The prevalence of HIV infection in sub-Saharan Africa is the highest in the world, yet most HIV-1-infected persons in this region do not know their infection status.1,2 Persons who present late in the course of their HIV disease have significant short-term mortality.3 Early diagnosis of HIV infection is increasingly understood as a critical gateway to appropriate antiretroviral therapy (ART) provision and effective prevention. Furthermore, most HIV transmission occurs from persons unaware that they are infected.4 Early recognition of HIV infection provides enormous personal and public health benefit.
In the United States and Europe, active provider-assisted partner notification has become a key HIV prevention strategy leading to increased HIV counseling and testing among sexual partners of patients with new HIV diagnoses.5-7 Generally, 3 methods of partner notification are available as follows: passive referral, contract referral, and provider referral.8 With passive referral, the patient is encouraged to disclose the exposure of their partner(s) to HIV by themselves. Under contract referral, health care providers allow the index patient a short period of time to contact, notify and refer sexual partners, after which a health care provider advises the contact of their exposure although maintaining the anonymity of the index case. Under provider referral, a health care provider contacts the partners immediately and directly but with anonymity.
In sub-Saharan Africa, the effectiveness of partner notification strategies has not been evaluated.9 Passive referral, the standard of care in Africa, has had minimal success.10 Use of active partner notification has been limited by concerns regarding privacy protection and social harm and apparent lack of community and political support. However, the potential benefit of partner notification is evident. In antenatal and postpartum clinics, disclosure of HIV status by women has improved prevention behaviors including condom use,11-13 uptake of prevention of mother-to-child transmission activities,12,13 and decision-making regarding subsequent pregnancies.11 Clearly, the potential public health benefit of partner notification in sub-Saharan Africa is substantial.
We compared patient referral, contract referral, and provider referral among patients with newly diagnosed HIV in a sexually transmitted infections (STIs) clinic setting in Malawi.
Persons with newly diagnosed HIV infection at Kamuzu Central Hospital and Bwaila Hospital outpatient STIs clinics in Lilongwe, Malawi, were recruited. Lilongwe is the capital city of Malawi with a population of approximately 900,000. All patients presenting to these STIs clinics are tested for HIV under an opt-out protocol that includes group pretest counseling, rapid tests (Determine HIV-1/2, Abbott Laboratories, Chicago, IL and Unigold, Trinity Biotech, Bray, Ireland) and individual posttest counseling. Patients from Lilongwe who had a positive HIV test result for the first time, were 18 years or older, had been sexually active in the last 90 days, were willing and able to provide locator information for their sexual partners, and agreed to be randomized to a method of partner notification were eligible to participate.
Index patients provided informed consent and answered a short questionnaire about recent sexual behavior, including the number, type, and locations of sexual partners in the past 3 months. All were provided referral cards to give to their partners, were counseled on the importance of safe sex behavior, staged using WHO clinical staging criteria, and had blood drawn for CD4 counts using flow cytometry (Epics-XL; Coulter, Brea, CA). Index patients were then randomized to passive, contract, or provider referral using a permuted block design with randomly allocated block sizes of 6, 9, and 12, stratified by sex and study site. The passive referral group was responsible for notifying their partners themselves. The contract referral group was given 7 days to notify their partners, after which a health care provider contacted partners who had not reported for counseling and testing and counseled them to visit the clinic although maintaining the anonymity of the index case. Notification in the provider referral group occurred within 48 hours. Community outreach workers who are trained HIV testing counselors or nurses performed notification. A standard protocol for community contact was followed. Randomization assignment was concealed in a sealed envelope until the end of the enrollment visit (after all partner data and locator information had been collected).
Index patients returned to the clinic 2 weeks after enrollment to receive CD4 test results and initiate HIV care. Index patients eligible for ART based on CD4 count or WHO clinical stage were referred to a convenient HIV treatment clinic. Index patients were also asked whether their partners were notified, how their partners were notified, and their knowledge of their partners HIV counseling and testing behavior.
Incoming patients were identified as partners if they presented a partner referral card or their name was found on the log of named partners during cross checking. Partners were tested for HIV under the opt-out testing protocol that is standard of care in the clinic. HIV antibody-negative or antibody-indeterminate specimens were tested for the presence of HIV RNA using the ultrasensitive Roche Amplicor Monitor HIV RNA assay.
