Implementation of couples’ voluntary HIV counseling and testing services in Durban, South Africa

The Canadian Sub-Saharan Africa HIV/AIDS Network sponsored the study, which was a collaboration between RZHRG and the HIV Pathogenesis Programme (HPP) in the Doris Duke Medical Research Institute (DDMRI) at the University of KwaZulu-Natal’s Nelson R. Mandela School of Medicine Campus. Five hospital-based clinics in Durban were selected for the pilot, and selection was based on large catchment population and proximity to HPP. All five clinics are located in the biggest township in Kwa Zulu Natal Province called Umlazi. Umlazi is 4,481 hectares in size and has a population estimated at 550,000 m to over 1 million with approximately 30 % of people living in informal housing [21]. It is divided into 26 community sections using alphabets from A to Z with additions of AA, BB, CC. HPP and Gateway clinic are both located on the grounds of Prince Mshiyeni Memorial Hospital (PMMH). Whereas HPP is a research center that follows up participants recruited in a variety of research studies, Gateway provides outpatient care services to people in the catchment of PMMH and mainly the V-section of Umlazi. D, H and V clinics are all small clinics within a 20 km radius of HPP and all provide outpatient care services to the respective communities in D, H and V sections of Umlazi. Except for HPP, all the other clinics are managed by nurses and counselors and have a visiting doctor who goes to the clinic to provide HIV care and treatment. HPP is has research staff that include a physician, nurses, counselors and administrative staff. All clinics are of average size (8–10 rooms used for services and waiting areas) with the exception of V-clinic which is smaller (about 4 rooms). A more detailed description of the clinics is provided in Table 1.

In February 2013, three RZHRG counselor trainers, each with over ten years of experience in providing CVCT and CVCT trainings, conducted didactic and practical CVCT training sessions for groups of local clinic counselors. These five day trainings were developed with US CDC [8] and adapted for site-specific use through a CDC Centers of Excellence grant. Whenever training is provided in a new setting, the CDC materials are translated and adaptation as needed to the local circumstances. Didactic trainings include modules on understanding HIV discordance; couple counseling skills; the CVCT intervention; data collection; laboratory guidelines; providing concordant negative results, concordant positive results, and discordant results; and providing counseling on support and prevention services. A portion of the last day of the training was used to discuss the trainees’ expectations and perceptions of the training curriculum and to address any concerns trainees may have regarding working in the field. All counselors were given pre- and post-tests assessments prior to and immediately following the didactic training, with a score of 75 % at posttest required in order to proceed with practicum training. RZHRG counselor trainers supervised CVCT counselor trainee practicums, which were conducted the week after the didactic training. Practicums required each counselor to observe at least two counseling sessions led by the RZHRG trainers and execute at least two supervised counseling sessions for each possible couple-level HIV serostatus result (concordant negative, concordant positive, discordant).

Simultaneous with the counselor trainings, two RZHRG promotions trainers with extensive experience promoting CVCT among communities in Zambia conducted a three day didactic training for CVCT promotions for local volunteers from the community. These volunteers were identified and screened based on previous experience working in health promotional activities. Didactic trainings include modules on an introduction to CVCT and the promotion of CVCT with delivery of invitations, the role of promoters, promotional strategies and promoter webs of influence, and CVCT promotional skills including public speaking and communication skills. Promoters were also trained in data collection during invitation delivery. A portion of the last day of the training was used to discuss the trainees’ perceptions of the training curriculum and to address any concerns they may have in working in the field. All community health volunteers were given pre- and post-training assessments. Scores of 65 % were required to be certified as competent. This threshold score for qualification for promoters is less stringent than that required for counselors because the promoter’s role is to provide basic information about the importance of HIV testing that encourages couples to access CVCT. Generally they have a lower level of education than counselors or nurses. The counselor’s role is more technical and the accepted score for certification is higher to ensure that only counselors that are likely to provide good quality service are allowed to practice CVCT. RZHRG promotions trainers then observed field practicums as trainee promoters began the process of inviting couples to attend pilot weekend CVCT services. Each promoter invite card was coded to allow for tracking and performance-based reimbursement. Promoters used one-to-one, door-to-door, group presentation, and clinic health talks to advocate for CVCT and encourage couples to use the service. They gave out invitations mostly in the communities they lived which are the catchment areas of the five clinics; however they were not restricted to these areas and were free to invite from other areas of Umlazi.

