For the first time to our knowledge, this cohort reports a snapshot on the clinical, immunological and virological outcomes of APHIV measured over 24-month in relation to HIV-disclosure in two West African pilot sites contributing to the IeDEA West Africa collaboration. We made several key findings: first, the COHADO project has encouraged the full HIV-serostatus disclosure to APHIV, with an increase from 46.1% to 74.2% after 24 months, but we still found that one in four APHIV remained not formally disclosed of his/her own HIV-serostatus at the endpoint while they all were aged above 13 years. Second, unlike the caregivers and the psychologists, doctors have very little involvement in the disclosure process. Third, we found that after 24-month of follow-up, the cumulative death rate was high, close to 3%, and only 45% of APHIV had a favorable 24-month combined outcome after two years of follow-up. Finally, in adjusted analysis, a favorable outcome was significantly associated with the site: APHIV from Lomé significantly increased the odds of a favorable 24-month outcome compared to those from Abidjan, but no significant association with HIV-disclosure.
Although the frequency of ALHIV fully disclosed of their HIV-serostatus varies according to studies in Africa, it remains overall low ranging from 16% to 39% [36–41]. In our study, we found that only 46.1% knew their HIV status at baseline, and 74.2% by 24 months of follow-up. Although this should be closer to 100% according to the WHO recommendations, this is higher than reported in previous studies, particularly for West Africa [25, 31, 42]. In Ghana, two separate studies reported the proportion of HIV disclosure to be 11.2% among children and adolescents aged 8-14 years in 2009 and 44% among ALHIV aged 12 -19 years in 2015 [25, 43]. Furthermore, we observed an increase in the HIV-disclosure process over the 24-month follow-up period. This is clearly visible in Lome where the proportion of APHIV disclosed had increased from 24.8% to 77.2%. One of our hypothesis is that this likely occurred after a 3-day training workshop on HIV-serostatus disclosure to ALHIV delivered in 2016, in Abidjan and involving the HIV health professionals from all the IeDEA-pWADA sites [44]. This workshop may have changed the health care workers’ perceptions and practices regarding HIV-disclosure and explain the significant increase of HIV-disclosure rate at the endpoint, but we also acknowledge that this is can be also only correlated to the adolescent age.
As a matter of fact, we found that few doctors are involved in the process of HIV-disclosure, and they tended to delegate the disclosure practice to others for several reasons including over clinical work load, or fear in doing it. Indeed, healthcare workers face structural issues including limited human and technical resources, whereas disclosure is a complex process which needs time and training [21, 28]. In 2018, there were too few counselors working in the current HIV-programs. Thus, when existing in the HIV-programs, psychologists are better trained to address mental issues such as HIV-related stigma and taboo, explaining their good in the process. But they became the only person being in charge in the full HIV-disclosure process and they are not always daily available. Therefore, we feel that the whole staff should be involved in the HIV-disclosure with a multidisciplinary approach and task-sharing. That approach should be set-up at each HIV program level to offer a comprehensive care, including the process of HIV status disclosure. It is important for national and regional programs, to tailor locally appropriate strategies to improve disclosure practice, such as the training of multidisciplinary team on disclosure mentioned earlier [34]. Furthermore, caregivers are also unprepared for HIV-disclosure and fear stigmatization. For many parents, their children are too young or are not ready to receive HIV-disclosure [26, 29, 45]. The role and benefits of having caregivers involved in the HIV-disclosure process remain unclear. While some studies have suggested that caregivers are in a much better position to disclose ALHIV serostatus, others have reported that according to ALHIV, health workers are better placed [46–48]. Nevertheless, a better understanding of what refrain caregivers to disclose their HIV-status to their child is important to support them accordingly.
We also found that APHIV HIV-disclosed to tended to have worse clinical and immunological conditions than those not disclosed. Unfortunately, we were not able to assess the exact timing of HIV disclosure as this is an evolving process that can take several clinical visits. We make however, two main hypotheses that could explain this observation. Either HIV-disclosure had a negative effect on the APHIV leading to poor ART adherence and thus poor clinical outcomes. Or, it is the poor immunological and clinical conditions of the patient that prompt their HIV-disclosure to improve urgently their health outcome via the improvement of their adherence to ART. We feel, in regards to the context of our study and factors correlated to HIV-disclosure, that this second hypothesis is most likely. This indication bias could have reduced the HIV-disclosure effect measured after 24 months, since APHIV disclosed have worse clinical and immunological conditions. Similarly, we noted the cohort effect: those disclosed to were significantly older, and therefore were most likely to be at a more advanced stage of the disease progression. Previous cohort studies, in both high and low income settings have reported that younger age of APHIV is associated with less frequent WHO 4 clinical stage compared to older APHIV [49, 50].
