Male partner voluntary counselling and testing associated with the antenatal services in Kinshasa,Democratic Republic of Congo: a randomized controlled trial

Study site

The study was conducted at Kingasani Maternity, Kinshasa, DRC. VCT in the context of PMTCT was introduced in 2002. About 10,000 deliveries occur each year in this clinic. More than 95% of all pregnant women attending ANC at Kingasani currently undergo VCT, but less than 2% of male partners invited for VCT have been tested over the past three years. The HIV prevalence among women presenting at ANC in Kinshasa is 4.3%. ¹⁶ In the present study male partner VCT was conducted in three types of venues: bars, churches and neighbourhood health centres. Neighbourhood, availability of space for VCT, frequentation by men and willingness of the site's owner were the main criteria for venue selection.

Participants

Participants were enrolled from 1 September 2006 to 31 January 2007. Pregnant women presenting for ANC at Kingasani Maternity were enrolled if they were at least 18-years-old and consented to participate in the study. Those women who reported not knowing the father of the unborn baby or where the father of the unborn baby did not live in the city were excluded from the study. The male partner in this study was the father of the unborn child who was identified by the woman and to whom she brought the invitation following a routine ANC visit. We invited all adult men (aged at least 18 years) whose female sexual partners had received VCT at Kingasani Maternity.

Intervention

We designed a three-arm RCT corresponding to the three venues where the male partners were randomly invited via their female partners. The three arms consisted of: (1) neighbourhood health centre-based VCT; (2) bar-based VCT; and (3) church-based VCT. Pregnant women who presented to the ANC were given general information regarding the study prior to consenting for the study. Simple randomization was used to prepare the allocation sequence. Sealed envelopes containing the invitations for VCT to the three different venues were prepared in advance by a study supervisor.

During a routine ANC visit, consenting women picked up an envelope containing an invitation with corresponding arm number. Consenting women randomly selected invitations. Trained study nurses working at Kingasani Maternity performed the enrolment interviews/counselling for all consenting pregnant women.

Considering that most men living in the Kingasani neighbourhood work far away from their home, we invited male partners to the venue after normal working hours between 15:00 and 20:00. Pregnant women did not come to any of the three venues when their husbands presented at a specific venue for VCT.

Once at the venue, confidential pre-test counselling took place in a private room where patient confidentiality could be guaranteed. Trained counsellors conducted a pretest group confidential counselling session, consisting of: explanations about HIV/AIDS, client risk evaluation, risk reduction plan and phlebotomy for rapid HIV testing. This was followed by post-test group counselling, 30 minutes after phlebotomy. Services were free of charge and a reimbursement for transportation expenses was provided.

During the pretest counselling, every male partner had the opportunity to ask confidential questions before consenting to voluntary HIV testing. A trained counsellor conducted each VCT session according to the national AIDS Control Program guidelines, including collection of sociodemographic information and HIV-risk evaluation. A peripheral venous blood sample (∼1.5 cm³) was drawn from consenting male partners for HIV rapid testing, using the Determine HIV 1/2 (Abbott Diagnostics, Wiesbaden, Germany), Double Check Gold HIV1&2 (Orgenics Ltd, Yavne, Israel) and Unigold HIV (Trinity Biotech PLC, Bray, Ireland). The HIV tests results were provided to the clients the same day, and an invitation for a couple counselling visit within a week was extended to each client.

Couple counselling was conducted in a specific venue in two steps by a trained counsellor. Step one involved an individual session consisting of knowledge sharing about HIV/AIDS; risk reduction plan, informed consent for disclosure and psychological support. The second step (couple counselling session) consisted of a short discussion of knowledge regarding HIV/AIDS. The counsellor used the opportunity to encourage each client to share her/his HIV result with the partner. HIV discordant and HIV-positive concordant couples were referred to HIV/AIDS clinics in their immediate neighbourhood for care and support.

Outcomes

The primary outcome of the trial was male participation in VCT. We also sought to identify factors associated with uptake of couple counselling and determinants of male participation and couple counselling.

