Scientific FACTS - you go check yourselves24 Aug 2024 18:35
I must say congratulations to Greend100 for some execllent posts, at least trying to understand the science and financial implications for himself and Cizzle. (Iβm sure he wonβt desire this accolade from persona non grata!)
However science is science.
For those who donβt like facts - please donβt read.
The Sensitivity/ Specificity/PPV/ NPV and Accuracy results presented by CizzleBio itself are POOR.
For those not versed in science, the results show NPVs (Negative Predictive Value*) of 94, 96 and 85 respectively - which is very good.
But a PPV (Positive Predictive Value*) of 34, 42 and 55 respectively - which is VERY Poor.
The overall ACCURACY of the the test was 42, 42 and 51 respectively - Accuracy measures how well a diagnostic test correctly identifies the presence or absence of a disease overall.
Itβs proportion of all test results (positive and negative) that are CORRECT.
51% is the best result?!!!
The PPV and NPV values are based on βprevalenceβ in a population. So this depends on the age and demographics of those tested.
Eg If it was a test for a Dementia protein - the accuracy of the test would be skewed towards those more elderly. Ie the true positives would be greater with an older cohort and lower with a younger cohort of testees. (Thus skewing the data)
So Cizzleβs Lung cancer Negative test accuracy (The NPV) is excellent, but the PPV and over Accuracy is terrible!
And PPV is probably the metric that healthcare professionals will use when deciding whether to use the test en masse!
Look at other tests by other companies for other conditions - eg Oxford Biodyn PSE test, accuracy, PPV and NPV of 94, 93 and 95% respectively. Look at other tests out there and compare.
I ask long-term investors again:
If you were starting from scratch today, using Ciz own data/ projections (Not me, not my spin, but their OWN data), would you honestly now invest in this?