Promotion of couples' voluntary counselling and testing for HIV through influential networks in two African capital cities.

BACKGROUND: Most new HIV infections in Africa are acquired from cohabiting heterosexual partners. Couples' Voluntary Counselling and Testing (CVCT) is an effective prevention strategy for this group. We present our experience with a community-based program for the promotion of CVCT in Kigali, Rwanda and Lusaka, Zambia. METHODS: Influence Network Agents (INAs) from the health, religious, non-governmental, and private sectors were trained to invite couples for CVCT. Predictors of successful promotion were identified using a multi-level hierarchical analysis. RESULTS: In 4 months, 9,900 invitations were distributed by 61 INAs, with 1,411 (14.3%) couples requesting CVCT. INAs in Rwanda distributed fewer invitations (2,680 vs. 7,220) and had higher response rates (26.9% vs. 9.6%), than INAs in Zambia. Context of the invitation event, including a discreet location such as the INA's home (OR 3.3-3.4), delivery of the invitation to both partners in the couple (OR 1.6-1.7) or to someone known to the INA (OR 1.7-1.8), and use of public endorsement (OR 1.7-1.8) were stronger predictors of success than INA or couple-level characteristics. CONCLUSION: Predictors of successful CVCT promotion included strategies that can be easily implemented in Africa. As new resources become available for Africans with HIV, CVCT should be broadly implemented as a point of entry for prevention, care and support.

Influence network effectiveness in promoting couples' HIV voluntary counseling and testing in Kigali, Rwanda.

OBJECTIVE: To identify predictors of promotion of couples' HIV voluntary counseling and testing (CVCT) in Kigali, Rwanda. DESIGN: Analysis of CVCT promotional agent [influential network leaders (INLs), influential network agents (INAs)], and couple/invitation-level predictors of CVCT uptake. METHODS: Number of invitations and couples tested were evaluated by INL, INA, and couple/contextual factors. Multivariable logistic regression accounting for two-level clustering analyzed factors predictive of couples' testing. RESULTS: Twenty-six INLs recruited and mentored 118 INAs who delivered 24 991 invitations. 4513 couples sought CVCT services after invitation. INAs distributed an average of 212 invitations resulting in an average of 38 couples tested/agent. Characteristics predictive of CVCT in multivariate analyses included the invitee and INA being socially acquainted [adjusted odds ratio (aOR) = 1.4; 95% confidence interval (CI) 1.2-1.6]; invitations delivered after public endorsement (aOR = 1.3; 95% CI 1.1-1.5); and presence of a mobile testing unit (aOR = 1.4; 95% CI 1.0-2.0). In stratified analyses, predictors significant among cohabiting couples included invitation delivery to the couple (aOR = 1.2; 95% CI 1.0-1.4) and in the home (aOR = 1.3; 95% CI 1.1-1.4), whereas among noncohabiting couples, predictors included invitations given by unemployed INAs (aOR = 1.7; 95% CI 1.1-2.7). Cohabiting couples with older men were more likely to test, whereas younger age was associated with testing among men in noncohabiting unions. CONCLUSIONS: Invitations distributed by influential people were successful in prompting couples to seek joint HIV testing, particularly if the invitation was given in the home to someone known to the INA and accompanied by a public endorsement of CVCT. Mobile units also increased the number of couples tested. Country-specific strategies to promote CVCT programs are needed to reduce HIV transmission among those at highest risk for HIV in sub-Saharan Africa.

Perinatal mortality associated with abruptio placenta in singletons and multiples.

OBJECTIVE: This study was undertaken to investigate the association among plurality (number of fetuses per pregnancy), abruptio placenta, and perinatal mortality. STUDY DESIGN: A retrospective cohort study on 15,051,872 singletons, 413,619 twins, and 22,585 triplets delivered in the United States between 1995 and 1998 was conducted. We compared the occurrence of perinatal death between pregnancies complicated by abruptio placenta and those without with the use of adjusted odds ratios. The generalized estimating equations framework was applied to adjust for intracluster correlations among multiples. RESULTS: Placental abruption occurred among 93,968 singletons (6.2 per 1000), 5051 twin (12.2 per 1000), and 353 triplet (15.6 per 1000) gestations ( P for trend<.0001). Placental abruption was associated with significant risk of mortality irrespective of the plurality subtype. Perinatal mortality was greatest among singletons (adjusted odds ratio [95% CI]=14.3[13.2-15.4]), followed by twins (4.4[3.9-4.9]) and least among triplets (3.0[2.0-4.6]) ( P for trend<.0001). CONCLUSION: As plurality increases from 1 to 3, the risk of placental abruption rises, whereas the risk of abruptio-associated perinatal mortality declines.