Sample size

Male participation was estimated to be 15% in the health centre-based VCT arm, and ∼20% in one of the two arms (bar VCT or church-based VCT). With 80% power, 0.05 significance level and a 5% difference in male participation between the health centre and one of the two other arms, the required sample size was ∼901 women per arm.

Data analysis

Epi-info software 2000 (Ear Information™ 3.4.3; Centers for Disease Control and Prevention [CDC]) was used for the analysis. Descriptive statistics were performed. The mean with standard deviation and median with quartiles were used for quantitative variables. Categorical variables were summarized using proportions.

Unadjusted analyses were performed to assess the association between the intervention and the outcome: odds ratios (OR) were used to estimate the effect sizes. Chi square (χ ²) test with a P value <0.05 were used to test differences in proportions.

Multiple logistic regression was used to identify determinants of male participation (including female characteristics) and uptake of couple counselling (including male and female characteristics). Variables found to be significant in univariate analysis (P < 0.10) were included in the multivariate model. The final model was obtained by stepwise backwards elimination and the interaction terms were tested.

Ethical issues

Informed consent was obtained from all participants in the study. Participants received no incentive to participate in the study. Only transportation fees were refunded for those who participated in the couples counselling. The study protocol was approved by the Institutional Review Board of the Kinshasa School of Public Health.

Demographic characteristics

A total of 2706 pregnant women were enrolled; their mean age was 27 years. Eighty-five percent of study women had completed at least a secondary educational level; 68% were housewives. Detailed characteristics of the women are described in Table 1.

Among 2706 invited men, 593 (22%) presented at the venues, and almost all (99.6%) consented to be tested. The mean age of tested men was 36 years; 37% had at least a secondary educational level and 69% were unemployed (no stable work with a regular salary). Most men (93%) were married or cohabiting with the mother of the unborn child; 19% reported having had sex at least once with a casual partner during the three last months; only 40% of these men had used a condom during this episode. Detailed characteristics of the men are described in Table 2.

Individual VCT and couple participations

A total of 591 of the 2706 (21.8%) men reported to one of the three study venues and underwent HIV testing; 26.5% (236/891) were randomized to the bar venue; 20.9% (189/906) to the church venue and 18.3% (166/909) to the health centre venue. Male participation was 1.61 and 1.18 times more likely to occur at the bar venue (95% confidence interval [CI]: 1.28–2.01 [P < 0.001]), and the church venue (95% CI: 0.93–1.48 [P = 0.163]), respectively, compared with the health centre venue (Table 3).

All 591 men tested were invited for the couple counselling with their female partners. Only 208 (35%) men presented for couple counselling (95% CI: 31–39%). The attendance at the church venue was 45.5%, over three times (OR = 3.36 [95% CI: 2.08–5.42], P < 0.001) greater than at the health centre venue (19.9%). Couple counselling at the bar venue was 37.7%, more than double (OR = 2.44 [95% CI: 1.53–3.87], P < 0.001) the rate at the health centre venue (Table 3).

Fifty (1.8%) of the 2706 women and 13 (2.2%) of the 591 men were found to be HIV-positive. Among HIV-positive men, eight were tested at the health centre-based VCT, two at the church-based VCT and three at the bar-based VCT. A total of 571 (96.6%) couples were concordant negative, two (0.3%) concordant positive and 20 (3.3%) discordant.

Determinants for participation

Male participation

Innovative strategies are still needed to increase male participation in VCT, as male involvement is a determinant for PMTCT uptake. Two studies conducted in Burkina Faso and Rwanda found an association between male participation and the acceptance of VCT during ANC attendance by pregnant women. ^3,17 PMTCT male participation has a positive impact on the acceptability of this intervention. ^18–22

Despite many efforts to improve male involvement in VCT during ANC visits, there has been little progress in this domain. ^9,23 In our study, we found a significantly higher male participation in VCT at the bar venue compared with the health centre, but did not find a significant difference between the church-based venue and the health centre. A randomized trial conducted in Zambia found that the male participation rate in VCT in a local clinic was lower compared with VCT provided elswhere. ²⁴ Studies from sub-Saharan Africa have found that male participation levels in hospital settings vary between 12.5% and 18.7%. ^9,19,25 A study in Uganda found that non-health service VCT strategies reached a greater number of clients in general. ²⁶ The World Health Organization (WHO) ²⁷ recommends VCT for clients and their partners to be done in health facilities, but also to explore if other places would be acceptable and accessible.

