Late initiation of antiretroviral treatment following diagnosis is contributing to the continuing high death rate among people who present with low CD4 counts in eight sub-Saharan African countries.
Data from South Africa show that more than half of people eligible for antiretroviral treatment (ART) at diagnosis waited at least 12 months to begin treatment, while data from eight African countries show that patients in rural areas, those identified with HIV outside a PMTCT programme and those who are not followed up when they miss appointments were most likely to start treatment at a very low CD4 count.
Earlier enrolment and better management leading to improved retention can play an important role in the linkage of diagnosis to antiretroviral treatment (ART) and care to improve mortality rates reported Ingrid Bassett and Denis Nash in two different studies presented at the Fifth International AIDS Society Conference on Pathogenesis, Treatment and Prevention in Cape Town.
In spite of World Health Organization guidelines recommending treatment initiation before the CD4 count falls below 200 cells/mm3, many continue to access care with low CD4 counts despite having been diagnosed some time before. Little is known about what happens to a patient after diagnosis and how programme factors can positively affect enrolment and retention in treatment and care.
Speaking at a conference symposium on when to start treatment, Dr Francois Venter, President of the South African HIV Clinicians’ Society told delegates that South Africa is doing very badly at retaining patients in care after diagnosis and starting people on treatment before they become seriously ill.
Data presented today from Durban show the extent of delays in initiating treatment and the extent to which patients are being lost to follow-up despite extensive uptake of HIV testing.
In a prospective study Ingrid Bassett and colleagues enrolled patients prior to HIV testing at two semi-private, partially government-subsidised, outpatient clinics in Durban, South Africa. McCloud serves a predominantly urban population and St. Mary’s a semi-rural one. Both offer comprehensive HIV care. Patients pay a fee for care.
The minimum observation time for patients included in the study was six months. CD4 counts ≤200cells/mm³ or an AIDS-defining condition determined antiretroviral eligibility. Patients were enrolled from November 2006 until October 2008 with follow-up until the end of June 2009.
Of the 3,401 patients screened 2,797 (82%) were tested. Of the 1,467 (52%) who tested positive 605 (41%) underwent CD4 count testing within 90 days of diagnosis. Of the 368 (61%) with a CD4 count
Fifty per cent of those eligible waited 3.5 months or longer before starting ART. Fifteen per cent (216/1,467) of HIV-positive patients died during the study period and 21% (76/368) of the ART-eligible cohort died before starting ART. In spite of outreach 30% of pre-ART patients were not reached.
Limitations noted by the authors include the fact that sites may not be representative of public sector hospitals in South Africa.
The problems of late treatment initation is not confined to South Africa.
In a second study Denis Nash and colleagues reviewed PEPFAR data on median CD4 count at the start of antiretroviral treatment for the period January 2005 until October 2008 for a total of 1,690 cohorts (121,154 patients) at 267 sites in eight sub-Saharan countries (Ethiopia, Kenya, Lesotho, Mozambique, Nigeria, Rwanda, South Africa and Tanzania).
Cohorts represented patients who had begun antiretroviral treatment at a particular site every quarter.
Cohorts with median CD4 counts
While the median CD4 count was 135 cells/mm³, it increased from 115 to 140cells/mm³ during the period under review. Taking into account calendar time, cohort size and country, programme level factors associated with low CD4 count included:
- Not being affiliated with a PMTCT programme resulting in a 4.4 higher risk of a low CD4 count at treatment initiation
- The programme being in a urban area.
Knowledge of HIV was not associated with having a low CD4 count when starting ART. However, knowledge of HIV and believing that having one sexual partner reduced risk let to a false sense of security and increased risks of late ART initiation.
Limitations noted by the authors include the fact that data on programme-level factors only indicate the availability of services and not the quality nor coverage.