Partners were considered “locatable” if the index was able to provide locator information, including name and where they could be found, during enrollment. Main partners were defined as spouses and live-in partners or boyfriend/girlfriend if the index did not name a spouse or live-in partner. Casual partners included regular casual partners, infrequent casual partners, sex workers, and boyfriend/girlfriend if the index already had a spouse. Partners were considered new HIV diagnoses if they were testing for the first time or their previous test result had been negative. The primary outcome was partner visit to the clinic during the 30 days after index enrollment.
Based on the assumption of 1 partner per index client and 15% of the partners in the passive referral arm presenting, the overall sample size of 240 index patients with 80 index patients in each arm was calculated before the start of the study to have 85% power to detect (α = 0.05, 2-sided test) an absolute difference of 25% between passive referral and the 2 active referral study arms (contract or provider).
Unconditional logistic regression with a cluster robust variance estimator14 was used to calculate 95% confidence intervals (CIs) for the proportion of locatable partners visiting by arm and risk differences and relative risks of visiting for the 2 active referral arms versus the passive referral arm. Preplanned subgroup analyses were performed by sex and type of partner (main partner vs. casual partner). Planned sensitivity analyses where the unit of analysis was the index case (rather than the named partner of an index case) and the primary outcome was defined as at least 1 partner visiting the clinic were conducted. Time to presentation among all locatable partners was analyzed using Cox Proportional Hazards Regression with robust CIs15 to account for clustering by index patient. The Wald χ2 test was used to compare the effect of method of partner notification on time to presentation. The proportional hazards assumption was evaluated using the Cox test and visually by plotting the ln [-ln (survival)] against ln (time). We used Stata version 10 (StataCorp, College Station, TX) for all analyses.
The Institutional Review Board at the University of North Carolina, Chapel Hill and the National Health Sciences Research Committee in Malawi approved the protocol. Informed consent was obtained from all participants before participation.
We recruited 240 newly diagnosed HIV-positive men and women between October 2, 2008, and September 2, 2009. Of 401 persons attending the clinics with a diagnosis of HIV infection, 267 met eligibility criteria. Reasons for ineligibility included living outside Lilongwe, no sexual partners in the previous 3 months, no sexual partners in Lilongwe, they came for testing as a couple, or had tested positive previously. Of those eligible, 240 (89%) enrolled in the study. Reasons for refusal to participate included not having time (26%), did not want to (17%) or afraid (3%) to notify partners, and wanting time to think about it (23%). Refusers were similar in sex (P = 0.2), age (P = 0.3) and marital status (P = 0.8) to participants but had more median years of education (10 years vs 8 years) (P = 0.04).
Among index patients, 58.3% were female and 71.2% were married (Table 1). The median CD4 count at HIV diagnosis was 317.5 cells per cubic millimeter (range 25-1254). Index patients named 1-11 sexual partners in the previous 3 months, although most named a single partner (86%). Three index patients randomized to the provider referral arm named 5, 8, and 10 sex workers as partners for whom they could not provide basic locator information. No other index named more than 3 partners.
Overall, 302 partners were named including 219 (73%) main partners. The index reported planning to have sex again with 220 (73%) of the partners. The median partnership duration was 24 months [interquartile range (IQR: 3-84)]. Condom use was low; only 15% of index patients reported condom use at last sex and 76% reported never using condoms with their partner (Table 2).
Among 302 named partners, locator information was available for 252 (84%). Compared with nonlocatable partners, locatable partners were more likely to be spouses (64% vs. 0%), male (56% vs. 10%), and have a duration of the partnership >1 month (79% vs. 7%).
