Not sure that I am reassured yet
MAC
Would you not think that a sample size sufficiently small to be the cause of the variation kind of reduces the validity of the program results.
So we should ignore the 90%/75% and go back to a more likely 82%/85% as the more reliable figures.
How does Triage work then?
If it is lower sensitivity/higher specificity (than Detect) it will give more false negatives and more cancers will be missed.
Best Wishes
Paper Tiger
Yes I really do not get that diagram
Quote:
Originally Posted by
MAC
Cxbladder(triage) is a purpose designed screening tool per the diagram below;
Perhaps we all will have to wait to be convinced by Cxbladder(triage) Tiger as the spec has not yet been released by Pacific Edge, although some including myself anticipate that it should have a higher specificity and performance than NMP22 BladderCheck else Pacific Edge may not have pursued it’s development for this application.
Attachment 5977
Whilst I understand that the person on the left hand side of page 8 of the annual report is not Chris Swann even though the right hand side is titled "Chairman's Report" that Triage thing baffles me.
The test would need to have a really good sensitivity - not many false negatives - for a GP to be happy to accept the result and send the patient home (so to speak) given that apparently they pack 80% off to the specialist currently.
But high sensitivity implies low specificity so lots of false positives and you have lots of to the specialist anyway.
Or if you can achieve high specificity as well then it would be all round better than CxB Detect.
But it could be something else entirely as well, so does the patient turn up at the specialists with a note from the GP "I do not know what is wrong but it is not bladder cancer".
I guess we just have to wait and see...
Best Wishes
Paper Tiger