Methodology
How a criterion earns a green. Two independent passes, one shared span, and no verdict in between.
Four steps per case
The engine parses once, then checks every criterion twice.
1. Parse the policy
Split the payer policy into discrete criteria, each with its own id and source line.
2. Policy to record
Take each criterion and search the chart for a span that meets it. Record the quote.
3. Record to policy
Read the chart independently and check the span back against the criterion it claims to satisfy.
4. Resolve
Green only when a pass cites a verbatim span that satisfies it. Otherwise the row is a gap.
Why two directions
A single read can miss a real match or talk itself into a weak one. Reading from both directions, policy to record and record to policy, surfaces more genuine support and still greens only on a verbatim cited span.
- A green is traceable to one verbatim quote, not a verdict. Every proven row carries the exact span a pass cited from the chart.
- Both passes are kept on the result for the audit trail, including the reasoning for a not-found.
- A disagreement is never split down the middle. It is a gap and the gap names what is missing.