Application of pooled testing in screening and estimating the prevalence of COVID-19

Abstract

The recent emergence of the COVID-19 pandemic has posed an unprecedented healthcare challenge and catastrophic economic and social consequences to the countries across the world. The situation is even worse for emerging economies like India. WHO recommends mass scale testing as one of the most effective ways to contain its spread and fight the pandemic. But, due to the high cost and shortage of test kits, specifically in India, the testing is restricted to only those who are symptomatic. In this context, pooled testing is recommended by some experts as a partial solution to overcome this problem. In this article, we explain the basic statistical theory behind the pooled testing procedure for screening as well as prevalence estimation. In real world situations, the tests are imperfect, and lead to false positive and false negative results. We provide theoretical explanation of the impact of these diagnostic errors on the performances of individual testing and pooled testing procedures. Finally, we study the effect of misspecification of sensitivity and specificity of tests on the estimate of prevalence, an issue, which is debated a lot among the scientists in the context of COVID-19. Our theoretical investigations lead to some interesting and precise understanding of some of these issues.

Evaluation of efficiency and sensitivity of 1D and 2D sample pooling strategies for SARS-CoV-2 RT-qPCR screening purposes

“…4 Because the original Cq distribution depends on the origin of the sample (hospital, care center, …) and the stage of pandemic, our observations do as well. Firstly, our results confirm the widely accepted idea that sample pooling methods show a higher efficiency when prevalence is low [1][2][3][4][5][6]13 and that, for 1D and 2D pooling methods, as prevalence increases, a threshold is reached after which smaller pool sizes become more efficient 1,6 . However, appraising the performance of a pooling method exclusively by its efficiency would ignore one of the major drawbacks of pooling: loss of sensitivity due to dilution of the target.…”

“…The prevalence, however, cannot be known precisely, and as a result, the prevalence must be estimated. We can do this either before adopting a pooling strategy by testing the individual samples and using the fraction of positive samples as an indication for the prevalence or when a pooling strategy is already in place by calculating it from the percentage of positive pools 2,17 . Last, the calculated efficiency gain is merely a representation of the number of individual RNA extractions and RT-qPCR reactions and does not evaluate the amount of labor or time-to-result.…”

“…Several pooling strategies in the frame of SARS-CoV-2 testing have been discussed in recent preprints. The most discussed method is referred to as one-time pooling [1][2][3][4] .…”

“…Researchers have explored many strategies, but two of them have been welcomed for their simplicity and effectiveness: one-time pooling (1D pooling) and two-dimensional pooling (2D pooling). In 1D pooling, the samples are pooled, pools are tested and samples in positive pools are tested individually (Figure 1) [1][2][3][4] . Labs worldwide have extensively evaluated 1D pooling strategies for SARS-CoV-2 testing in the lab [5][6][7][8] or using simulations 1 .…”

Application of Pooled Testing Methodologies in Tackling the COVID-19 Pandemic

“…For example, if hospital capacity for future COVID-19 cases is to be planned, the government needs to have an assessment of the prevalence rate of the same. For some details of the prevalence estimation methodologies, the interested reader is referred to Gastwirth and Hammick (1989), Zenios and Wein (1998), Stephens et al (2000), Vansteelandt et al (2000), Roy and Banerjee (2019) and Guha et al (2020).…”