ABSTRACT
Background: Adolescent girls and young women are the epicenter of the global HIV epidemic and in need of multilevel interventions to improve their health outcomes. Method(s): FANMI, a randomized-controlled trial, evaluated the effectiveness of community-based cohort HIV care versus standard of care (SOC) among adolescent and young adults living with HIV (AYALH) in Haiti. Females, 16-24 years who were newly diagnosed with HIV at clinic or community HIV testing sites, or defaulted >6 months from care, were randomized 1:1 to FANMI vs SOC. FANMI was designed to improve convenience, social support and stigma by grouping AYALH in cohorts of 6-10 peers to attend monthly HIV care sessions in a community center with integrated clinical care, group counseling, and social activities led by the same provider. National guideline changes during the study included switching participants to dolutegravir regimens and expanding SOC visits to 6 months. The primary outcome was 12-month retention defined as any visit 9-15 months from enrollment. Secondary outcomes included viral suppression (< 1000 copies/ml), risk behaviors, and acceptability using interviews. Result(s): 120 AYALH enrolled (60 per arm) between May 2018-January 2021. Median age was 21, 91% were newly diagnosed, and median CD4 count was 591 cells/mm3 (IQR 399-788). A total of 78.3% (47/60) FANMI participants vs 85.0% (51/60) in SOC achieved the primary outcome (unadjusted RR=0.92 95%CI 0.78-1.09, p=0.35). Excluding 9 participants who never attended a FANMI/SOC visit after enrollment, 12-month retention was 88.7% (47/53) in FANMI vs 87.9% (51/58) in SOC (RR =1.01 95%CI 0.88-1.15, p=0.90). Participants who presented for HIV testing vs community testing and achieved the primary outcome: 95% vs 70% (FANMI) and 83% vs 88% (SOC). Viral suppression among those retained at 12 months: 44.6% (21/47) in FANMI and 37.3% (19/51) in SOC (RR 1.20 95% CI 0.74-1.9, p=0.45). There were no differences in pregnancy and risk behaviors. Providers preferred FANMI reporting increased time for counseling and peer support. FANMI participants reported high acceptability, decreased stigma, and increased social support with no confidentiality breaches. Limitations included interrupted study operations during the COVID-19 pandemic. Conclusion(s): FANMI was not more effective for AYALH in Haiti but was preferred by providers and highly acceptable to participants. It offers promise as a complementary program for high-risk AYALH in low-income settings facing barriers to clinic-based care.
ABSTRACT
Introduction: Stress is associated with multiple adverse health outcomes, including hypertension. The impact of stress on health may be moderated by social support. The distribution of stress, support, and their association with hypertension have not been well described in low-income countries that face severe poverty. Over the past decade, Haiti has suffered massive natural disasters including the 2010 earthquake, cholera outbreak, COVID-19 pandemic, and recurrent civil instability-all of which may act as prolonged stressors.Hypothesis: We assessed the hypotheses that 1) there are high levels of both stress and support in Haiti, and 2) high support would moderate the relationship between high stress and hypertension. Method(s): We measured stress and social support using validated instruments in a population-representative cohort of adults living in urban Port-au-Prince, Haiti between March 2019 and April 2021. Stress was measured using the Perceived Stress Scale, while social support was measured using the Multidimensional Scale of Perceived Social Support. For visualization, continuous scores were categorized using equal-width groups (stress: low (1-5), moderate (6-10), high (11-16);support: low (7-21), low-moderate (22-35), moderate (36-49), moderate-high (50-64), high (65-77)). Linear regression models were used to quantify the associations between: 1) stress and support adjusting for age and sex, 2) blood pressure and stress adjusting for age and sex. A formal moderation analysis was conducted to assess if support moderated the relationship between stress and blood pressure. Result(s): Among 2,817 adults, 59.7% female and the median age was 40 years (IQR 28-55). The majority had an income of less than 1 US dollar a day (69.7%). The median stress score was 8 out of 16 points, and median support score was 61 out of 77 points. Stress was higher with older ages (60+ years versus 18-29 years: +0.79 points, 95% CI 0.51 to 1.08) and in females (+0.85 points, 95% CI +0.65 to +1.06). Support was higher in males (+3.29 points, 95% CI 2.19 to 4.39). Support was inversely associated with stress, adjusting for age and sex (-0.04 points, 95% CI -0.04 to -0.03). Stress was not associated with systolic or diastolic blood pressure after adjustment for age and sex. Support did not moderate the association between stress and blood pressure. Conclusion(s): In this urban cohort of Haitian adults living with chronic civil instability, stress was moderate and support was high. While support was associated with lower stress, it did not moderate the relationship between stress and blood pressure. Despite the high levels of instability in Haiti, participants displayed resiliency through high levels of support, which may be an underutilized resource in reducing stress and long-term negative health outcomes.
