hh-health-ai/healthcare-equity/guideline-inclusion

Guideline inclusion tracker

Track guideline-body decisions as dated adoption catalysts and translate them into coverage and volume consequences.

Workflow

1. Identify the relevant body and the exact mechanism for the product:

• Preventive service or screening test → USPSTF (A/B grade triggers no-cost-share

coverage under ACA §2713, generally effective for plan years beginning after the recommendation).

• Vaccine → ACIP vote, then CDC director adoption, then the VFC resolution.
• Oncology therapy or off-label use → NCCN category, because payers reference

the compendia.

• Chronic disease management → specialty society standards (ADA Standards of Care

each January; ACC/AHA; GOLD; KDIGO), which move practice but carry no coverage mandate — a slower and weaker catalyst that is often over-modelled.

1. Get the calendar. These bodies publish meeting dates, draft-comment windows and

publication schedules in advance. Draft recommendations are posted for public comment before finalisation, which usually gives one to three months of visibility on the likely outcome — the single most useful and least used feature of this catalyst class.

1. Read the draft, not the rumour. The draft's evidence statement and the comment

volume tell you how contested it is. Monitor the page with scripts/page_snapshot_diff.py so a change is detected the day it happens.

1. Trace the coverage consequence explicitly. Which payer types are obliged, from

which plan-year date, and with what cost-sharing change. Then hand it to a reimbursement engine → coverage-check for the Medicare position, which does not follow automatically from a USPSTF grade.

1. Apply the governance-risk overlay. State plainly that the 2025 disruption to

ACIP and USPSTF composition weakened the mechanical reliability of this linkage, and handicap the coverage consequence accordingly rather than assuming it.

1. Quantify through the funnel. A screening recommendation acts on the diagnosed

step of the epi-demand funnel, not on share. Re-run the funnel with the new diagnosis rate rather than applying a growth rate to revenue.

1. Check the lag. Guideline-to-practice change is slow and uneven: coverage changes

at the plan-year boundary, clinician behaviour over quarters to years. Model a ramp, never a step, unless the mechanism is pure cost-sharing removal at a fixed date.

1. Emit the brief.

Not-automatic

A guideline recommendation does not license a revenue step change. It licenses a change to the coverage terms and to one specific step of the demand funnel, with a lag, and now with governance risk attached.

Reference: references/guideline-bodies.md. Contract: ../../references/evidence-brief.md.