London South East prides itself on its community spirit, and in order to keep the chat section problem free, we ask all members to follow these simple rules. In these rules, we refer to ourselves as "we", "us", "our". The user of the website is referred to as "you" and "your".
By posting on our share chat boards you are agreeing to the following:
The IP address of all posts is recorded to aid in enforcing these conditions. As a user you agree to any information you have entered being stored in a database. You agree that we have the right to remove, edit, move or close any topic or board at any time should we see fit. You agree that we have the right to remove any post without notice. You agree that we have the right to suspend your account without notice.
Please note some users may not behave properly and may post content that is misleading, untrue or offensive.
It is not possible for us to fully monitor all content all of the time but where we have actually received notice of any content that is potentially misleading, untrue, offensive, unlawful, infringes third party rights or is potentially in breach of these terms and conditions, then we will review such content, decide whether to remove it from this website and act accordingly.
Premium Members are members that have a premium subscription with London South East and have access to Premium Chat. You can subscribe here.
London South East does not endorse such members, and posts should not be construed as advice and represent the opinions of the authors, not those of London South East Ltd, or its affiliates.
@CautiousOptimist - I hear you ! Why oh Why do they need to use the same terms in different contexts !?
When I was involved with drafting the guidelines - each guideline had it's own glossary to define terms. After a few years the same term cropped up over and over in different guidelines glossaries with just subtle differences. Eventually, the job of gathering all the guidelines glossary entries together was given to the group defining the Common Technical Document. They struggled for so long that in al lot of cases they just duplicated glossary entries in the end to keep everyone happy.
This reminds me of that. - Anyway - I have read and re-read your description and explanation and I find myself in complete agreement with your writing and conclusions so somewhere along the line - how ever our wires got crossed - right or wrong I think I agree with what you described.
Hi Ophidian, thanks for taking the time to check it and reply. Indeed, not knowing the finer nuances of the terms in other settings, all I am going on is the clinical definitions, as described in Alistair’s video.
Perhaps a phrase in the explanation you cited that could muddy the waters for some is “actual ... results”. “Actual positives” (P, or TP+FN) is distinct from “positive results” (TP+FP). Hence, SENS = TP/P = TP/(TP+FN).
I think we’re on the same page in the end. But if this can cause differences in opinion among us scientists, even after watching Al’s excellent video, you can see why grappling with this causes problems for most of the general public!
Have a good week ahead.
@CautiousOptimist - to move on:
Negative Predictive Value NPV = TN/(TN + FN) so since a false Negative result is a negative result that SHOULD register as a positive I accept that the Sensitivity of the test has a bearing on the NPV.
Crikey this is tough for a sunday evening.
@Cautious Optimist - you've sent me scurrying back to the definitions. I have to go careful because to me sensitivity, selectivity and specificity are terms used to describe the suitability of an Analytical method and so the fact that terms are also used in a clinical sense often catches me out. However - not here I hope.
The clinical sensitivity of a test is defined as the ratio of True positive results (TP) to the actual number of positive results in the group being tested (P) SENS = TP/P
The clinical Specificity of a test is defined as the ratio of True Negative results (TN) to the actual number of negative results in the group being tested (N) SPEC = TN/N
Thee is a very good summary slide in the most recent Avacta video
So no - I didn't mean to type sensitivity
Great video Timster, watched it already and confirms my points. What isn’t written on the slide of definitions in the first minute, but leads a lot of people to confusion if you don’t grasp it, is:
Watch this all is revealed by big Al
Even I understood it..
I keep coming back to this as well CO.
Another way of putting it might be?
A high specificity gives a low number of False positives.
This would be desirable in low prevalence mass screening test.
This is not my field though so I'll defer to yourself and Ophidian to clarify.
Ophidian, when you said “A HIGH specificity gives a Low number of False negatives and so hence - a HIGH NPV (Negative Predictive Value)”.
Did you mean to type sensitivity here?
My reasoning, apologies if it’s like teaching you to suck eggs, but might be useful for others... If you’ve actually got the virus, sensitivity is about whether the test detects it or not, and so determines the fraction of actual positives that are either detected as true positives or not detected, I.e. read as false negatives. Conversely, specificity controls rates of true negatives and false positives for the actual negatives.
Negative Predicitvie Value (NPV) is a function of false negatives, true negatives and prevalence, so specificity and sensitivity both have an impact on it, albeit sensitivity is more important.
NPV will always be between (1 minus prevalence) and 1 - otherwise you wouldn’t use the test, so a high number for NPV is to be expected for low prevalence, but you want the test to push it closer to 1.