During follow-up, the cumulative death rate reached 3%. While this seems high, it is lower than that reported in other studies conducted in ALHIV in sub-Saharan Africa, where it ranged from 4-6% [33, 51]. Although these rates remain in the same order of magnitude, our relative lower rate could be explained by the shorter follow-up duration.
Overall, 45% of APHIV had a favorable combined 24-month outcome that is rather sub-optimal in terms of quality of ART response. A favorable outcome was not associated to HIV-disclosure but more frequent in Lomé than in Abidjan. However, APHIV in Lomé were in poorer clinical and immunological condition compared to those from Abidjan, and therefore had more leeway to reach a favorable 24-month outcome. Although, we were not able to document different sub-types, we do not support a difference between the two sites to explain the difference. Our finding highlights rather differences of care practices and resources available between West African cities, which must be considered to improve APHIV care delivery. Nevertheless, our outcomes are observed in the context where about 80% of APHIV were receiving a NNRTI-therapy with a limited access to second-line therapy, common in West-Africa. The sub-optimal virological response in our population is in line with the results reported in a snapshot-study conducted in Lomé in 2016, with high rates of virological failure and drug resistance [16].
Many studies have reported on how HIV-disclosure improves ALHIV clinical, immunological and virological outcomes as well as retention in care [28, 32, 33, 51, 52]. In COHADO, we found that HIV-disclosure during COHADO tend to reduce the odds of having a favorable 24-month outcome, though this was not significant. This result could partially be explained by the disclosure indication bias induced in COHADO, previously mentioned, where disclosure is prompted by the advanced stage of the disease. We advise caution in the interpretation of our observation. Because the disclosure data collection was not standardized, nor we did not have the date of disclosure for all APHIV, with a substantial proportion of APHIV already HIV-disclosed before the inclusion in COHADO cohort and without HIV-disclosure date. So, the association between HIV-disclosure and the 24-month favorable outcome is evaluated at a time too soon for some APHIV and too far for others.
It has been previously reported that ALHIV not disclosed of their HIV-serostatus increases virological failure by a 5-fold [53]. The identification of virological failure is possible where viral load measurement is available, which was not routinely done in West Africa. The COHADO study is one of the first cohort to provide viral load data in West-Africa, with measurements substantially more available after 24 months of follow-up. Viral load access is a real issue in ALHIV care monitoring in Africa, and the prioritization of viral load measurement is universally recommended since 2016 particularly for ALHIV who have high risk of virological failure[53]. In the COHADO clinical sites, the annual viral load measurement is a recent opportunity that should be scaled up and supported in every HIV-care facility, as virological success could be also an indicator used to encourage treatment adherence and therefore outcomes of APHIV. At the contrary, identifying early enough virological failure would be helpful in reinforcing more closely the treatment adherence to re-suppress viral load in these vulnerable population.
Our study met several limitations. First, we enrolled less than half of APHIV who visited the sites during the inclusion period due to logistical issues, mainly related to health care workers overwork, with little time to propose the study and get formally the parent’s consent. Although those included did not differ significantly from those not included, the small sample size and relatively short follow-up may have limited the statistical power of our analysis. This flaw could have overshadowed the link between the 24-month health outcomes in APHIV, and several variables such as age, sex or other relevant variables. Second, we selected our APHIV in urban referral center for adolescent HIV care where APHIV are likely receiving better standard of care compared to the general population in these cities. Third, the few time points in the follow-up did not allow us to perform a longitudinal analysis for instance. However, the COHADO cohort reported a small proportion of lost to follow-up reflecting high quality of follow-up in the selected sites. Nevertheless, our study provides original data documenting the feasibility and indirectly the positive effect of accompanying actively the HIV disclosure process using health care worker training on this topic, and routine monitoring of this crucial event. Second, to our knowledge, this study gives a representative and longitudinal assessment of the virological response in APHIV on lifelong ART that could be further used as a reference.