Scholars' Blog

Sample pooling as a strategy to contain the transmission of Covid-19

Covid-19 is currently spreading at an alarming rate. Despite showing signs of slowing down, the total number of infected individuals stands at around nine million as of late June 2020, which is astounding compared to previous instances of the coronavirus including SARS and MERS. In light of this, a strategy of mass testing, tracing and containing is paramount in keeping Covid-19 in check. Sample pooling is one such strategy to keep the virus at bay. Currently utilised for reverse transcription polymerase chain reaction (RT-PCR) tests, its ability to pool together multiple samples has promising implications for the efficiency of future testing.

Currently, the two most popular tests for Covid-19 are the polymerase chain reaction (PCR) tests and the antibody tests. The PCR test is the most common and most accurate test as of now. It starts with a healthcare worker taking a throat swab or sputum sample from a patient, amplifying that specimen in a machine hundreds of thousands of times, and then looking for unique coronavirus genetic material in a small segment of the RNA. The PCR test is currently recommended by the World Health Organisation (WHO) due to its accuracy, but it is costly to run compared to other tests. On the other hand, the antibody test does not detect the virus itself, but detects whether an individual has antibodies to fight off the virus. An individual’s finger is usually pricked to obtain a blood sample, after which a trained expert examines the blood specimen for antibodies specifically targeted to the coronavirus. The accuracy of the antibody tests is under doubt, and is currently only recommended as a tracker for the spread of Covid-19 instead of the main diagnostic tool. Hence, the strategy for all countries now ought to be maximising the number of PCR tests to be done to accurately diagnose all those who have been infected.

Sample pooling for RT-PCR entails five to ten genetic samples being grouped and tested together. Only if the test comes up positive will all samples need to be individually tested. As such, sample pooling strategy can significant increase testing efficiency and decrease costs for the RT-PCR test. It is especially useful in the mass testing of at-risk individuals, including patients with chronic illnesses, the elderly, especially those in nursing homes, and frontline hospital staff. Sample pooling has seen multiple successes in replicable experiments. In Germany, a team of scientists have published an article in the prestigious journal The Lancet Infectious Diseases, detailing the results of their sample pooling experiment. Quoting one of the lead scientists, “Our molecular biological test method (RT-PCR) was so sensitive that we were able to combine samples from up to 30 individuals in a single test tube. Even if only one person in that group of thirty was infected, we still obtained a positive result for the pooled test”. Reducing the number of RT-PCR tests to be done by a magnitude of 30 will massively increase efficiency and test rates. Additionally, pooling has been done for many diseases other than Covid-19 and has had a track record of success, including for syphilis screening, and testing blood donors for HIV, hepatitis B, and hepatitis C. In a study about malaria published in the Journal of Clinical Microbiology, pooling samples were able to cut over 50% of reactions and halve the cost of testing. The study argues that “pooling samples prior to diagnostic testing for low-prevalence gene targets in a population promises an opportunity to conserve resources without sacrificing diagnostic certainty”. Indeed, sample pooling has the potential to revolutionise and greatly cut costs of Covid-19 testing if done well.

The successes of sample pooling are not limited to only the laboratory. Currently, sample pooling is being done in multiple countries around the world, including China, Germany, India and Pakistan. It has been highly successful in Wuhan, where five to ten samples were pooled and tested together in a single test and an astonishing nine million people were tested in ten days. Testing a million people a day is no easy feat, but will do wonders for the Covid-19 pandemic if possible to replicate in other countries. Sample pooling was also used to test residents and staff at 131 nursing and residential care homes in Saarland, Germany. 22,000 people were to be tested, presenting a large logistical challenge but also an opportunity to showcase the effectiveness of pooling. On the other side of the world, India and Pakistan have also adopted sample pooling with high degrees of success.

Despite promising initial case studies, sample pooling is not without its problems. One such problem is the lack of sensitivity of sample pooling. Sensitivity is crucial for its success as samples get diluted when mixed with many other samples, which may lead to false negatives. As a journal published in the Journal of Clinical Microbiology points out, “pooling can decrease the sensitivity of assays due to the dilution of the samples, which is problematic in clinical diagnostics”. This problem is compounded for individuals with low viral loads who have just come into contact with the virus. False negatives are highly dangerous, as an infected individual could become an unwitting “super-spreader” when they mistakenly believe they are virus-free. This is especially problematic given the uniquely long incubation period (the time between exposure to the virus and symptom onset) for Covid-19, which is an average of five to six days, meaning an infected person could have spread the virus for a long time before realising they had the virus. Hence, a sensitive test is crucial to prevent dilution from distorting the test results. A second problem of sample pooling is cross-contamination. This can a) happen during sample collection, when some residue genetic material contaminates another test kit, or b) happen during the laboratory test when many samples are tested at the same time. For example, an RT-PCR test could test 96 samples, or even several hundred samples at once, increasing the chances of cross-contamination. This is what happened in Canada, when traces of carryover from sample to sample in the RT-PCR machines lead to false positives. At the same time, sloppy medical practices can also create cross-contamination. In the U.S.’s Centers for Disease Control and Prevention, researchers entered and exited coronavirus laboratories without changing their coats, and even assembled ingredients in the same room where other researchers were working on positive coronavirus samples. This also produces the problematic result of false positives. Certain jobs require the demonstration of a virus-free status, while some countries require proof that you are healthy before you are allowed to travel. In addition to the stigma associated with Covid-19, false positives can be devastating for individuals who may not have the virus. Hence, sensitivity and cross-contamination are two problems that must be dealt with before sample pooling is deployed en masse.

