Quentin's profile
Activity for Quentin Hanley
No posts found.
No challenges found.
Improving COVID-19 Testing Efficiency using Guided Agglomerative Sampling
Proofs of the effectiveness of pooling in real samples are indirect proofs that smaller samples and volumes can be used. Investment in low-cost small-volume methods is needed. Cost and scale remain as limiting features of population scale testing.
Pooling mathematics apply to the test as an immutable basic unit.
In a world where the only show in town is 96-well plates, large volumes, and real-time PCR it makes some sense to pool. If moving on from methods that were adopted in emergency circumstances is possible, there are better returns on investment in my opinion. Existing technologies can be applied and properly systems engineered to bring costs down.
PCR Pooling: an application to manage pooling of samples increasing COVID-19 testing 8x
NHS Mass PCR Screening Pillar - Pilot for screening all 8000 staff at East Kent: 3CR end-point PCR as a mass PCR screening covid
Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
This will inform the pool size which can also allow projection of the frequency staff/patients can be tests for particular resources.
Finally, It will allow us to consider clustering as proposed in the documents Fayyaz has provided. We can check. This may also allow an increase in pool size.
Have I understood correctly.
Increase Covid Testing by 3, 9 or 15 fold
This is a pooling related proposal. There are others:
https://testingmethods.crowdicity.com/post/3166770
https://testingmethods.crowdicity.com/post/3172560
There are already papers indicating pools of 32 can work and suggest 64 are feasible:
https://www.medrxiv.org/content/10.1101/2020....9438v1.full.pdf
Other fully published work.
https://jamanetwork.com/journals/jama/fullarticle/2764364
In this one, there is a link to an R-script that attempts to estimate optimal pool size for any expected prevalence.
https://testingmethods.crowdicity.com/post/3172560
There is additional code in support here:
https://github.com/foxtrotmike/AS
Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
I have written an R-script that I think provides simulations to find the optimal number of pooled samples for a given prevalence. I have attached the script and an example graph. Please see if you can throw a brick through it. It is important to get this right.
The rough rule of thumb is that the optimal samples size are given by 1/prevalence. Practically due to variance related issues, I recommend rounding down or subtracting one from this number. For example, for 20% prevalence, the samples size should be 5. In a small operation, subtract 1. This is most beneficial for large number of samples.
Figures were produced as follows:
Pool5 had 20% prevalence (0.2), 20,000 samples, and 100 trials.
Pool100 had 1% prevalence (0.01), 20,000 samples and 100 trials.
The simulations make clear the issues with variance.
I use a binomial random number generator. The Rbinom call can be replaced with other distributions if you prefer these.
-Quentin
Improving COVID-19 Testing Efficiency using Guided Agglomerative Sampling
https://www.medrxiv.org/content/10.1101/2020.03.26.20039438v1
Obviously it hasn't been through peer review, but your experience could certainly provide some of that review for this platform. There are at least 3 similar proposals involving pooling.
Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Code (not a language I know but no matter):
https://github.com/foxtrotmike/AS
Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
https://testingmethods.crowdicity.com/post/3166770
Improving COVID-19 Testing Efficiency using Guided Agglomerative Sampling
Quentin Hanley voted on a comment on the idea PCR Pooling: an application to manage pooling of samples increasing COVID-19 testing 8x
Quentin Hanley voted on a comment on the idea NHS Mass PCR Screening Pillar - Pilot for screening all 8000 staff at East Kent: 3CR end-point PCR as a mass PCR screening covid
Quentin Hanley voted on a comment on the idea Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Quentin Hanley voted on a comment on the idea Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Quentin Hanley voted on a comment on the idea Samba II, developed by Cambridge University Hospitals with Diagnostics for the Real World, https://drw-ltd.com/
Quentin Hanley voted on a comment on the idea Samba II, developed by Cambridge University Hospitals with Diagnostics for the Real World, https://drw-ltd.com/
Quentin Hanley voted on a comment on the idea Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Quentin Hanley voted on a comment on the idea Improving COVID-19 Testing Efficiency using Guided Agglomerative Sampling
Quentin Hanley voted on a comment on the idea Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Quentin Hanley voted on a comment on the idea Pooled samples staff monitoring in Covid-19 care areas as part of Triple lock - starve the virus [regular covid frontline staff screening + all inpatient admission screening + every 7 days inpatient screening]
Quentin Hanley voted on a comment on the idea Samba II, developed by Cambridge University Hospitals with Diagnostics for the Real World, https://drw-ltd.com/
