19 February 2021
Last week the results of a SARS-CoV-2 antibody survey conducted by the South African National Blood Service (SANBS) were published. But the results were only for four provinces. It turns out that bureaucratic delays by the South African Health Products Regulatory Authority, the National Health Laboratory Service or Roche (or some combination of these three organisations), prevented SANBS from obtaining results from the other five provinces. Illustration: Lisa Nelson
A deadlock over regulatory approval of an antibody testing machine has cost South Africa crucial information about the spread of Covid-19.
On 12 February the results were published of the South African National Blood Service (SANBS) seroprevalence survey, which tested the blood of nearly 5,000 donors from four provinces: Eastern Cape, Northern Cape, Free State and KwaZulu-Natal.
In early December, we reported that a number of seroprevalence surveys were planned, including the SANBS survey, which had been announced in July 2020, but had not yet started. As reported at that time, the Western Cape Blood Service was a partner to SANBS.
But SANBS is a national blood service — so what happened to testing the Western Cape, Gauteng, North West, Mpumalanga, and Limpopo?
In short, the South African Health Products Regulatory Authority (SAHPRA) has not approved the use of a testing assay on a particular model of machine that is used to process blood from those provinces. This despite the test being submitted for consideration in August.
Roche’s Elecsys Anti-SARS-CoV-2 assay was used on Roche’s cobas e411 platform in the four provinces which produced results. In the five other provinces, the same Elecsys assay would be run on the Roche cobas e801 machine. SAHPRA has approved the assay for the e411, but not the e801.
To put it simply: This is exactly the same assay for two very similar machines. The one machine is simply bigger and faster than the other, and it is for this machine that the assay has not yet received approval.
According to experts familiar with these machines, the reagents (chemicals) used for the two machines are identical but simply have different catalogue numbers.
A survey of international medical literature concerning the use of the e801 machine for SARS-CoV-2 antibody detection suggests that it is highly effective.
According to SAHPRA, “The test is currently with the National Health Laboratory Service (NHLS) for in-country validation.”
Only once the NHLS completes its validations, can SAHPRA review the performance of the test and issue final recommendations. SAHPRA also referred us to the NHLS regarding questions about how long this validation has been going on, and when it might be expected to be completed.
In response to GroundUp’s questions, the NHLS told GroundUp that, “the validation process has not really been started”.
“The NHLS wishes to state on record that the test in question was not sent to the NHLS through the correct channels. The company stated that they were going to organise the validation through a sister laboratory in KwaZulu-Natal and then did not come back to the NHLS with a request to oversee the process.
“For this reason, the validation was not performed in line with the SAHPRA requirements and the NHLS has been trying to liaise with the company in question to expedite the process … The NHLS is currently discussing the way forward with the company to see how we, as the NHLS, can assist them to expedite and complete this process.
“Again, the company has not been forthcoming with the requirements for validation. Concerning validations, the NHLS has built a solid record when it comes to expediting validations. However, we cannot validate a test without kits, analysers, and controls.”
But Roche disputes this. In response to GroundUp, the company said: “Roche’s antibody test received conditional approval under Section 21 regulation (emergency use authorisation) from SAHPRA in August 2020. As part of this process, supplied products have to be evaluated by the national reference laboratory for performance in-country. Roche responded fully and timeously with these requirements.”
When SARS-CoV-2 antibody tests were first announced, there was a justifiable hesitancy for approving such tests from SAHPRA, particularly as they could mislead individuals about their exposure to Covid-19. However, the use of antibody tests for epidemiological purposes is quite different to using them for diagnostic purposes – it is one of the only ways to understand at a population level the true spread of the virus. Without such data, we cannot effectively plan vaccine distribution, resource allocation, and other public health decisions.
Whether fault lies with NHLS dragging their feet, Roche’s lack of compliance with NHLS requirements, or SAHPRA failing to expedite this approval, the end result is that we know less than we should about the spread of the pandemic in South Africa.