referral, and partner notification services: A cost–benefit analysis. Archives of Internal Medicine, 153, 1225–1230. Kassler, W. J., Alwano-Edyegu, M. G., Marum, E., Biryahwaho, B., Kataaha, P., and Dillon, B. (1998). Rapid HIV testing with same-day results: A field trial in Uganda. International Journal of STD and AIDS, 9, 134–138. Mbizvo, M. T., and Bassett, M. T. (1996). Reproductive health and AIDS prevention in sub-Saharan Africa: The case for increased male participation. Health Policy and Planning, 11, 84–92. Mbizvo, M. T., Machekano, R., McFarland, W., Ray, S., Bassett, M. T., Latif, A., and Katzenstein, D. (1996). HIV seroincidence and correlates of serconversion in a cohort of male factory workers in Harare, Zimbabwe. AIDS, 10, 895–901. McKenna, S. L., Muyinda, G. K., Roth, D., Mwali, M., Ng’andu, N., Myrick, A., Luo, C., Priddy, F. H., Hall, V. M., Von Lieven, A. A., Sabatino, J. R., Mark, K., and Allen, S. A. (1997). Rapid HIV testing and counseling for voluntary testing centers in Africa. AIDS, 11, S103–S110. Sangiwa, G., Balmer, D., Furlonge, C., Grinstead, O., Kamenga, C., Coates, T., and the VCT Study Group. (1998). Voluntary HIV counseling and testing (VCT) reduces risk behavior in developing countries: Results from the voluntary counseling and testing study. Presented at the XII World AIDS Conference, Geneva, June 1998 [Abstract 33269]. Shaffer, N., Cuachoowong, R., Mock, P. A., Bhadrakong, C., Siriwasin, W., Young, N. L., Chotpitayawunondh, T., Chearskul, S., Roongpisuthipong, A., Chinayon, P., Karon, J., Mastro, T. D., and Simmonds, R. J. (1999). Short-course zidovudine for perinatal HIV-1 transmission in Bangkok, Thailand: A randomised controlled trial. Lancet, 353, 773–780. Sweat, M., Sangiwa, G., and Balmer, D. (1998). HIV counseling and testing in Tanzania and Kenya is cost-effective: Results from the voluntary HIV counseling and testing study. Presented at the XII World AIDS Conference, Geneva, June 1998 [Abstract 33277]. Van der Straten, A., King, R., Grinstead, O., Serufilira, A., and Allen, S. (1995). Couple communication, sexual coercion and HIV risk reduction in Kigali, Rwanda. AIDS, 9, 935–944. Wasserheit, J. N. (1992). Epidemiological synergy: Interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sexually Transmitted Diseases, 19, 61–77. Wawer, M. J., Sewankambo, N. K., Serwadda, D., Quinn, T. C., Paxton, C. A., Kiwanuka, N., Wawire-Mangan, F., Li, C., Lutalo, T., Nalugoda, F., Gaydos, C. A., Moulton, L. H., Meehan, M. O., Ahmed, S., the Rakai Project Study Group, and Gray, R. H. (1999). Control of sexually transmitted diseases for AIDS prevention in Uganda: A randomised community trial. Lancet, 353, 525–535. Wiktor, S. Z., Ekpini, E., Karon, J. M., Nkengasong, J., Maurice, C., Severin, S. T., Roels, T. H., Kouassi, M. K., Lackritz, E. M., Coulibaly, I. M., and Greenberg, A. E. (1999). Shortcourse oral ziduvudine for prevention of mother-to-child transmission of HIV-1 in Abidjan, Cote d’Ivoire: A randomised trial. Lancet, 353, 781–785. Wilkinson, D., Wilkinson, N., Lombard, C., Martin, D., Smith, A., Floyd, K., and Ballard, R. (1997). On-site HIV testing in resource-poor settings: Is one rapid HIV test enough? AIDS, 11, 377–381.