The 10 Most Common OASIS Scoring Errors (and What They Cost You)

Most OASIS errors do not announce themselves. They sit in a submitted record for weeks, travel through the billing cycle, and show up as a denial, a QA return, or a suppressed quality score. By that point, the clinician who completed the assessment has moved through two more caseloads, and the visit itself is barely retrievable from memory.

The errors on this list are not obscure edge cases. They are the patterns that MAC auditors at CGS and Palmetto GBA cite consistently in ADR guidance, the ones that QA teams flag most often in chart review, and the ones with the clearest financial consequence under PDGM.

For each error, this post covers what it is, why it happens, what it costs, and how to stop it, both through better clinical practice and, where applicable, through upstream automation that catches the condition before the clinician ever opens the record.

The 10 Errors

1. OASIS Responses That Contradict the Clinical Narrative  [ADR Trigger]

If the OASIS scores a patient as requiring minimal assistance with bathing and the visit note describes moderate assistance with soap and water, the chart has an internal contradiction. Reviewers at CGS have cited inconsistency between the OASIS and the clinical narrative as one of the most common reasons for ADR issuance in home health.

The contradiction usually does not come from carelessness. It comes from timing. The clinician scores the OASIS hours after the visit from memory, while the visit note was written closer to the encounter. The two documents capture different mental reconstructions of the same event.

2. Inaccurate Functional Scoring Under Section GG  [PDGM Payment Error]

Section GG functional items, specifically those covering self-care and mobility, feed directly into the PDGM functional impairment level calculation (source: CMS.gov, PDGM Overview). The three impairment levels: low, medium, and high, determine a component of the case-mix weight for every 30 days. An inaccurate GG score at SOC does not just affect payment for one visit. It affects payment for the entire 30-day period and, when carried through to recertification without review, the second period as well.

The most common pattern: clinicians score GG items based on what the patient reports rather than what they observe during the assessment. The OASIS guidance manual is explicit that scoring should reflect the patient's actual performance, not their self-report or their caregiver's report, when observed performance is available.

3. Homebound Status Not Clearly Supported in the Record  [Medical Necessity Denial]

A patient who does not clearly meet homebound criteria should not be admitted to home health. But the documentation problem is not usually about patients who fail to qualify. It is about patients who qualify, where the documentation does not make that case clearly enough to survive review.

Homebound documentation requires specific language: the taxing effort required to leave the home, the medical condition that makes leaving inadvisable, or both. Vague entries such as 'patient is homebound' or 'patient has difficulty ambulating' do not satisfy the standard. MAC reviewers at Palmetto GBA have cited inadequate homebound documentation as a leading cause of medical necessity denials since the Review Choice Demonstration expanded.

4. Diagnosis Sequencing Errors at M1021–M1023  [Wrong Clinical Group]

Under PDGM, the primary diagnosis at M1021 determines the clinical group, which is one of the five dimensions used to calculate the case-mix weight. There are six clinical groups: musculoskeletal rehabilitation, neuro rehabilitation, wound care, complex nursing interventions, behavioral health, and MMTA — and the payment rates differ meaningfully across them.

A patient admitted primarily for wound care after a surgical procedure belongs in the wound care clinical group. If the primary diagnosis entered at M1021 is a musculoskeletal code because the clinician or coder used the most familiar code rather than the one that best represents the primary reason for home health, the patient lands in the wrong group at a different payment rate.

5. Incomplete or Unreconciled Medication Lists  [ADR Flag / Care Gap]

M2001 through M2003 cover medication management, and the medication list at SOC must reflect what the patient is actually taking in the home, not just what the discharge summary lists. Hospital discharge summaries are frequently incomplete, outdated, or include medications the patient stopped taking before admission. The SOC documentation guide covers this in more detail, but the core requirement is clear: reconcile the list at the bedside against what the patient shows you.

Discrepancies between the hospital medication list and what the clinician finds in the home are not errors. They are expected. The error is failing to document those discrepancies and the reconciliation process.

6. Vague or Missing Wound Documentation at M1300–M1342  [Clinical Group and Quality Error]

Wound items, particularly M1306 (unhealed pressure ulcer at Stage 2 or higher) and M1342 (status of most problematic surgical wound), carry significant weight in both PDGM clinical group assignment and quality measure calculations. Vague documentation — entries like 'wound healing well' or 'wound intact' without measurement, tissue description, exudate, or periwound assessment — does not meet the standard required for either accurate PDGM classification or defensible documentation under audit.

M1342 in particular is a consistent QA return trigger. The item requires the clinician to select one of five status options based on specific criteria in the guidance manual. Selecting a status without documentation to support it is the most common pattern flagged during chart review.

