Pooled spits!💦All sorts of models have been suggesting what labs should be doing to increase sample throughput. But knowing that things aren’t always as easy as they seem on paper, we thought it best to try this out for ourselves (with saliva, of course!) medrxiv.org/content/10.110…
We pooled together one positive saliva sample with negative samples to form total pool sizes of 5, 10 or 20. We extracted RNA & tested for SARS-CoV-2 in RT-qPCR to check out how Ct values changed. While pooling saliva decreased virus detection, we still detected most infections. 
Based on the samples tested, we conservatively estimated that pool sizes of 5, 10, and 20 would lead to detection sensitivities of 93%, 89%, and 85% as compared to that of unpooled samples (though we do wonder how this would be viewed by the #FDA😉).
Using our lab data @EliFenichel modeled the best pooling strategy, depending on the expected prevalence of virus in the population being tested. We were excited about the resource-savings benefits of pool sizes ≥10, but above 3% prevalence, pooling by 5 actually proved better.
To put this in real world terms: at 0.5% prevalence, 10,000 people can be tested with just 1,318 tests-including the retesting of all samples from test-positive pools. If tests cost $30 each, this saves $260,453 vs. individual testing while still identifying 43-50 infections🤑
Also, pooled testing strategies do not have to be fixed! As the virus prevalence fluctuates, adaptive test strategies should be considered to maximize the costs-savings benefits. In turn, this will allow for continued surveillance against virus resurgence as communities reopen.
We did also explore (A) pooling of RNA extracted from saliva: results overall (B+C) were similar to when pooling samples before RNA extraction. While pooling pre-extraction did lead to more variable results, we still like this upfront approach due to its resource-savings benefits
Saliva & SARS-CoV-2💦:
- sensitive detection vs NP swabs✅
- stable detection after prolonged periods at elevated temperatures without fancy preservatives✅
- doesn’t need an expensive RNA extraction step✅
- can also be pooled✅
- incredible team effort still going strong✅✅✅
- sensitive detection vs NP swabs✅
- stable detection after prolonged periods at elevated temperatures without fancy preservatives✅
- doesn’t need an expensive RNA extraction step✅
- can also be pooled✅
- incredible team effort still going strong✅✅✅
Thank you @aewatkins6 @EliFenichel @WeinbergerDan for continuing to improve my knowledge of stats (and for your patience). And team @VogelsChantal @DougBrackney @IsabelOtt @ChaneyKalinich @XtinaHarden @arnaucasanovas @DelaCruzYaleMed @ShelFarFar @VirusesImmunity @NathanGrubaugh👏