Overall, 107 (35%) partners visited the clinic. Partner presentation, including nonlocatable partners, was 20 of 93 (22%; 95% CI: 13% to 30%) in the passive referral arm, 45 of 94 (48%; 95% CI: 38% to 58%) in the contract referral arm, and 42 of 115 (37%; 95% CI: 28% to 45%) in the provider referral arm. Restricting the analysis to locatable partners, the proportion of partners visiting was 20 of 82 (24%; 95% CI: 15% to 34%) in the passive referral arm, 45 of 88 (51%; 95% CI: 41% to 62%) in the contract referral arm, and 42 of 82 (51%; 95% CI: 40% to 62%) in the provider referral arm (Table 3). None of the partners classified as nonlocatable returned. Among locatable partners, those in the contract and provider referral arms were both 2.1 times as likely to visit the clinic compared with those in the passive referral arm (contract versus passive: risk ratio: 2.1; 95% CI: 1.4 to 3.2; P < 0.001; provider versus passive: risk ratio 2.1; 95% CI: 1.4 to 3.2; P < 0.001). The proportion of partners visiting the clinic was 27% higher in both the contract and provider referral arms, as compared with passive referral (contract versus passive: risk difference: 27%; 95% CI 13% to 41% P < 0.001; provider versus passive: risk difference 27%; 95% CI: 13% to 41%; P < 0.001). The proportion of index patients with at least 1 partner visiting the clinic for counseling and testing was 26% (95% CI: 16% to 35%) in the passive referral arm, 55% (95% CI: 44% to 66%) in the contract referral arm, and 51% (95% CI: 40% to 62%) in the provider referral arm.
Time to presentation among partners was associated with method of partner notification (P < 0.001) (Fig. 1). The hazards were not proportional over time; so hazard ratios were estimated separately for partner visit in the first 7 days and after 7 days. In the first 7 days, locatable partners in the contract referral arm returned at a similar rate to the passive referral arm [hazard ratio 1.4 (95% CI: 0.7 to 2.6)], but locatable partners in the provider referral arm returned at a higher rate [hazard ratio (HR): 2.1 (95% CI: 1.1 to 3.7)] compared with partners in the passive referral arm. After 7 days, locatable partners in both the contract referral [HR: 6.6 (95% CI: 2.3 to 18.8)] and provider referral arms [HR 4.3 (95% CI 1.4 to 13.0)] returned at a higher rate than partners in the passive referral arm. The median time between enrollment of the index and partner presentation among those who visited the clinic was 3 days in the passive referral arm (IQR: 2-7 days), 7 days in the contract referral arm (IQR: 3-11 days), and 4 days in the provider referral arm (IQR: 2-8 days). In the contract referral arm, 30 (67%) partners who reported for counseling and testing were traced by a community counselor.
The acceptance rate for HIV testing among partners seen in the clinic was high. Overall, 104 (97%) of partners accepted HIV testing, and 67 (64%) tested HIV positive; 1 partner was identified as acutely infected based on HIV RNA in the blood, lack of HIV antibodies and subsequent seroconversion. Twelve partners (15%; 95% CI: 7% to 22%) in the passive referral arm, 21 partners (24%; 95% CI: 15% to 33%) in the contract referral arm, and 21 partners (26%; 95% CI: 16% to 35%) in the provider referral arm were new HIV diagnoses; and 13 partners had previously tested positive. The median CD4 count among partners was 344 (range: 47-940). Twenty-eight percent of partners were eligible to start ART based on the current Malawi treatment guidelines (CD4 < 250 cells/mm3) (Table 4). Most partners reported only a single sexual partner, with 85 (82%) reporting one sexual partner in the previous 3 months.
Index patients and partners reported 2 social harms. In 1 instance, a female index reported her male partner (spouse) abandoned her when she disclosed her HIV status. In the other, a female partner called the police when the community counselor visited the home. The situation was quickly resolved, and the partner later sought counseling and testing at the clinic.
The HIV pandemic in Africa has gone unabated for more than 20 years, despite massive prevention efforts.16 The introduction of ART in recent years has undoubtedly benefited many patients, but frequently patients receive therapy too late for maximal benefit. Currently, access to ART is increasing, and ART as a prevention tool has been supported.17 However, optimal treatment and prevention require that infected people know their status. To achieve this goal, novel strategies of massive household testing18 and couples counseling19 have been investigated. Partner notification, a logical and potentially critical intervention, has not been evaluated.
In Malawi, we observed that provider-assisted methods of HIV partner notification are feasible, acceptable, and effective among STIs clinic patients. A high proportion of eligible patients participated and provided accurate partner locator information. Provider-assisted partner notification was implemented without difficulty and was supported by clinic staff. Provider-assisted partner notification resulted in more partners receiving counseling and testing services than passive referral, the current standard of care.
Partner notification increased early referral to care. About one quarter of infected partners were eligible to begin ART based on current Malawian national guidelines and half of all infected partners had a CD4 count of 350 cells per cubic millimeter or less. Given that mortality is significantly increased in late presenters and baseline CD4 count is a strong predictor of response to ART and mortality,20 the population of partners identified in this study are highly likely to benefit from knowledge of their status.
