How it works

Legible OASIS QA, step by step

MsJaneHH guides the assessment, checks it against CMS rules, routes it through QA, and previews the export. Below is what the tool does today — described only where it's actually built.

Step 01

Guided assessment

Clinicians move through OASIS-E2 items one at a time. Built-in skip logic hides items that don't apply, so the form stays short and relevant. Voice dictation is available as a typing convenience — the clinician still reviews and enters every answer.

Step 02

Validation with CMS citations

Deterministic edit checks run against the answers and flag required-field gaps, invalid codes, and item-to-item inconsistencies. Each flag shows the specific reference behind it — e.g. Ch. 3 §G Functional Status — so it's checkable, not a black box.

Step 03

Correction workflow

Flagged assessments land in a reviewer queue showing each one's flag count and category breakdown. A reviewer can assign a specific flag back to the clinician with a note. Once the clinician responds, MsJaneHH re-runs the rule itself before the fix is allowed to close — a fix isn't accepted just because someone says it's done. The reviewer gives final sign-off after that check passes.

Step 04

Export, gated on real errors & audit trail

Export is blocked automatically while any flagged error is still open — not a policy, a hard gate in the tool itself. Once every error clears, preview and export the CMS-shaped file. Every create, edit, validation run, assignment, and sign-off is recorded on the assessment's audit trail.


Payment-impact review

Some OASIS items feed directly into PDGM case-mix scoring. MsJaneHH flags the ones worth a second look before submission — not to maximize payment, but so the record is complete, consistent, and supportable. These flags surface alongside the same citations as every other rule.


A readiness score you can check the math on

Every assessment gets a readiness score — the share of applicable required checks that actually pass. The formula is published, not hidden: it's a percentage, not a black-box confidence number, and it moves for a reason you can point to.


What happens after submission

QA is the front door. A rev cycle recovery module — in development — extends the same rules-based approach to denials: matching codes to CMS rules, drafting appeals, and tracking deadlines. See the rev cycle module →


What makes a flag trustworthy

A flag is only useful if you can verify it. Every rule in MsJaneHH carries a citation object — source document, section, spec version, and effective date — and the whole ruleset is versioned.

Rule referenceexample
M1021 · Primary diagnosis

Primary diagnosis is required.

source: OASIS-E2 Guidance Manual · §I Active Diagnoses · specVersion: OASIS-E2 · effective: 2026-04-01
ruleset oasis-e2-v2026.04.01

In progress: the ruleset is under active clinical review to confirm every citation against current CMS source documents. Rules whose reference is still being verified are marked for review rather than published as final — accuracy over coverage.

See the current item counts, rule counts, and last CMS source sync →

Want to see it on your own assessments?

Request pilot access