Testing for COVID-19 at scale: extending capacity for disease identification
By Nick Collier, CTO
In the five weeks since the WHO announced the name for coronavirus disease COVID-19, it has escalated into a global pandemic. The situation is evolving rapidly, putting huge pressure on healthcare and economic systems worldwide.
One area that has come under scrutiny is how and when people are tested for the disease.
Failure to scale-up testing will exacerbate the challenge
As it stands, most people with symptoms will not be tested, even if COVID-19 is suspected. Afterall, treatment decisions are driven by the symptoms not the diagnosis.
Yet, given current self-isolation guidance in the UK, people need to take seven or 14 days off work if they have a cough or high temperature, depending on their home circumstances. If they only have the common cold or seasonal flu, they could subsequently become infected with COVID-19, triggering a further period of self-isolation.
Multiply this scenario across the entire population and the financial, social and logistical repercussions for individuals, public services and business operations will be immense.
On March 17, the UK’s chief scientific advisor Sir Patrick Vallance acknowledged:
“The quicker we can get to a true community-based test the better.”
So, what’s holding this back?
Current COVID-19 testing practice
Since the beginning of the outbreak, most testing has been performed using the reverse transcription polymerase chain reaction (RT-PCR) technique. For this molecular test, swab samples need to be contained in a stabilising media, then transported to a central laboratory for analysis. A core benefit of this technique is that established technologies and workflows enable it to be conducted at a relatively large scale.
RT-PCR requires several hours of sample preparation and amplification, and the throughput of the available laboratory instruments is low compared to antibody-based tests. Some labs conduct manual sample preparation followed by thermal cycling, whilst others have fully automated instruments. The capability and capacity of these instruments varies greatly; the BD-Max system is capable of processing 24 samples in three hours whilst the Roche Cobas system can provide 96 results in three hours and 384 results in eight hours.
Central laboratories therefore have limited capacity to perform COVID-19 testing using molecular techniques. In the UK, a molecular test developed early on was initially handled by one lab in London capable of 100 tests per day. This was increased to 12 labs with a combined capacity of approximately 1,500 tests per day, then further extended to provide up to 8,000 tests per day. To date, around 0.7 people per thousand in the UK have been tested; scaling the testing base to the volume required is a challenge.
It’s clear that additional testing techniques are required. Since molecular testing only gives a positive result when the virus is present, there is a limited window of opportunity to test people. This approach is good for identifying infected patients and confirming recovery. But it won’t tell us who has had the virus already.
How to further scale COVID-19 testing
The million-dollar question is how to quickly and effectively scale COVID-19 testing to cover wider communities. The answer could lie with serological tests, which are simpler, quicker and cheaper than molecular tests. However, it can be difficult to distinguish the antibodies produced against other pathogens, so a careful approach is required to avoid false positives.
Serological tests look for the antibodies generated in response to an infection, meaning it is possible to identify those who have previously been infected and recovered. This is very useful for virologists and epidemiologists who want to understand the path of a virus, for example how many people were infected and who had mild symptoms or no symptoms. It is this data that will allow accurate estimates to be made surrounding case fatality rate.
In the initial stages of an infection there is a delay while the adaptive immune system mounts a specific response. In this period, molecular based tests have higher sensitivity. However, once antibodies are produced, they can increase rapidly to reach high concentrations in the blood. As the patient recovers, the antibody concentration decays but it can remain at a detectable level for some time. The length of time depends on the sensitivity of the technique and perhaps the severity of the illness, but it is typically a matter of months. This enables identification of asymptotic cases as well as those involving people who stay at home and don’t get tested during the illness.
Given the universal recommendation to self-isolate for seven to 14 days if symptoms develop, serological testing becomes an attractive option; after five days, antibodies should be detectable.
This would allow people who have been ill to know whether they have had the virus and act accordingly. They can avoid transmitting the infection or having to self-isolate many times, given the expectation that the epidemic will last for many months.
Serological testing can be performed on immunoassay platforms that have very high throughput. For example, the Abbott Architect i2000SR can perform up to 200 tests per hour and the Roche Cobas e801 module can perform 300 tests per hour. Furthermore, test facilities often have multiple instruments to cope with the normal demand for immunoassay-based testing.
