Resources
Start with your number.
See a directional estimate of what your mid-cycle is leaving on the table — then dig into how leading finance teams are closing the gap.
Lost Revenue Calculator
Three inputs. Built from public industry benchmarks.
50900
2,00060,000
Estimated annual lost revenue
$13.2M
What's slipping through documentation, coding & denials
Undercoded severity & missed Dx$6.6M
Preventable denials$4.6M
Unfought / written-off appeals$2.0M
Recoverable with Medically$9.2M
Directional estimate from public benchmarks (~$11M / 250-bed baseline). We'll validate it on your real data — no obligation.
Guides & reports
Go deeper on the mid-cycle.
Guide
The CFO's guide to mid-cycle leakage
Where revenue actually slips between documentation and payment — and how to find it before the claim goes out.
Estimate yoursReport
Why payers are winning the denials race
Denials approaching 20%, 60%+ never appealed. What payer automation means for providers — and how to get ahead of it.
Request itBrief
Responsible AI you can defend to an auditor
Evidence citation, anti-upcoding guardrails, and human-in-the-loop — why defensibility is a feature, not a footnote.
Read our approachThe RFP checklist
What every buyer now expects from AI-driven CDI, pre-bill, and denials.
These are entry requirements in a 2026 evaluation — missing any one usually ends it. Use the list on any vendor, including us. The right column is how Medically answers.
| Requirement | What to look for | Medically |
|---|---|---|
| Coverage & evidence | ||
| 100% chart coverage | Every inpatient encounter reviewed pre-bill, not a sampled or prioritized subset. | YesEvery discharge, inside the bill-hold window, with no DNFB impact. |
| Whole-record reading | Structured data (labs, meds, vitals, orders, flowsheets) plus notes, radiology, and pathology. | YesLabs, medications, vitals, orders, flowsheets, microbiology, imaging and pathology results, notes, and prior codes. |
| Evidence-linked findings | Each suggestion cites source data with timestamps; the reviewer sees why in one view. | YesVerbatim, located chart spans on every finding. No citation, no finding. |
| Both directions | Flags unsupported diagnoses and downgrades, not only upgrades. Essential for compliance defense. | YesTwo-way findings in one list, each with a defensibility score judged against the payer's own criteria. |
| Human approval & audit trail | Accept or reject with reason codes, logged. Nothing auto-billed. | YesRequired reason codes and an append-only ledger with model and prompt versions. Nothing reaches the claim without sign-off. |
| Financial & quality math | ||
| DRG impact math | MS-DRG and APR-DRG shift, SOI/ROM, and dollar value per finding, used to prioritize work. | YesMS-DRG, APR-DRG SOI/ROM, and dollars on every finding, including downside exposure. |
| Quality capture | PSI/HAC, POA, Elixhauser and CMS risk variables, and HCC where relevant. | YesPOA, HCC/RAF with MEAT evidence, and quality-sensitive codes flagged for review. |
| Prioritized worklists | Ranked by dollar impact, denial risk, and bill-hold deadline, with auto-assignment. | YesAll three rankings, with auto-assignment to your reviewers. |
| Realized-revenue reporting | Dashboards that attribute billed dollars, not projected ones; CMI, CC/MCC capture, acceptance rate. | YesA finding-to-payment ledger: surfaced, accepted, coded, billed, paid, and denied-after-billing. |
| No DNFB delay | Reviews finish within the bill-hold window. | YesPre-bill review runs inside your existing hold; concurrent findings recompute mid-stay. |
| Integration, security & commercial | ||
| EHR integration | Epic first, then Oracle Health and MEDITECH, via HL7/FHIR; findings reachable inside the workflow. | YesRead-only HL7 v2 / FHIR R4 connectors for Epic, Oracle Health, MEDITECH, and others. |
| Security certifications | SOC 2 Type II and a HIPAA BAA at minimum; HITRUST helps in the public sector. | HIPAA · BAASOC 2 Type II · roadmapHIPAA-compliant safeguards and a BAA today; SOC 2 Type II and HITRUST are on our roadmap. Details |
| Fast implementation | Live in 8 weeks or less, with a retrospective look-back that proves value before go-live. | YesTarget go-live in six weeks or less, with a free 90-day look-back first. |
| Outcome-aligned pricing | Contingency or at-risk pricing, at least as an option. | Yes$0 pilot, pay only on results — or a flat subscription if your census swings. |
| Appeals that finish the job | Generated letters are table stakes; look for submission, deadline tracking, and payer analytics. | YesDrafted with payer policy cited, submitted electronically, tracked to the deadline, overturns reported by payer. |
| Beyond table stakes | ||
| Payer-specific clinical criteria | Sepsis-2 vs Sepsis-3, ASPEN vs GLIM, KDIGO — chosen per payer, not one algorithm for all. | YesA versioned criteria registry per payer; the version applied is recorded on every finding. |
| Clinical-validation defensibility score | A reproducible measure of whether a diagnosis will survive payer review — with its arithmetic. | YesDeterministic 0–100 score; breakdown and scorer version stored with it. How it works |
| Model governance | Published precision and acceptance rates, versioned models, drift monitoring, documentation you can review. | YesRates by finding type in your console; shadow mode before release; model documentation under NDA. |
Want this as a scoring sheet for your evaluation? Ask us and we'll send the checklist with room for every vendor's answers.
Ready to see your real number?
The calculator is directional. The demo is exact — run on your own charts.
Book a demoHave questions first? Contact us →