ABSTRACT
Background: Challenges to retain patients with HIV in Haiti were worsened by civil unrest and the COVID-19 pandemic. To support patient retention, GHESKIO, one of the largest HIV care centers in the Caribbean, set up 11 community distribution points (CDPs) for antiretroviral therapy (ART) pickup and viral load testing at satellite sites in Port-au-Prince neighborhoods, and offered home delivery to patients. Methods: The choice to pick up ART at CDPs was offered to all patients by 5/2019. Nurses at CDPs referred patients to GHESKIO clinics if they were symptomatic or due for physician visit. Data on all ART pickups in 5/1/2019- 10/23/2020 from GHESKIO's electronic health records were described. Multivariable logistic regressions were used to identify patient characteristics associated with having ≥1 non-clinic visit (i.e. at CDP or home). Results: 16,234 patients completed ≥1 drug pickup visits during the study period (41.2% male;mean±SD age 41.8±13.3 years;14% newly initiated ART since 5/2019;6.0±4.1 years since ART enrollment as of 5/2019 among previously enrolled patients). 39.3% of patients had ≥1 non-clinic pickup (31.8% had ≥1 CDP visit, 12.7% had ≥1 home visit). Patients attended 77,514 visits (4.8±2.2 per patient), 16.2% and 3.3% of which were CDP and home visits, respectively. Since 9/2019, when nationwide lockdown due to political unrest began, 21.9% of visits were at CDPs and 3.9% at home. After 3/2020, when the first COVID-19 case was detected in Haiti, 15,183 patients completed 35,564 visits (2.3±1.2 visits per patient);proportions of visits at CDPs and home increased to 27.5% and 4.7%, respectively. Of patients with visits since 3/2020, 2,824 (18.6%) patients relied solely on non-clinic ART pickups (13.7% only at CDPs;3.3% only at home). Regression suggests male sex, higher education, higher income, age <18 years, longer time since ART initiation, and non-single civil status were associated with having ≥1 non-clinic visit. Patients living in Carrefour, a neighborhood blocked from GHESKIO clinics during civil unrest, were more likely to have ≥1 non-clinic visit than patients from other neighborhoods. Conclusion:Community distribution of ART builds resilience in health systems and supports continuity of care when access to clinics is limited. These services may be especially preferred by younger patients with longer time since ART initiation, higher income and education, and living in areas with limited access to medical clinics.
ABSTRACT
Background: Delays in ART initiation for TB testing are associated with high rates of loss to follow-up. There are limited data on outcomes with same-day testing and treatment for patients with TB symptoms at HIV diagnosis. Methods: We conducted a randomized trial comparing same-day and rapid (7 days) TB testing and treatment initiation among adult patients with TB symptoms at HIV diagnosis at GHESKIO in Haiti. The same-day group received Xpert Ultra results and initiated either TB medication or ART on the day of HIV diagnosis. The rapid group received Ultra results within the first week and started ART on Day 7 if not diagnosed with TB. Dolutegravir (DTG) replaced efavirenz (EFV) as the first-line anchor drug in December 2018. The primary outcome was 48-week HIV-1 RNA <200 copies/mL. Results: Between November 2017 and December 2019, 500 participants were randomized to rapid (n=250) or same-day treatment (n=250) (Table 1). 234 (46.8%) were female, median age was 37 (IQR: 30, 45), and median CD4 count was 278 (134, 421). In the rapid group, 40/41 (97.6%) participants diagnosed with TB started TB drugs;244 (97.6%) started ART. In the same-day group, 45/45 (100%) diagnosed with TB started TB drugs;250 (99.6%) started ART. There were no statistically significant differences in 48-week outcomes between groups. In the rapid group, 224/250 (89.6%) were retained in care, and of these, 171 (76.3%) had HIV-1 RNA <200 copies/mL. In the same-day group, 219/250 (87.6%) were retained in care, and of these, 155 (70.8%) had HIV-1 RNA <200 copies/mL. The primary outcome (48-week HIV-1 RNA <200 copies/mL) was achieved by 171/250 (68.4%) in the rapid group and 155/250 (62.0%) in the same-day group (p=0.133). Outcomes were superior among participants who initiated ART with DTG instead of EFV, with HIV-1 RNA <200 copies/mL in 82.4% vs. 68.1%, respectively, (p=0.001) among those receiving viral load testing, and 75.3% vs. 60.4% among those randomized (p<0.001). Conclusion: Among patients with TB symptoms at HIV diagnosis, both rapid and same-day treatment are associated with near-universal initiation of TB treatment and ART, with no significant difference in 48-week outcomes. Viral suppression rates were lower than anticipated, which we attribute to high rates of transmitted EFV resistance, political instability with a national lockdown, and the SARS-CoV-2 outbreak in Haiti during the study period. Viral suppression rates are superior with DTG, supporting the rapid transition from EFV to DTGbased ART.