Research question
What are the clinical and immunological outcomes of HIV infected adults and children in a rapidly expanding, decentralized, HIV Treatment and Care Programme in rural Africa; how does HIV treatment influence disease progression and outcomes at an individual level?
Fit with Africa Centre research portfolio
This research fits in Objective 1 of the Africa Centre’s current research activities – understanding the impact of ART roll-out at individual, household, clinic and community level.
Data sources used
ARTemis, the database of patients in the HIV Treatment and Care Programme, developed at the Centre for use of programme monitoring as well as clinical research. We have ethics approval to link data from ARTemis at an individual patient level with ACDIS data for people who live in the Africa Centre surveillance area. In addition we have established a prospective clinical cohort in two of the busiest clinics in the HIV Treatment and Care programme to chart more details clinical information on patients on treatment than is possible in ARTemis.
Methods/study design
This body of work is possible through the Centre’s partnership with the Department of Health in establishing and implementing a comprehensive HIV Treatment and Care Programme in the sub-district. Data on all patients initiating treatment are stored in ‘ARTemis’ (an operational database maintained at the Africa Centre), including baseline data, patient characteristics, clinical staging, subsequent 6-monthly laboratory results (CD4 counts and viral loads) and outcomes (including death and loss to follow-up). ARTemis also stores CD4 counts on individuals who have not yet met eligibility criteria for initiation onto ART (6-monthly for adults and more frequently for children depending on clinical stage/immunological status). Approximately 40% of the treatment cohort reside within the DSA, and clinical and population data are linked. In addition, we are extending our data collection in a prospective clinical cohort, collecting detailed, longitudinal, morbidity data on patients initiating treatment has been implemented at two of the busiest clinics in the DSA, providing a fairly unique, primary care, community cohort. We are undertaking a series of related analyses on patient outcomes, serious morbidity, adverse drug reactions, hospitalisation, adherence.
Findings to date - adults
Delivery of HIV treatment and care to rural communities presents unique challenges and current ART delivery models may significantly limit the accessibility of ART. To have the greatest impact on public health, HIV treatment programmes will have to be decentralised and integrated into the existing primary health care (PHC) system. Monitoring of treatment outcomes is essential to identify constraints or deficiencies in programme performance, and viral load monitoring may provide warning of adherence problems and potential development of acquired antiretroviral resistance. To evaluate the decentralised model of ART delivery to a rural community we analysed the scale-up of Hlabisa ART programme and explored trends in both the characteristics of individuals accessing treatment and the outcomes within the first year of treatment. Clinical and immunological outcomes have been examined in 5,719 adults who initiated ART between 2004 and September 2008: 67.9% were female, and median baseline CD4 count was 116 cells/mm3 , lower in males than females (91 cells/mm3 vs 128 cells/mm3 respectively, p<0.001). Overall retention in care at 12 months was relatively high at 84.0%; 10.9% died and 3.7% were lost to follow-up. Approximately 77% of active patients had a suppressed plasma viral load (<25 copies/ml) one year after treatment initiation. Mortality rate was 12.6 deaths per 100 person years, peaking in the first 3 months following ART initiation at 30.1 deaths per 100 person years, reducing thereafter to 6.2 deaths per 100 person years.
Although retention in care of those on ART is good, it is less so for those people in care but not yet treatment eligible. Of 4,223 adults, who had attended clinics in 2007 for CD4 count monitoring, and not yet eligible for ART, 83.9% were female. Overall retention, defined as a return monitoring visit within 13 months, was 44.9% with median time to return of 201 days. Male sex and higher CD4 counts were significantly associated with lower odds of retention, whilst older age was significantly associated with higher odds of retention.
