Steps in this run
| Step |
Calls |
Tokens in |
Cache hit |
Cost |
|
response generation
|
1 |
2,633 |
|
$0.01514 |
All 1 API call — tap to expand
$0.015144
Est. cost (USD)
Result preview
Ran the Bayesian math on this exact system last week. At average-risk prevalence (roughly 20-40 cases per 100,000 person-years), REDMOD's 88% specificity still generates around 120,000 false positives per million screened, with a PPV of approximately 0.18 percent, meaning roughly 1 true cancer found per 555 flagged patients. Each false positive triggers an EUS or contrast MRI cascade costing $3,00