Implementing viral load testing in resource-limited settings: one size does not fit all

Routine viral load can be done in low-resource settings, Helen Bygrave of Médecins Sans Frontières' Southern African Medical Unit told the 17th International Conference on Aids and Sexually Transmitted Infections in Africa (ICASA) last week, but choosing the appropriate testing platform, training clinicians, and increasing treatment literacy for patients are essential to ensuring that viral load tests are done correctly, and results acted upon.

Bygrave noted that viral load is an important component in what is referred to as the 'treatment cascade' – that is, keeping people in care, and moreover, ensuring that their viral load is undetectable. The World Health Organization (WHO) recommends that people on ART receive viral load tests once yearly. But in resource-constrained settings, viral load is not routinely available due to the cost of testing and complexities surrounding the test itself, as well as logistical issues such as sample transport and limited human resources.

Bygrave says that in low-resource settings, considering human resources, transport, lab accessibility, and type of facility are important in determining which sample type and technology to use.

Viral load testing in Malawi and Zimbabwe

She described viral load testing practices at MSF treatment facilities in Thyolo district, Malawi, and Buhera district, Zimbabwe. Both sites have implemented routine viral load testing since 2011.

In Buhera, given lack of frequent transport from clinics to the district laboratory and challenges in transporting samples to the capital laboratory using cold chain, dried blood spot (DBS) testing – using EDTA venous blood – is used. At the primary healthcare level, nurses collect blood samples. Viral load testing is first conducted at month three, and then again yearly.

In Thyolo, blood for viral load testing is sampled and shipped using dried blood spots. "When you move to dried blood spots, you do lose something on the sensitivity, but we showed that the specificity remained good," said Bygrave, who noted that a systematic review on DBS is soon expected, and that the WHO will give guidance on viral load methods early next year.

While there were concerns from the conference floor about the use of DBS, Bygrave notes that venous DBS is "well validated in many countries", and that in many resource-constrained settings, "without DBS, getting viral load done anywhere outside of central hospitals is not possible. The reality of scaling up viral load in rural settings without DBS – I simply don't see it, given the challenges with sample transport."

In Malawi, MSF is currently considering outcomes when sample preparation is shifted to lay workers. Viral load is first conducted at month six, and then again twice yearly.

In order to reduce costs, the project utilised 'viral load pooling', which includes testing five DBS samples together: if a result comes back negative (that is, below 1000 copies/mL), then it can be assumed that all the patients have a viral load below 1000 copies/mL. If it comes back positive, then – using an algorithm – all patients undergo further viral load testing. Bygrave says that this method reduced the number of viral load tests required by 30%, saving the district up to $207,000 a year. The Ministry of Health is now considering using this method nationally.

Bygrave notes that rolling out viral load testing does not happen overnight: although a concerted effort to scale-up viral load testing started in Zimbabwe in 2010, a scale-up wasn't seen until 2012. In Buhera, after 24 months of using DBS, 83% of the cohort had a viral load result. In Thyolo, after 24 months, only 27% of the cohort had a viral load result. Of those having a first routine viral load test in 2012, 12.6% in Buhera and 7.1% in Thyolo had a viral load above 1000 copies/mL. Bygrave notes that implementing viral load may negate the need for CD4 monitoring. In Zimbabwe, CD4 monitoring was stopped; in Malawi, it was never started.

Bygrave says that simply conducting a viral load test isn't enough. She observes a 'viral load cascade', in which patients are lost between getting a viral load test, receiving follow up care, getting a second viral load, and being switched to second line treatment if needed. "Viral load is not a magic bullet, and unless we supervise this algorithm, it's not going to work…We have to get results to patients, and we have to be very sure about supervision of this cascade so that we don't pour a lot of money down the drain to implement a new test that's not going to be acted on."

Of those in Zimbabwe whose viral load remained high despite enhanced adherence counselling, only a little over 50% were switched to second-line treatment. Bygrave notes that decision making about switching patient's regimens is "highly centralised" and doctor-oriented in most countries, and that there are concerns over switching patients to more expensive second-line treatment if they are considered to have poor adherence.

Viral load testing in South Africa

Suzanne Jed of I-TECH South Africa, which supports the Department of Health, presented findings from 151 patient files in two provinces, which were mined for data on ART start date, ART regimen, viral load specimen collection date, and viral load result in order to consider how many patients were receiving viral load tests as per Department of Health guidelines, and what those results found. National guidelines require viral load testing to be done six months after initiation, and once yearly thereafter.

I-TECH found that 71% of the 151 patients were eligible for viral load monitoring (meaning they had started ART at least six months ago). Of those eligible, 46% had viral load obtained as per the national guidelines; 22% did not have viral load results obtained within the past year; and 32% had no documentation of viral load.

Of those who had their viral load measured, 47% were undetectable; 26% had between 41 and 400 copies/mL; 3% had 401 to 1000 copies/mL; 19% had greater than 1000 copies/mL; and 5% had pending results, or were not documented. For those who had viral loads above 40 copies/mL, the majority had started on ART six months before, and this was their first follow-up viral load test. The majority of patients were on first-line treatment.

Jed notes that the study’s power was limited due to small sample size. She said that point-of-care viral load and the roll out of fixed-dose combinations may positively impact testing rates, and adherence.