Skip to content

IgniteHealth · Methodology

Every number on this site has a footnote. This is where they resolve.

We label every claim with a confidence — high when the source is direct and sourceable, medium when the figure is directional, projection when the figure is modelled. We never publish a number without telling you which it is.

medium

80%+ — of routine refills marked Met against the clinical protocol

Engineering target rate, measured against the protocol-eligible denominator from our internal evaluation set. PrefillX marks Met or Flagged — it does not approve refills. The pharmacist of record signs off on every Flagged refill. Not yet validated against production traffic.

ref: auto-approval-target

projection

+5–13 pp — PDC lift across MAC, MAD, MAH

Range of plausible PDC lift across the three Part D triple-weighted measures, derived from outreach-volume modelling. Actual lift varies by baseline adherence and member mix.

ref: pdc-lift-methodology

projection

$5–7.5M — in QBP bonus per 100K Medicare lives

Quality Bonus Payment range based on a 0.5-Star uplift at typical 2026 benchmark and rebate share assumptions ($1,150 PMPM benchmark, 65% rebate share). Plan-specific results depend on bid, region, and CMS rate announcement.

ref: qbp-bonus-math

medium

3–5 min — saved per routine refill review

Time-motion baseline for human refill review (3–5 min routine, 10–15 min complex). PrefillX runs the criteria check on the routine path, leaving Flagged refills to pharmacist review.

ref: review-time-baseline

projection

11 weeks — projected payback

Projected breakeven on PrefillX deployment for a 100K-life MA plan, assuming target Met rate and outreach reach. Modelled, not observed against a live deployment.

ref: payback-modelling

projection

4.2 → 4.7 — projected Star Rating uplift

Illustrative half-star movement scenario for a plan moving from 4.2 to 4.7 Stars on Part D adherence measures, post-Tukey-outlier-deletion 2027 cut points.

ref: stars-uplift-illustration

high

100% — reasoning-trace coverage by design

Every PrefillX decision produces a reasoning trace logged to the audit store before any outbound action. The system is engineered so a missing trace blocks the downstream message — there is no "untraced decision" path. To be re-verified by external audit once we are in production traffic.

ref: audit-completeness-sampling

FDA Clinical Decision Support. PrefillX operates under FDA enforcement discretion (21st Century Cures Act §3060). It is not a 510(k) device. The pharmacist remains the decision-maker of record on every script.