Case study: 47% denial reduction at a 12-provider cardiology group

When the CFO of a 12-provider cardiology group in the US Midwest engaged our team in early 2026, the numbers on the page told the story before the meeting started. Denial rate: 14.8%. Days in AR: 52. Net collection rate had drifted to 91.4%. Working capital was tight enough that the practice had deferred two provider hires and one imaging equipment upgrade.

Ninety days later, denial rate was 7.9%. AR days were 38. Net collection rate had recovered to 96.1%. The practice had banked $1.21M in annualized net recovery. This is a detailed account of what changed, in what order, and why.

The baseline: where the leakage was

A denial rate is a headline. The composition is what actually tells you where to fight. We ran a 12-month denial analysis on their submitted claims, and 71% of denials concentrated in just five CARC codes:

• CO-16 (Claim/service lacks information) — 23% of denials
• CO-97 (Payment adjusted because benefit for this service is included in payment for another service) — 18%
• CO-50 (Non-covered services because deemed not medically necessary) — 12%
• CO-197 (Precertification/authorization/notification absent) — 10%
• CO-11 (Diagnosis inconsistent with the procedure) — 8%

Two payers — UHC Medicare Advantage and Cigna commercial — accounted for 61% of the CO-16 and CO-197 volume. That concentration told us the fix was not "improve billing broadly." It was "close five specific failure modes across two specific payers."

The 90-day plan

Four workstreams, sequenced deliberately so each depended on the previous:

Workstream 1 (Days 1–14): Front-end eligibility hardening

CO-16 is almost always a registration problem masquerading as a billing problem. We instrumented real-time eligibility at check-in for the top 8 payers, added payer-specific validation rules for subscriber ID formats and group number requirements, and introduced a hard-stop registration QA check for the fields most commonly missing on rejected claims. CO-16 volume dropped 62% in the first three weeks.

Workstream 2 (Days 7–30): Payer-specific scrubber rules

Cardiology-specific bundling rules (CO-97) required deep specialty knowledge. Our coding team built 34 payer-specific scrubber rules covering the actual UHC MA and Cigna edit patterns — echocardiogram-with-catheterization bundling, TEE-with-cardioversion sequencing, ICD-10 laterality and specificity requirements. Every rule was tested against a 90-day retrospective sample before going live.

Workstream 3 (Days 14–60): Daily denial standup with coders

The single behavioral change with the highest ROI. Every morning at 8:30 AM, the coding team, an AR follow-up lead, and the practice's revenue cycle manager reviewed the previous 24 hours of denials — root cause, submitting user, and prevention action. Denials stopped being a monthly retrospective and became a real-time feedback loop.

Workstream 4 (Days 21–90): Weekly retraining loops

Denial patterns that persisted beyond one week were routed into structured retraining — 30-minute sessions on Friday afternoons with the specific coders or front-desk staff involved. Not blame. Repetition. By week 6, the top three recurring denial types from week 1 had all fallen out of the top ten.

The results, quarter over quarter

Metric	Baseline	Day 90	Change
Denial rate	14.8%	7.9%	−47%
First-pass acceptance	84.1%	93.6%	+9.5 pts
Days in AR	52	38	−27%
Net collection rate	91.4%	96.1%	+4.7 pts
Cost to collect	7.1%	4.4%	−38%
Annualized net recovery	—	$1.21M	—

Where the $1.21M actually came from

Deconstructing the recovery number helps translate this into something other CFOs can benchmark against their own practices:

• $472K — Denials prevented and paid on first pass instead of written off (65% of denied claims are never reworked; preventing the denial captures the full claim value).
• $318K — Working capital release from a 14-day AR compression on ~$50M annual gross charges.
• $241K — Rework cost avoided (average $118 per reworked claim × prevented denials).
• $179K — Improved contractual capture from cleaner coding on bundled cardiology services.

The full $1.21M is annualized from the day-90 run rate. The recovery in the actual 90-day window was $310K in incremental cash — enough on its own to fund the engagement three times over.

Why this worked — and what to watch for

Three factors distinguished this engagement from ones that stall:

• The CFO participated in the daily standup for the first three weeks. Executive presence early signals that this is the operation's priority, not a side project.
• We fixed the top five CARCs, not the top fifty. Focus beats breadth every time.
• Every rule change was tested before deployment. Scrubber rules deployed without back-testing cause new denials as often as they prevent them.

The pattern is repeatable. Most specialty groups running above 12% denial rates have similar CARC concentration and can achieve similar results in a comparable window. What varies is executive commitment to the operating cadence — the daily standup is the load-bearing element.

What the practice did with the recovered capital

The two deferred provider hires closed within four months. The imaging upgrade was ordered in Q2. Perhaps more strategically, the practice negotiated its next UHC contract from a position of measurable operational strength — clean claims, low denial rate, a documented performance improvement. Payer contracts, ultimately, are negotiated on the numbers you can defend.

If you're evaluating a similar sprint for your practice, our Healthcare Solutions team runs this exact 90-day model as either a co-sourced or embedded engagement. Read our companion pieces on why first-pass acceptance is the only RCM metric that matters and the CFO's guide to outsourced RCM, or book a discovery call for a scoped assessment.