Prevention of HIV transmission within serodiscordant partnerships is an important HIV prevention strategy. In our study, 45% of tested partners were in a serodiscordant relationship. Serodiscordant couples receiving couples counseling report increased condom use and lower rates of seroconversion.21,22 Partner notification may be an effective strategy to facilitate both individual behavior change among uninfected individuals and increase couples counseling in the region.
Provider-assisted partner notification is an important method to increase testing among male partners. Although 50% of male partners sought evaluation in the contract and provider referral arms, only 15% did in the passive referral arm. Extending provider-assisted notification to antenatal clinic settings may be a novel way to increase male involvement in prevention of mother-to-child transmission (PMTCT) programs. To date, male involvement in PMTCT has been low in sub-Saharan Africa23,24 and continues to be difficult to implement. When male partners are involved or couples counseling is provided during PMTCT, HIV testing uptake is higher and women are more likely to implement PMTCT treatment and care interventions.25
The potential for social harms is a key concern in partner notification programs. However, the index patients and partners in our study reported only 2 social harms throughout the entire study period, a 0.5% cumulative incidence. Experience elsewhere in the region suggests social harms are not increased among women in couples antenatal testing compared with women who do not disclose to their partner,26 and in South Africa, men and women who disclosed their HIV status reported an increase in social support.27 Experience in the United States suggests partner notification does not increase partnership dissolution.28 However, prior history of abuse in a relationship after disclosure of HIV status is a strong predictor of reported physical or emotional abuse after disclosure.29 Screening for intimate partner violence and emotional abuse could be incorporated into posttest counseling, and further operations research will be necessary to investigate the effect of provider assisted partner notification on social harms in a variety of African settings.
The STIs clinic population in this study may not be representative of all new HIV diagnoses. Partners may be more motivated to respond to notification messages because of potential for STIs treatment. However, partners of persons testing positive, although seeking treatment for STIs are important to target for increased counseling and testing, as infectiousness is high in HIV-infected individuals with a concurrent STIs. Implementation of HIV partner notification in STIs clinics should go along with partner notification for STIs treatment.
Successful partner notification is contingent on index patients providing locator information for their partners and community counselors successfully locating partners. In our study population, index patients often did not know the name or location of one-time or short-term partners. These partners are unlikely to be notified by the index and are unable to be traced by community counselors. Unfortunately, this group may represent high transmitter populations. As provider-assisted partner notification techniques are further refined in this setting, techniques to elicit more accurate locator information and find partners will be improved, and a larger proportion will be expected to be located and receive counseling and testing.
The 50% partner return rate compares favorably with other evaluations of HIV partner notification in developed countries6 and is a conservative estimate of the number of partners who received testing services as it is possible partners may have tested elsewhere. Counseling index patients to advise their partners to visit the study clinics minimized partners testing elsewhere, and the partner cards clearly described the location of the study clinics. Partner cards also included instructions for area testing centers to return the cards to the study clinics if partners presented the cards at those locations and study staff periodically visited other major testing sites in Lilongwe to determine if partners had tested there. No index patients reported partners who tested at other sites.
Early evaluation of a partner is preferred because of prevention opportunities. Provider referral led to more rapid clinic visits. Partners who visit on their own volition were seen in the first week; most partners who returned in the passive referral arm did so within the first week; and a third of partners sought counseling and testing on their own when contract referral was used. Accordingly, provider-assisted referral can be used to improve return after about one week.
This study provides the first evidence of the effectiveness of partner notification in sub-Saharan Africa. Partner notification, including provider-assisted strategies, has recently been implemented in Cameroon and more than 2000 partners have been evaluated,30 further supporting the feasibility of partner notification in non-Western settings. The passive referral approach to partner notification has not been successful. Active partner notification strategies, such as contract or provider referral, are required. More aggressive partner notification has the potential to rapidly and efficiently expand HIV treatment and prevention. Active partner notification is an opportunity that cannot continue to be missed.
1. Measure DHS. Malawi Demographic and Health Survey 2004
. Calverton, MD: National Statistical Office (Malawi) and ORC Macro; 2005.