Therefore I think NPV doesn’t tell you much for low prevalence settings, e.g. population screening. Most important to remember is: high sensitivity reduces false negatives, which will help to detect and keep more infected people isolated. High specificity reduces false positives, meaning fewer people have to isolate for nothing while awaiting PCR confirmation, and less strain on the test and trace system.
My last post tonight (perhaps). But the conversations that you can have walking slowly across a vast landscape over 5 or 6 hours without the pressure to maintain a dialogue.
Never really been there before but driven through it quite a lot but you've got to get 'up there' to look down on it to realise how special it is. We're really very lucky in the UK to have such diversity in such a comparatively small and densely populated land.
My friend is a headmaster and had been to a talk recently about Covid and one speaker impressed him immensely. A long talk about Covid but in the summary she did the big build up to the 'Top 10' virus killers of all time. Of course she pointed out the current prevalent biggest viral killer only comes in at .......... Number 5 and of course that's HIV.
Covid just doesn't feature. We're obsessed because it's killing white people in the developed World. She maintained that if this was happening in just Sub Saharan Africa we'd probably be hardly aware of it.
Makes you think doesn't it?
CaptainStainley that is a beautiful part of the world! Been up there many times!
Fair enough, I’ll leave it to you to point out each one. I’m gambling on 100% of them, if by a miracle I miss out so be it - the RNSs are all that is important.
Yes! What a result. Even better if he went to GDR full time to upset their board was not nice there today. Just wish they would stop trying to diminish our company name buy changing the v to a w just to annoy people.
Bet I get a response though ;-) Translucent filter....
I filter no one and read everything albeit not real time today.
A lovely circular walk up 'Hay bluff' with wind and scattered torrential rain - which missed us all day long. Windy on the top like you wouldn't believe. Fantastic.
I do think it's important to point out when posts are particularly self important whilst simultaneously being apparently devoid of any true content. It'll be interesting to see if there is any answer as to what that post means or just insults to try to obfuscate the issue......
Over to you Orphidian.
CS, this was a suggestion made by CO yesterday and was a good one. His mounted lordship has latched on to that in the hopes he’ll be able to say it was his idea to soothe his fragile ego. My best advice to you is to put him on ignore, I have as I’m bored on his insane paranoia. All you’ll miss out on is vaguely educated guesswork. The waffly cheerleader has been binned too. Much better board today! Try it, you might like it...
Thank you for that detailed defence of your post Orphidian.
"The defence rest m'lord"
whatever works for you Captain - take your meds and let nurse put you to bed. Don't worry your little intellect over things you can't possibly understand.
"@minimil - A hospital only version of the LFD could even encompass the use of additional solvents potentially should that be required allowing for tailored optimisation to Hospital diagnostic use. All in time I suspect. let concentrate on getting the mas market one out first.
Does this actually mean anything or is it just pseudo scientific sounding 'mumbo jumbo'?
@minimil - A hospital only version of the LFD could even encompass the use of additional solvents potentially should that be required allowing for tailored optimisation to Hospital diagnostic use. All in time I suspect. let concentrate on getting the mas market one out first.
Ophidian - I also believe they need two types of LFT. If this is the case then I would have expected a faster drive towards approval/marketing the general population one (as this is the bigger $ earner). If the higher prevalence version requires more time to develop (as implied in latest RNS) why sacrifice sales if the lower prevalence is ready to go? Unless of course the lower prevalence is still not ready.
Thank you for taking the time to explain it helps a lot of us to hold on to this investment through the dips.
I think in my case though 'a little knowledge is dangerous'
@PAH00 - without fully understanding the compromises in the affimers per se - I don't think anyone outside the company can really answer that one. I'm not so sure that optimisation necessarily means playing one off against the other. More likely it is favouring one over the other in the mechanics of how the basic chromatography on the strip works. I could see how you might want to speed up the result for example and that might be by changing the viscosity on the strip perhaps which might mean changing the concentration of the affimer which could impact the sensitivity or selectivity more or less.
Just idle speculation as I don't have the necessary data to be able to say.
Just as a thought - who says they only have to produce one LFD. I could easily see one for low prevalence use for general population screening and another one for use in Hospitals and other higher prevalence areas. Why not ?
Apologies when I say lab test. I mean the professional use LFD.
Starting to realise that sensitivity and specificity importance is linked to the prevalence of the disease in the setting.
Would there be an argument that the lab test would favour sensitivity if used in a high prevalence environment. Whereas 1 or 2 % sensitivity may be sacrificed in the mass screening test to increase specificity?
Just to clear up the question from earlier. A HIGH specificity gives a Low number of False negatives and so hence - a HIGH NPV (Negative Predictive Value) which is what is required for getting the country back on it's feet.