Post-Op Vitals — Baseline vs. Retest

Re-evaluate new-patient acquisition, collections, and cost. Compare against baseline. Same instruments, same locations, same finance team in the room. The numbers either moved or they didn't — and where they didn't, this sheet says so first.

VITALS — DAY 0 vs RETEST · SAME INSTRUMENTS · DESK RATES STILL HAVE NO PUBLISHED BENCHMARK

Eleven, not twelve. The group runs 12 locations; 11 were trading at Day 0. Parkway-11 opened March '26 and has no Day-0 reading — it never will. Every delta above is same-store across those 11. All-in rows exist so the CFO sees the whole P&L; they are not growth and are not presented as growth. Per location that is 19.4 new patients/loc/mo at Day 0 and 22.2 at retest, both on 11 sites. Methodology, sheet 3 §02③.
Six moved for the group, two against it, one not at all — and one desk on sheet 4 didn't move either. A retest in which every vital improved would mean the instruments were rebuilt to flatter the work, which is exactly what freezing the definitions at Day 0 exists to prevent. Reactivation (patients dormant 18+ months) is excluded from both sides of every acquisition figure.

"Same instruments" is not a sentiment. It is a frozen definitions document, a re-run of the same extract against the same systems, and a written disclosure every time reality forced a definition to move.

A retest is only a retest if the thing being re-measured is the thing that was measured. Before a single figure was pulled, the definitions document issued with Form 247-B on 25 July '25 was read aloud against each current source system and each instrument logged as HELD or CHANGED. Seven held. Three changed, and all three are expanded on the next sheet with both readings printed. Nothing on this list was quietly adjusted to make a delta look better; where a definition moved, the less flattering reading is the one that carries the headline.

Why "held" is a finding and not a formality. Seven of ten definitions survived fourteen months of an engagement in which we were actively changing the systems those definitions read from. That is not luck; it is the reason the definitions document exists as a separate artifact with its own change log rather than as a paragraph in a report nobody re-opens. The three that moved are disclosed rather than absorbed, because an instrument that changes silently produces a delta that means nothing — and a delta that means nothing is worse than no delta, since somebody will act on it.

03 · SAMPLE SIZE AND WINDOW — STATED FOR EVERY RATE ON SHEET 1

Why a large n can still fail to carry a decision. Broken appointments are measured across 34,900 scheduled appointments — the n is not the problem. The vital is a six-point monthly series, and the Day-0 series ran 21.0% to 23.6% with a standard deviation of 0.93 points. A 0.8-point move sits inside one standard deviation of the instrument's own noise, so we do not claim it and we do not let anyone else claim it either. The last three months of the retest window average 20.4% against 22.5% for the last three of the Day-0 window. That is suggestive. It is also not the instrument we froze, so it stays in this footnote rather than in the delta column, and it gets re-read at the annual retest when the frozen window has moved past the fix.

The best desk did not move, and that is the correct result. North-04 answered 96% of new-patient-opportunity calls at Day 0 and answers 96% now. There was nowhere for it to go: the ceiling on this instrument is not an industry chart — dentistry publishes none for desk answer rate — it is this network's own best desk, and North-04 is that desk. The gains came from the bottom of the distribution, which is where they were available. Best-to-worst booking spread closed from 24 points to 17 and remains the largest single variance in this file.
Denominators, so the two desk documents can be read together. The answer rates above cover new-patient-opportunity calls only, at three desks, in one week. Form 247-K reports Riverside at 81% for week 29 — that is all calls on all lines. Both use the frozen definition; the populations differ, and a reader comparing 76% here to 81% there is comparing two correct numbers measuring two different things.

Four of five in the same direction, by the same mechanism. Contractual adjustments up, production per patient down, collected value per patient down, FFS/cash mix down: that is one thing, not four. A larger share of first visits are unestablished and insured patients, and unestablished insured patients write down more, accept less, and produce less. It is the standing price of buying volume rather than a defect in the buying, and it is printed at the same weight as the volume gain on sheet 1 because a CFO who learns it from us in September is a different conversation from one who finds it in a schedule in March.

NO CONTINUE / CONCLUDE RECOMMENDATION APPEARS IN THIS DOCUMENT. IT IS A READING, NOT AN ARGUMENT FOR OUR OWN RENEWAL. THE DECISION DOCUMENT IS FORM 610-Q, AND IT CARRIES THE CASE FOR CONCLUDING PRE-FILLED.