2. Bunnell R, Opio A, Musinguzi J, et al. HIV transmission risk behavior among HIV-infected adults in Uganda: results of a nationally representative survey. AIDS
3. Girardi E, Sabin CA, Monforte AD. Late diagnosis of HIV infection: epidemiological features, consequences and strategies to encourage earlier testing. J Acquir Immune Defic Syndr
. 2007;46(Suppl 1):S3-S8.
4. Marks G, Crepaz N, Janssen RS. Estimating sexual transmission of HIV from persons aware and unaware that they are infected with the virus in the USA. AIDS
5. Hogben M, McNally T, McPheeters M, et al. The effectiveness of HIV partner counseling and referral services in increasing identification of HIV-positive individuals a systematic review. Am J Prev Med
. 2007;33(2 suppl):S89-S100.
6. Landis SE, Schoenbach VJ, Weber DJ, et al. Results of a randomized trial of partner notification in cases of HIV infection in North Carolina. N Engl J Med
7. European Partner Notification Study Group. Recently diagnosed sexually HIV-infected patients: seroconversion interval, partner notification period and a high yield of HIV diagnoses among partners. Qjm
8. Mathews C, Coetzee N, Zwarenstein M, et al. Strategies for partner notification for sexually transmitted diseases. Cochrane Database Syst Rev
9. Mathews C, Coetzee N, Zwarenstein M, et al. A systematic review of strategies for partner notification for sexually transmitted diseases, including HIV/AIDS. Int J STD AIDS
10. Medley A, Garcia-Moreno C, McGill S, et al. Rates, barriers and outcomes of HIV serostatus disclosure among women in developing countries: implications for prevention of mother-to-child transmission programmes. Bull World Health Organ
11. Kumar A, Waterman I, Kumari G, et al. Prevalence and correlates of HIV serostatus disclosure: a prospective study among HIV-infected postparturient women in Barbados. AIDS Patient Care STDS
12. Farquhar C, Kiarie JN, Richardson BA, et al. Antenatal couple counseling increases uptake of interventions to prevent HIV-1 transmission. J Acquir Immune Defic Syndr
13. Farquhar C, Mbori-Ngacha DA, Bosire RK, et al. Partner notification by HIV-1 seropositive pregnant women: association with infant feeding decisions. AIDS
14. Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics
15. Lin DY. Cox regression analysis of multivariate failure time data: the marginal approach. Stat Med
16. Cohen MS, Hellmann N, Levy JA, et al. The spread, treatment, and prevention of HIV-1: evolution of a global pandemic. J Clin Invest
17. Granich RM, Gilks CF, Dye C, et al. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet
18. Bunnell R, Mermin J, De Cock KM. HIV prevention for a threatened continent: implementing positive prevention in Africa. JAMA
19. 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. BMC Public Health
20. Hogg RS, Yip B, Chan KJ, et al. Rates of disease progression by baseline CD4 cell count and viral load after initiating triple-drug therapy. JAMA
21. 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
22. 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
23. Msuya SE, Mbizvo EM, Hussain A, et al. Low male partner participation in antenatal HIV counselling and testing in northern Tanzania: implications for preventive programs. AIDS Care
24. Theuring S, Mbezi P, Luvanda H, et al. Male involvement in PMTCT services in Mbeya Region, Tanzania. AIDS Behav
. 2009;13(Suppl 1):92-102.
25. Becker S, Mlay R, Schwandt HM, et al. Comparing couples' and individual voluntary counseling and Testing for HIV at Antenatal Clinics in Tanzania: a Randomized Trial. AIDS Behav
26. Semrau K, Kuhn L, Vwalika C, et al. Women in couples antenatal HIV counseling and testing are not more likely to report adverse social events. AIDS
27. Wong LH, Rooyen HV, Modiba P, et al. Test and tell: correlates and consequences of testing and disclosure of HIV status in South Africa (HPTN 043 Project Accept). J Acquir Immune Defic Syndr
28. Hoxworth T, Spencer NE, Peterman TA, et al. Changes in partnerships and HIV risk behaviors after partner notification. Sex Transm Dis
29. Gielen AC, McDonnell KA, Burke JG, et al. Women's lives after an HIV-positive diagnosis: disclosure and violence. Matern Child Health J
30. Tih PM, Forgwei G, Welty T, et al. Integrated HIV contact tracing and partner notification in Cameroon: a feasible HIV infection risk reduction intervention for resource-poor settings. Paper presented at: 5th IAS Conference on HIV Pathogenesis and Treatment; 2009; Cape Town, South Africa.