Client-level health indicators were collected for tested couples as a component of routine CVCT service operation including: age, gender, years cohabiting, pregnancy status, previous HIV testing, previous CVCT, ART use, neighborhood, and HIV serostatus of both partners. The HIV testing algorithm was based on the South African algorithm using two rapid HIV test kits (Alere Determine HIV-1/2 and Uni-Gold Recombigen HIV-1/2) run in parallel for each individual. The RZHRG team has developed data-collection tools to systematically and accurately collect these indicators. These tools have been used extensively by government clinic staff in Zambia and were easily adapted for use in this pilot. Collection of this data using electronic databases already established in RZHRG’s Zambia and Rwanda programs were used to record client indicators throughout the CVCT pilot. Counselors were trained on collecting data and quality control of data in the field for all data collection instruments including CVCT monitoring and evaluation log books, aggregate data sheets, and promoter information log books. Data entry into Microsoft Access databases, data quality control, and data cleaning was performed by a data manager in the Durban office after training by the RZHRG data manager.

Over twelve weekends (Saturday and Sunday) from March to June 2013, CVCT services were offered in the five pilot clinics from 8 am to 5 pm. Couples presented at the clinic with an invitation from a promoter and either attended a group session i.e. if more than one couple was available or went straight into pre-test counseling. They were then tested for HIV and given results and posttest counseled together. RZHRG trainers were available to support the clinics during this period in all aspects of the program (i.e., counseling, promotion and data management). The timeline of implementation events is shown in Table 2.

Client-level indicators (age, age-disparity between partners, cohabitation status and duration, pregnancy status, previous HIV testing, previous CVCT, ART status, neighborhood, and test site) were analyzed with descriptive statistics (means and standard deviations (SD) for continuous variables; counts and frequencies for categorical variables). All variables were stratified by couple HIV serostatus results (concordant positive, discordant man positive, discordant woman positive, and concordant negative) and differences in the distribution of client-level indicators by HIV serostatus categories were evaluated with Chi-square (or Fisher’s exact) tests or ANOVA, as appropriate. We also explored previous testing and ART use among women by pregnancy status. Analyses were conducted with SAS v9.4 (Cary, NC) and all p-values are two tailed.

The study was approved by the Biomedical Research Ethics Committee of the University of KwaZulu-Natal. Written informed consent was not required from couples obtaining CVCT services. Couples provided oral consent to be tested as a couple after the group counselling and individual pre-test counseling sessions.

Trained recruiters recruited couples throughout the whole of Umlazi and referred them to one of the five clinics. Recruiters were not limited to a catchment area, though focused on recruiting from homes, markets, and taxi ranks around the clinics.

Twenty-one individuals (18 females and 3 males) that completed the training, provided CVCT to the recruited couples in the five clinics. The majority were counselors and there were 2 nurses and 1 registered midwife. All had at least matriculation level of education (i.e., completion of the final year of high school and all qualifications necessary for university entrance) and their median years of experience was 6 years (range from 1–10 years).

Of the 907 couples (1,814 individuals) that underwent CVCT, the prevalence of HIV was 41.8 % and the prevalence of HIV serodiscordance was 29.5 % (19.3 % M-F+, 10.3 % M + F-). 245 (27 %) of couples were concordant positive and 56.6 % of all couples had at least one partner who was HIV positive. Of all HIV positive persons, 35.4 % had HIV negative partners. The pilot identified at least 313 new HIV positive individuals, and possibly more if previous tests were HIV negative (we did not ask clients their HIV results from previous testing).

Men were on average 30.1 (SD = 8.4) years of age, women were on average 28.4 (SD = 7.9) years of age, and most participants were aged between 25–34 years. Concordant positive couples were significantly older on average than concordant negative couples; discordant M + F- couples were also significantly older on average than concordant negative couples (p < 0.001). The mean age difference between partners was 3.7 (SD = 3.2) years, and age disparity was significantly higher in concordant negative couples (mean 4.4 (SD = 3.9) versus other couples).

Most (67 %) couples were not cohabiting; those that were had been cohabiting for 4.5 (SD = 5.2) years on average. Concordant negative couples were the least likely to be cohabiting while M + F- couples had been cohabiting the longest. Pregnancy status did not differ by serostatus, and 11 % of all couples were pregnant at the time of testing. Previous individual HIV testing was low (50 % for men versus 63 % for women, p < 0.001). Only 4 % of couples reported previous CVCT with no difference by couple serostatus.

Among the N = 420 HIV positive women, 10 % were pregnant and 62 % had previously tested for HIV. Among HIV positive women, previous HIV testing was not significantly more common among pregnant versus non-pregnant women (70 % versus 61 %, p = 0.27). Of positive women who had been previously tested for HIV, pregnant women were not significantly more likely to be on ART (93 % versus 89 %, p = 0.52).