kingrollo
Guru
Perhaps we invite staff in from overseas - free movement of people and all that...
Page may contain affiliate links. Please see terms for details.
Perhaps we invite staff in from overseas - free movement of people and all that...
winjim
Smash the cistern
Yeah, it's literally what we* do all day every day. The quality system should be in place to catch any errors.
*I'm not a registered BMS but I do a very similar job. I don't do PCR though.
vickster
Legendary Member
Are they not busy doing testing in their own countries though?
newfhouse
Resolutely on topic
I’m not sure now is the time to be poaching skilled people from other health services, much as I agree with the sentiment about free movement.
kingrollo
Guru
What u mean other countries might have trained extra staff* to help with the increased testing workload ? - its a thought I suppose !!!
*And they probably had more to start with !
....Anyway were ramping up testing the next week ! (Phew !!!!)
winjim
Smash the cistern
Has anybody read the IBMS piece? They've been cross training staff and prioritising workload. Staffing is not currently a major issue. For example, and I've been isolating so not at work so I don't know the exact situation, but I expect the specialist work I do to have gone down to a bare minimum of urgent cases only, freeing up staff to work in other sections of the lab. There is a certain amount of flexibility and staff can be redeployed.
Remember this will be covering not only COVID19 testing, but other testing relating to acute admissions, as well as regular lab work. It's not all about PCR.
Remember this will be covering not only COVID19 testing, but other testing relating to acute admissions, as well as regular lab work. It's not all about PCR.
Is any work being down around looking at using AI to help with the increased volume of testing ?
I understand that AI is being looked at for breast screening and is being used for reviewing some cardiac cases.
vickster
Legendary Member
No idea. I just don’t think it’s right to poach staff from other countries during a global crisis.
And I’m certainly not looking for any sort of discussion on testing
winjim
Smash the cistern
I believe in the case of breast cancer screening that AI is useful because there can be some degree of subjectivity in the interpreration, and an AI can be trained to spot certain patterns. I don't know much about PCR interpretation but in my field, biochemistry, I'm not sure it would be too useful for individual results. Maybe in spotting overall clinical patterns, but we do have algorithms in place for some things like that. But even all our automated processes are checked and authorised by a BMS or Clinical Scientist before release.
Anyway, even if it were for some reason desirable, just the time it would take to procure, train, test, troubleshoot and validate a system would be prohibitive. As you can imagine we have a very rigorous quality management system it would need to get through and I don't think it could be done.
That's my take on it too I know it was used to track the spread as it took over most of world and spotting trends.
More than likely it's helping track it across the county as the moment. Once this is over a whole lot of new data will need to gone though without AI it going to be one difficult job.
I guess most of my question was more a general question really. Like you say it can't take over from skilled staff but when done well it can enhance them.
A different tribute to the NHS last night from Transdev in Harrogate (but not Arriva)
View: https://twitter.com/sunrise2409/status/1245792605922828288
View: https://twitter.com/sunrise2409/status/1245792605922828288
Only in the Uk in the middle of a national emergency do we still carry on with all the photo opps , all the big wigs and all the other nice bits of a formal opening ceremony. Of an emergency hospital ... Nice to see they have the priorities right.
I noticed talk of old Florence glossed over the bad bits.
I noticed talk of old Florence glossed over the bad bits.
winjim
Smash the cistern
Of course, labs are moving increasingly towards automation. And if you can get software to do a lot of the drudgery, it frees up staff time for the more interesting and unusual cases.
Machine Learning is used in many areas of medicine and particularly diagnostics. There are hundreds and hundreds of machine learning papers about diagnostics. It's something that will change a lot the next 10-20 years.
The use machine learning could probably be most useful for in this outbreak would be immunity passports and antibody test data. Our government really are wary towards antibody tests (you can still derive useful information from machine learning in theory even if the test isn't as accurate as you like, which is one reason why I strongly support the test). Aside you could very easily have too much data if they got their arses in gear. That's a not particularly easy job of prioritising who gets tests/what it means.
Machine learning I hope is being used collectively on the roughly 10,000 tests a day and the complete mess over retesting the same patients multiple times for inconclusive results.