Before our study, male participation during PMTCT activities at Kingasani Maternity was <2%. During the study period male participation in the health centre increased to 18%. This was probably due to: the fact that the men were invited between 15:00 and 20:00, and by the use of the letter of invitation. In the DRC, ANC activities have traditionally been scheduled in the morning. This scheduling has had a negative impact on the willingness of men to participate. Employed men have difficulties accompanying their wives to the ANC during work time. ^12,13,19,28 Revision of the ANC schedule in order to get more men involved in ANC services would be a good starting point.

Even with the increased male participation at the health centre in our study, participation rates still remained low. The quality of health services will determine whether services (including VCT) are used. If waiting times at the clinics are long, male partners may refuse to accompany their wives to the ANC. ²⁸ Capacity reinforcement and salary improvement for health service providers is necessary to improve the quality of health services. ^28–30

Couple counselling

Couple counselling is effective in reducing HIV transmission among stable couples. ^31,32

We considered male participation in the couple counselling session as a sign of their being engaged in the health of their family. During couple counselling, partners were asked to disclose their serostatus. VCT of partners of pregnant women is not only for their own advantage but also to the advantage of their families. It will lead to more support for women to increase uptake of antiretroviral therapy, family planning and nutritional advice. ³²

The high participation in the church-based VCT may be due to the fact that the church is among the places where couples are accustomed to go regularly on a weekly basis, without social, economic or cultural constraints. Most participants (70%) in our study were members of a church. The participation rate in our study was higher than in studies conducted in Rwanda and Zambia where the couple counselling participation rates were 26.9% and 9.6%, respectively. ³³ Our study suggests that churches and bar settings could be effective alternative sites for scaling-up couple counselling. ^31,32

The proportion of men who tested HIV-positive in churches (1.1%) and bars (1.3%) were close to the 1.3% HIV prevalence in the general population (results from the Demographic Health Survey conducted in the DRC in 2007). ³⁴ The higher proportion of HIV-positive men tested at the health centre venue suggests that the health facility VCT strategy may result in identifying a larger number of HIV-infected people. In Uganda, hospital-based VCT identified a greater number of people living with HIV including those with lower CD4 counts, than with other strategies. ²⁶

Determinants for participation

In our study and also in studies in Rwanda, Zambia and Cameroon, older women and men were more likely to participate in VCT and couple counselling. ^8,33 In a study in Burkina Faso cohabitation was also more likely to be associated with male VCT participation. ³⁵ The reasons for the difference in participation rate of persons affiliated with the Roman Catholic Church when compared with those affiliated with independents churches are unknown.

A weakness of our study is the lack of data concerning the characteristics of men who did not come for VCT. Interviews with the women could not yield sufficient information about their partners. Thus data of these absent men were not included in the analysis. A second limitation was the lack of data concerning the follow-up of the participants. The study was limited to five months. Follow-up data beyond this period were not collected, e.g. data related to childbirth and PMTCT prophylaxis uptake. We did not explore how the information that male partners received during the VCT session might have promoted men to become more fully engaged in the health of their family. This issue merits additional research.

We were not able to collect data concerning the number of the men in each arm who actually received the letter of invitation from their female partner. Although we used sealed envelopes, some women may have opened the envelopes and biased the uptake of the counselling at the different venues by not giving the letter to their male partner. We are not sure that the VCT venue mentioned in the invitation could have influenced women to give or not to give the invitation to their partner, and thus are not sure of the extent of this bias.