Hi AICp,
From my correspondence with Tony:
"The 95% confidence intervals are calculable from the numbers of patients and the PPA/NPA provided in our last Australian release. The lower bound of the 95% CI for both PPA and NPA for all diseases in the Australian paediatric results were above the 75% target that we had for the US study"
I used ASX Results:
https://www.resapphealth.com.au/wp-content/uploads/2017/06/1684564.pdf
And these sites:
https://www.medcalc.org/calc/diagnostic_test.php
https://www.researchgate.net/figure...ercent-agreement-PPA-NPA-and-overall-rates-of
Happy to be corrected (could be one or two small errors), but double checked figures and I'm 98.5% confident they're correct. I could go into detail about calculations if anyone needs.
Cheers n1
- I didn't calculate for "cough only" as I figured those results are redundant going forward, as the 'best' model uses four features/not one
Math problem with results, page-216
Currently unlisted. Proposed listing date: 4 SEPTEMBER 2024 #