Diagnostic testing to identify patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) plays a key role to control the coronavirus disease (COVID-19) pandemic. While several countries have implemented the use of diagnostic testing in a massive scale as a cornerstone for infection control and surveillance, other countries affected by the pandemic are hampered by its limited testing capacity. Reference Deckert, Barnighausen and Kyei1 Diagnostic identification of all individuals infected by SARS-CoV-2, including screening asymptomatic persons in the incubation phase, is crucial to limit viral spread. Innovations supporting population-based mass testing are required.
Pooled testing was first introduced in the 1940s and is now used for screening in blood banks. Reference Abdalhamid, Bilder and McCutchen2 Testing is done by pooling multiple individual samples together, which are later tested with usual reverse transcriptase polymerase chain reaction (RT-PCR) systems. Only in the case of a positive pool test would individual samples of the pool be tested, substantially reducing the number of tests needed. Reference Bertolotti, Deaner and Deshpande3 It requires no additional training, equipment, or materials with several studies for SARS-CoV-2 already done in the United States, Reference Abdalhamid, Bilder and McCutchen2,Reference Hogan, Sahoo and Pinsky4 Israel, Reference Yelin, Aharony and Tamar5 and Germany. Reference Lohse, Pfuhl and Berkó-Göttel6
Previous studies show that individual positive samples of SARS-CoV-2 can still be detected in pools of up to 32 samples, and possibly even 64 samples, provided that additional PCR amplification cycles are done. Sensitivity for a pool size of 16 samples was 96% with an estimated false negative of 10%. Pooled testing could also potentially be applied prior to ribonucleic acid (RNA) extraction, thus saving invaluable time and resources. Reference Yelin, Aharony and Tamar5 Similar results were obtained by Abdalhamid et al., which were an increase in testing capacity by at least 69% when prevalence rates are 10% or less. Reference Hogan, Sahoo and Pinsky4
Hogan et al. tested 2888 individual nasopharyngeal or bronchoalveolar lavage samples grouped in pools of 9 to 10. Two positive samples were identified for a positivity rate of 0.07% and only 1 false positive reading was observed. Reference Hogan, Sahoo and Pinsky4 In Germany, pooled testing of 1191 samples in pool sizes between 4 and 30 samples resulted in only 267 tests required to detect 23 positive individuals (positive rate of 1.93%) with all positive samples easily identified. Reference Lohse, Pfuhl and Berkó-Göttel6
In Germany, 2 pooling techniques were compared, a “routine high throughput” approach where random samples are pooled together for testing or a “door to door” approach where groups of similar people (ie, families, neighbors, etc.) are pooled together for testing. While both approaches save substantial resources, the “door to door” approach was found to carry more benefit, reducing tests by 56% to 93%, whereas the “routine high throughput” resulted in 24% to 86% fewer tests. In low to moderate infection levels, even a pool size of 5 would reduce the number of tests needed by 5-fold (78%). In countries with infection levels over 20%, a pool size of 10 would still result in a considerable reduction in the number of tests required (up to 50%). Reference Deckert, Barnighausen and Kyei1
These studies showed that pooled testing is able to detect positive samples with sufficient accuracy. It is best used for population-wide screening, contact tracing, and the monitoring of essential workers and asymptomatic individuals with unidentified risk, such as in airports, versus being much less effective if used in settings with high clinical suspicion, such as patients showing symptoms.
While concerns exist that low positive samples such as those found in convalescent patients could escape detection with increasing pool size, additional amplification cycles could be employed to allow better detection of larger pools. Reference Abdalhamid, Bilder and McCutchen2,Reference Yelin, Aharony and Tamar5 Nevertheless, pooled testing shows great potential in increasing testing capacity with existing resources with minimal loss of accuracy.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Authors’ Contribution
Both authors contributed equally to conceptualizing and drafting this manuscript.