HIV remains the leading cause of death amongst young adults in this population. An important cause of in-patient mortality for patients with HIV is cryptococcal meningitis. A retrospective case note study showed an in-hospital mortality rate of almost 41.3%, with only 12.2% of patients documented to be alive on ART in our programme beyond 12 months. In a further study at the provincial hospital at Empangeni, with both greater resources and improved access to the drug amphotericin B, early mortality was still high with 28-day mortality of 32.3%. 52/186 (28.0%) patients died within 14 days of diagnosis and 60/186(32.3%) had died by day 28. Overall mortality rate was 2.13 deaths per 100 person days (95% CI 1.66-2.72), with 14 day and 28 day mortatity rates of (2.54 per 100 person days, 95% C.I. 1.94-3.33) and (2.12 per 100 person years 95% C.I. 1.65-2.73) respectively. In multivariable cox regression analysis, focal neurology (aHR 11 95%C.I. 3.08-39.3,P<0.001),diastolic blood pressure <60mmHg (aHR 2.37 95%C.I. 1.11-5.04,P=0.025), concurrent treatment for tuberculosis(aHR 2.11 95%C.I. 1.02-4.35) and use of fluconazole monotherapy(aHR 3.69 95% C.I. 1.74-7.85,P<0.001) were associated with increased mortality at 14 and 28 days.
Despite the scale-up of HIV services the fragility of underlying health systems makes management of HIV-related complications very difficult in this setting. However, an audit of admissions to Hlabisa hospital has demonstrated a significant decline in the number of medical admissions subsequent to the roll-out of ART in the sub-district.
Lactic acidosis and symptomatic hyperlactaemia are amongst the most common toxicities seen with ART in sub-Saharan Africa; lactic acidosis has been reported with the use of stavudine which remains a component of first-line ART in most of sub-Saharan Africa. We investigated the virological and immunogical outcomes on ART following a diagnosis of stavudine-related lactic acidosis. There was no evidence of poorer long-term virological or immunological outcomes as a result of substitution of AZT for d4T following lactic acidosis/symptomatic hyperlactaemia. Lactic acidosis was associated with female sex, older age, higher body weight and higher baseline CD4 counts (the latter contrary to previous reports). Short-term mortality was lower than in reports from similar settings, possibly due to earlier clinic identification in our decentralised service.
Findings to date - children
HIV treatment, following South African treatment guidelines, has been offered to children since the inception of the HIV Treatment and Care Programme in June 2004. In addition to paediatric care and treatment being offered at all clinics, a family clinic was established in 2007 at the busiest Primary Health Clinic. At the start of the programme few children were commenced on ART. Relatively modest, but effective strategies to increase identification of eligible children, have resulted in increasing numbers of children commencing ART; a cumulative total of 477 by June 2008 and over 1000 by September 2009. Clinical outcomes of the first 477 children on ART show that children were relatively old at treatment initiation (median age 74 months, range 4-180) with only 58 children commencing treatment under 18 months. Most children already had advanced diseases; a quarter had tuberculosis; and a third demonstrated moderate to severe malnutrition. Mortality of children compared well to other cohorts from the continent; 32 (6.7%) children died over 732 child-years of follow-up (43.7 deaths per 1000 child-years; 95% CI 32.7-58.2), 17 (53.1%) within 90 days of treatment initiation. Children with baseline haemoglobin ≤8g/L were more likely to die (adjusted HR 4.5; 95%CI 1.6-12.3) as were those aged <18 months compared to >60 months (adjusted HR 3.2; 95%CI 1.2-9.1). Timely access to paediatric HAART for HIV-infected children is a clinical and research priority. Using facility and population-based data and a deterministic model, we estimated by the end of 2007 that half of the number of children predicted to require treatment had started HAART.
Preliminary findings examining weight gain of children on ART, in relation to timing of treatment initiation, shows that overall mean weight-for-age z-score (WAZ-score) increased after the first 6 months on treatment in all age groups. However, increase in crude mean WAZ-score was significant among children aged under 2 and 2-5 years, but not in children over 5 years, suggesting a critical window of HAART initiation.
Policy implications
The strengths of this cohort include its size, the quality and quantity of data to describe it. We have the ability to model the impact of the programme on the population due to the detailed, longitudinal information available about the community and the linkage between clinical and surveillance data. We are thus well placed to offer complex and wide ranging analyses on several aspects of ART in sub-Saharan Africa. In an era when the consequences of rapid ART roll-out are starting to be felt, and the challenges of late presentation with HIV persist, along with high and increasing rates of infection such as TB, large rural decentralised cohorts are needed to describe these challenges, and the effect of interventions to address them.