A running theme in the previous paragraph is the apparent mutual exclusivity of accuracy and efficiency. However, one possible solution to this dichotomy is the bacterial viral filter (BVF) produced by several companies in the United States. BVFs prevent cross-contamination by reducing the amount of bioburden (the number of bacteria living on a surface that has not been sterilised) by passing it through a filter, then putting it again through a second filter to make sure any bacteria or virus is not detectable. This may be able to successfully reduce the incidence of cross-contamination when collecting different swabs and samples. Another possible solution to the aforementioned problems is the Enhanced RT-PCR (ERT-PCR) test kit, developed by Professor Albert Yu of Peking University. The ERT-PCR has ten times the sensitivity of the conventional RT-PCR test due to an extra set of primers and an additional round of amplification. The increased sensitivity means that even patients with low viral loads, who are usually asymptomatic or pre-symptomatic, could be detected by this enhanced kit. In the context of sample pooling, the ERT-PCR could deal with the lack of sensitivity and false negatives the conventional RT-PCR could create. This technology has immense potential, as there would no longer be a trade-off between accuracy and efficiency if the ERT-PCR is successful. It remains to be seen whether the enhanced test kits live up to its potential, but the success of initial experiments suggests that we could be cautiously optimistic. Other possible solutions to the two problems include selecting well-performing primers to create reactions with higher amplification efficiency, or simply using aerosol-resistant pipette tips and designated work areas and pipettes for pre- and post-amplification steps to prevent cross-contamination. The solutions outlined above will hopefully increase both the accuracy and efficiency of RT-PCR tests.

Another concern is about the optimal number of samples to be pooled together. According to Professor Cheng of China’s Alfred Health, “if 5 to 10 per cent are positive, then you can only pool a few [samples]. Otherwise you'll need to re-test many individual specimens”. To alleviate the concern that a single positive specimen in a group may make it necessary to repeat tests on all the individuals in that group, leading to more tests being needed, an algorithm has been devised. This algorithm, developed by researchers from the University of Lahore, Pakistan, helps narrow down the number of repeat tests until the positive specimens are identified. The current strategy proposed, named “divide and conquer”, repeatedly splits up groups of samples to isolate the positive one. If a group of eight people is tested at the same time with a positive combined result, the eight will be divided into groups of four each, then tested again. If one group tests negative, the other group will be further divided in half and tested, and so on until all the positive cases are found. This algorithm will hopefully prevent instances where the pooling procedure is even more complicated than testing every single individual, maximising the efficiency of sample pooling. A similar solution to minimise the harms of decreased sensitivity is a pooling matrix, Tapestry Pooling, developed by IIT Bombay researchers from India. A technician enters how many samples they want to test and the prevalence rate, after which they are given a pooling matrix. “Samples are pooled according to this pooling matrix. Pools are tested and the quantitative reading from the tests is entered into our algorithm, which reconstructs the list of positives, negatives, and their viral loads,” says Manoj Gopalkrishnan, a researcher from IIT Bombay. These two examples of technology utilisation will arguably push sample pooling to even greater and more accurate heights.

It is unlikely that an effective vaccine for Covid-19 will be developed anytime soon. Similar viruses, such as SARS or even the common flu exist without a cure or with a vaccine that has to constantly evolve to face new strains of the virus. Additionally, it has not been proven that the antibodies produced by infected individuals after recovery actually prevents them from becoming ill again from Covid-19. In fact, antibodies could very well have the opposite effect. Antibody-dependent enhancement (ADE) is a phenomenon where antibodies actually increase the likelihood of you contracting the virus again in the future or suffering from more severe symptoms the second time you are infected. This has been observed with the Dengue virus, Zika virus, HIV and even certain strands of coronavirus. As a result, the most likely outcome is that we will have to learn to live with the Covid-19, similar to how we currently live alongside the common flu. This means that the strategy of trace, test and contain is all the more important in limiting its spread and potential victims. A vision where all citizens are tested periodically in private clinics and public hospitals and positive cases isolated, with everyone going about their normal lives, is not a utopia, especially with the ever-improving technology of RT-PCR tests. There is still light at the end of the tunnel.

 

About the author
Ayo Yi Ngok Chan went to Blavatnik School of Government as one of the inaugural Kwok Scholars in 2012. This article was published on 17 July 2020 by Blavatnik School of Government.