Strategy for pooled testing to scale up covid-19 testing capacity in The Gambia

The National Public Health Laboratory (NPHL) is the nerve center of the Gambia’s national anti-covid19 strategy. In recent weeks we have worked closely with colleagues at the MRC Unit the Gambia at LSHTM to develop COVID-19 diagnostic testing capacity at NPHL. As a COVID-19 testing center we can contribute towards increasing national testing capacity. However, with only two testing centers in the Gambia and limited resources to meet the potential demand for high numbers of tests, there is a dire need to explore innovative solutions for scaling up COVID-19 testing capacity in a short space of time. We wish to explore pooled testing as a strategy for rapidly scaling up COVID-19 testing capacity in the Gambia.
This project will include three main work packages: a training in COVID-19 epidemiology and pooled testing strategies, a validation stage to ensure that results generated from the pooled testing strategy replicate results from testing individual samples and finally an innovation component that will seek to implement the pooled testing strategy using sputum specimens as an alternative to nasopharyngeal swabs. We anticipate that successful implementation of these strategies will allow us to identify 5-10x more positive samples relative to individual testing with a fixed number of kits, while also considerably reducing the time between sample collection and positive sample identification.
Colleagues at the Harvard TH Chan School of Public Health and the Broad Institute of MIT and Harvard in Boston, MA United States, have developed a simple framework for COVID-19 pooled testing, which we seek to implement and expand on at the NPHL. The strategy involves combining multiple clinical specimens into a single pool and testing for COVID-19 using standard quantitative PCR assays. The aim is to classify individual samples as being COVID-19 positive or negative using fewer tests than would be required to test each sample individually. Pooling samples can be highly efficient when the majority of samples are negative (prevalence is low), as multiple negative samples may be classified after testing a single pool.
The strategy is flexible: the number of pools each individual sample is placed in as well as the overall number of pools can be varied depending on the number of samples to test and the available testing capacity. A standardized excel template, which can be tailored to different resource constraints and settings, guides the user on how to generate the pools and identifies presumptive positive samples from the pools for retesting; each sample is randomly labelled, and the same sample numbers are placed in the same combination of pools each time. Because the method can be carried out systematically following a pre-determined template, the logistical cost and potential for user error is reduced. In a low prevalence setting like the Gambia, up to 192 samples can be tested using 20 or less qPCR testing kits. This strategy has been written up, is available freely to the public as a preprint and is currently under peer review for publication. Our colleagues in Boston have tested and validated this approach in their laboratories. They will offer technical assistance in implementing this strategy in the Gambia and we will invite them to participate in training workshops for NPHL staff via Zoom.

Before deploying the pooled testing strategy, it will need to be validated in our labs. We will engage Dr Abdul Karim Sesay of the MRC Unit the Gambia for technical support in designing and executing the validation workflow. The validation will include two phases, an initial phase to determine the limit of dilution for pooled samples (how many samples may be pooled before sensitivity becomes too low) and a second phase to ensure that the results generated by the pooling strategy are concordant with results from individual testing. Briefly, we will select three archived samples that tested positive for SARS-CoV2 through qPCR testing, which represent a range of CT values as a proxy for viral load: border line CT (low viral load), average CT (Average viral load) and low CT (High viral load). Each sample will be pooled with a number of archived samples that tested negative for SARS-CoV2 through qPCR testing, using volumes in table 1. Nucleic acid extraction will be done for each pool followed by a SARS-CoV2 detection step using standard qpCR test kits. The limit of dilution will be defined as the largest pool size at which a positive sample can be detected.