A number of organisations have reported the development of serological assays for COVID-19. Euroimmun is developing a commercial assay which is in the process of being CE marked, a team at DukeNUS developed a test used for contact tracing in Singapore and the Wuhan Institute of Virology also deployed a serological test1.
Sample collection and preparation
Serological tests analyse blood, so sample acquisition and transport could pose a bottleneck. The volume of blood required varies between instruments. As an example, the Roche Cobas series requires 12-20uL of serum for its existing serological assays, however this ignores the steps of serum separation and sample handling in automated instruments. Typically, testing services request 3-4mL of blood and need a minimum 0.15mL of serum. Such volumes require venepuncture performed by trained professionals.
One area to explore is whether these tests might be performed using smaller capillary blood samples collected at home via a finger prick into a microtainer (with attention duly paid to the pre-analytical errors this might generate). An alternative might be to implement a drive-through testing arrangement for sample collection. This avoids the risk of potentially infectious people congregating in one place, but a sophisticated appointment system would be needed to make it workable.
Is there scope for a Point of Care test?
The ability to provide Point of Care (PoC) testing can deliver great improvements to the efficiency and effectiveness of disease management at the individual, local and macro level. PoC tests already exist for influenza and the same technologies could be applied to COVID-19.
For instance, the Abbott ID Now is a rapid molecular flu test analysing samples obtained using a nasopharangeal swab to provide diagnosis within 13 minutes. The use of such PoC tests varies greatly between different countries; they are most prevalent in the US.
Any PoC test instrumentation ideally needs to be widely available, so it’s only the manufacture of the final product that needs to be scaled up. Simplicity is also key, and existing solutions such as the Abbott ID Now might provide an effective blueprint, being based on minimal sample preparation and constant temperature RNA amplification. However, it is necessary to determine whether these instruments have the necessary sensitivity for COVID-19 before scaling up development and deployment.
Another potential area for development is the proven antibody-based PoC tests for flu. These offer the simplest possible format, being based on lateral flow strips (similar to a pregnancy test). They are low-cost and would be easy to manufacture in high volumes. However, even for flu, their sensitivity can be inadequate and the same is likely to be true for COVID-19. Using a test with insufficient sensitivity might well be worse than not testing at all.
Barriers to rapid scale-up
There are of course challenges to rapidly scaling any medical procedure, such as the regulatory requirements stipulated by the US Food and Drug Administration (FDA) and Centers for Disease Control (CDC). Nevertheless, these are exceptional times. In the US there has been a push to move testing away from the CDC and state labs to hospitals and commercial companies. On 29 February, the FDA changed its regulations, allowing diagnostic labs that have previously met federal quality standards to modify the CDC protocol or design their own kits. Subsequently, on 9 March, the CDC revised its COVID-19 guidelines to demand the testing of only one patient sample, not two, halving reagent needs.
Clearly the need for a rapid response to the crisis must be balanced with the need for high quality, reliable and repeatable tests. Additional considerations range from the availability of skilled people to obtain samples or conduct tests to the accessibility of materials and instruments required for the end-to-end process, including sample collection, storage and transportation as well as preparation and analysis.
Stepping up to the COVID-19 testing challenge will require a concerted and collaborative effort involving multiple organisations, agencies and countries. Taking parallel development paths which converge at the opportune moment is the best way to generate effective new solutions in a tight timeframe.
Existing techniques and technologies such as molecular, serological and antibody-based tests hold much promise. Evolving and adapting them could provide the answers we’re looking for. Knowledge sharing and resource pooling are vital. If scientists, technologists and medical specialists work together, we can learn quickly and protect people from this unprecedented threat to human health and life.
Sagentia has revolutionised the development of our clients’ lab testing instrumentation by driving costs down and reducing timescales. For more information contact Nick Collier at +44 1223 875200 or email firstname.lastname@example.org
ISO 9001 and ISO13485 accredited.
Sagentia is a global science-led product development consulting company that specialises in the design and development of IVD systems, instruments and consumables from point of care devices through to large scale central laboratory systems. We cover applications in clinical chemistry, haematology, immunochemistry, molecular diagnostics and genomics, working extensively on diagnostic consumables, microfluidic approaches for small sample sizes, low-cost solutions and multiplex systems.