Without improved infant HIV diagnosis, early treatment recommendation may have limited impact

Early infant HIV diagnosis (EID) is becoming more frequent, according to a retrospective multi-country analysis in Cambodia, Namibia, Senegal and Uganda in 2009, researchers announced today at the Eighteenth International AIDS Conference in Vienna.

The 2010 World Health Organization (WHO) guidelines recommend all infants infected with HIV under two years of age begin antiretroviral treatment regardless of CD4 count or disease stage. Without antiretroviral treatment an estimated 50% of children with HIV will die before the age of two.

HIV DNA testing is necessary to make a definitive diagnosis in children below 18 months because of the persistence of maternal antibodies up until this age.

The use of dried blood spots (DBS) has simplified sample collection as it is a less invasive procedure for infants, and facilitates storage and transportation to laboratories equipped to carry out DNA testing using polymerase chain reaction (PCR) testing, so providing a means in resource-poor settings for early infant diagnosis.

The study reported on Tuesday was led by the respective ministries of health with technical support from UNICEF.

The researchers reviewed a selection of 18 to 25 collection sites in each country, covering the geographic range of health centres, HIV service availability and time since the start of EID services; transportation and central laboratory components were also reviewed.  A standardised questionnaire was used to look at sample volume and programme practice and key informant interviews at the national level looked at programme management and scale-up.

Future planning must take into account the effects of decentralisation of EID services, noted Matt Barnhart of UNICEF.

Out of a total of 84 EID collection sites reviewed, with over 21,000 infants tested, geographical coverage varied from 9% in Senegal to 86% in Namibia. While sample collection has increased significantly in all four countries, service use remained under 50%, ranging from 30 to 40% at most sites.

In spite of decentralisation, service use was clustered in all four countries, representing just 2%, 8%, 9% and 9% of all potential sites in Namibia, Cambodia, Senegal and Uganda, respectively. Dr. Barnhart suggested this was due, in part, to lack of training and supervision in many sites. The samples were simply not being sent for testing. The issue that needs to be addressed, he stressed, is that of organisation of services and care at the sites. Additionally, he noted the possibility that patients found certain sites more accessible and allowed for anonymity.

While all national algorithms used encourage testing at six weeks, late age at diagnosis was common. Fewer than one half of infants tested were tested in their first two months of life, Dr. Barnhart noted. He added, “So coverage of the optimal service – early testing – is even lower.”

The average ages were 5.3 months, 4 months, 4.4 months and 7.2 months in Cambodia, Senegal, Namibia and Uganda, respectively. However, some variations in age were seen in more established EID programmes that had been in operation for more than two years; in Uganda the average age in January 2008 was 7.4 months and in October 2009 6.1 months; in Namibia in January 2006 the average age at diagnosis was 6.2 months falling to 3.3 months in August 2009, indicating an improvement as the programme matured.

Of those infants who ever tested positive the numbers who were enrolled in care, remained alive and on antiretroviral treatment were low, ranging from 25 to 45%. 

These figures, Dr. Barnhart noted, underline the missed opportunities for early testing that include prevention of mother-to-child transmission (PMTCT) follow-up appointments after delivery, and scheduled vaccination visits to the clinic.

A focus on the timely use of testing services for EID leading to comprehensive treatment and care services is essential to further the success of treatment in HIV-infected infants and children, Dr.Barnhart stressed.

He also highlighted the differences in ease of scale-up which depends on pregnant women knowing their status. For example, coverage of early infant diagnosis is significantly higher in Namibia where, according to the presenters, 80 to 100% of HIV-positive pregnant women know their status, compared to Cambodia and Uganda where only an estimated 66 to 68% of women know theirs.

While these findings provide further opportunities to improve early infant diagnosis services, many improvements for HIV-exposed infants have been made: PMTCT coverage is increasing; the availability of infant testing services is growing; and – of critical importance – children do respond extremely well to antiretroviral therapy if tested early enough and so treated early, said Dr Barnhart.

He acknowledged the many challenges that are severely limiting the impact of EID: for example, late age at testing; centralisation of services; and slowly rising service coverage. However, with these findings, he added, ministries of health have already begun to plan, pilot and address the gaps to improve the survival of HIV exposed infants. 

Dr Francois Venter, in an earlier satellite session on early infant diagnosis, explained that where he works, in a busy urban setting in Johannesburg, “Over the past few years the numbers of infants getting tested has increased fivefold, yet the numbers on treatment hasn’t budged one iota.” He cautioned, “If the system is not there to support the area, it doesn’t matter how good the test is.” He suggested that the outcome, that is the numbers of children on treatment [and remaining in care] is a useful barometer for how well a public health system serves the most vulnerable and those most in need.

Under current circumstances, with relatively low utilisation of infant HIV diagnosis, adoption of the new WHO guidelines on immediate treatment of all children under two with HIV infection will not significantly increase the numbers of children with HIV found to be eligible for treatment, according to another poster presentation of research undertaken in Uganda using a real-life cohort followed since 2003.

The researchers found that only half of all children eligible, no matter the criteria used (CD4 cell count, disease stage or age), would be put on treatment. As with the four-country analysis, the crucial component, according to Martina Pennazzato and colleagues, is improving timely access to early infant diagnosis.

The researchers applied WHO guidelines of 2006, 2008 and 2010 separately to a cohort of 985 children with a median age of five years, ten months, followed since 2003.

The proportion of those eligible for antiretroviral treatment was identified, as well as the probability of starting antiretroviral treatment over time.

In accordance with the guidelines of 2006, 2008 and 2010, 40%, 57% and 66% of all children, respectively, would have been eligible for treatment at enrolment. The probability of not being on treatment two years later following these criteria would have been 24%, 16% and 12%. 

No matter the criteria used (CD4 cell count, clinical stage or age) to identify eligibility, only half of all eligible children ever started antiretroviral treatment. After 2008, age was the least used criteria to identify eligibility (OR 10.5 95% CI: 3.8 to 3.11).

Martina Pennazzato noted that, in this implementation stage of the guidelines:

• paediatric programmes linked to prevention of mother-to-child transmission (PMTCT) programmes were more likely to enrol infants and young children.

• the wide dissemination and implementation of the guidelines was critical; together with being able to counter the scepticism of some of those working in the field who, while knowing it is the right thing to do, often think twice about putting asymptomatic infants on treatment because of the challenges involved.

Future planning requires looking at how to improve current EID services as well as scaling-up to ensure infants and children are started on treatment early enough to make a difference in health outcomes.

Tripathi S. et al. Increasing uptake of HIV early infant diagnosis (EID) services in four countries (Cambodia, Namibia, Senegal and Uganda). Eighteenth International AIDS Conference, Vienna, abstract TUPD205, 2010.

Penazzato M et al. The evolution of paediatric antiretroviral treatment guidelines: what’s the impact on the ground? Eighteenth International AIDS Conference, Vienna, abstract TUPDB204, 2010.

Abstracts and a presentation for this session is available on the official conference website.