7. Skip Pattern Violations Under OASIS-E1 Logic  [Submission Error]

OASIS-E1 introduced updated skip patterns, including the revised M0102 path, following the removal of M0110. Skip pattern violations, where a clinician or EMR completes an item that should have been skipped, or skips an item that should have been completed, generate iLogic errors on submission. Depending on the error type, iLogic errors can cause the OASIS to be rejected outright or to require correction before the claim can be submitted. The OASIS-E1 change guide covers each updated skip pattern in detail.

8. Inconsistent PHQ-2 Screening and D0150 Completion  [Quality Measure Suppression]

The Patient Health Questionnaire depression screening at D0150 follows a specific completion sequence. D0150A and D0150B are completed first. If either response is coded 1 or 2, the full nine-item PHQ interview follows. If both are coded 0, the interview stops. The OASIS-E1 revision corrected the item language that had created confusion about this path since 2023.

The error pattern: clinicians complete or skip the full interview based on the old item language rather than the guidance manual instructions. This creates inconsistency across clinicians in the same agency, which suppresses the agency's depression screening quality measure score on Care Compare.

9. SOC Assessment Completed Outside the Five-Day Window  [Technical Denial]

CMS requires the OASIS Start of Care assessment to be completed within five days of the first billable visit date. The date entered at M0090, the date the assessment was completed, must fall within that window. An M0090 date outside the five-day window is a technical error that can result in claim denial, independent of the clinical documentation quality.

The most common cause is not a missed deadline. It is an incorrect M0090 date entry. A clinician who completes the assessment on day two but enters the date on day six generates a false violation. This is entirely preventable at the data entry level.

10. Prior Episode Data Carried Forward Without Review  [Multiple Downstream Errors]

At recertification or resumption of care visits, clinicians sometimes carry forward OASIS responses from the prior episode without confirming the current status at the bedside. This is one of the most operationally understandable errors on this list; the prior data is right there in the EMR, the patient looks similar to the last episode, and the visit is already running long. But OASIS responses must reflect the patient's current status, not their prior status. The OASIS documentation automation guide explains the distinction between pre-population (presenting prior data as a starting point for review) and carry-forward (submitting prior data without review).

The financial consequences of carry-forward compound. A functional score carried forward from a prior episode where the patient was more impaired understates current function, suppresses improvement scores at discharge, and misrepresents the patient's clinical trajectory on Care Compare quality measures.

The Common Thread

Seven of the ten errors on this list share a root cause: documentation that happens long after the clinical encounter, from notes and memory rather than from real-time observation. The clinician sees the patient at 10 am. She documents the OASIS at 9 pm. In those eleven hours, details compress, specific observations blur into general impressions, and the record she produces does not match the assessment she conducted.

Three errors: homebound status, wound documentation, and medication reconciliation — require bedside clinical judgment that no technology can replace. Those three will always need clinical training, peer review, and QA oversight.

The other seven are fundamentally data problems: wrong timing, wrong sequence, wrong source, or no source at all. Those are the errors that pre-visit automation addresses directly, by moving documentation work upstream, enforcing OASIS-E1 rules before the clinician opens the record, and presenting prior data as a draft to confirm rather than a blank to rebuild.

Common Questions

• Which OASIS errors most commonly trigger ADRs?

MAC guidance from CGS and Palmetto GBA consistently identifies four ADR triggers: inconsistency between OASIS responses and the clinical narrative, inadequate medical necessity or homebound documentation, face-to-face encounter issues, and missing or late OASIS submission. Of these, the OASIS-narrative contradiction is the one most directly tied to documentation workflow rather than clinical knowledge.

• Can OASIS scoring errors affect HHVBP scores?

Yes. HHVBP performance in 2025 is based on improvement between SOC and discharge OASIS on functional and clinical measures. An inaccurate SOC baseline, whether scored too high or too low, distorts the improvement calculation at discharge. An understated SOC score inflates apparent improvement; an overstated one suppresses it. Either way, the score does not reflect actual patient outcomes.

• How does a diagnosis sequencing error affect PDGM payment?

Under PDGM, the primary diagnosis at M1021 determines which of six clinical groups the patient is assigned to. Each clinical group has a different base case-mix weight. A sequencing error that places a wound care patient in the musculoskeletal group, or a complex nursing patient in the MMTA group, means the 30-day payment is calculated from the wrong starting point. Depending on the clinical group gap, the payment difference can be several hundred dollars per period.

• What is the difference between pre-population and carry-forward?

Pre-population presents prior episode data as a starting draft for the clinician to review and confirm during the current assessment. Every field still requires a clinician's sign-off based on current findings. Carry-forward copies prior data into the current record without that review step. CMS requires OASIS responses to reflect the current patient status. Carry-forward, when used as a substitute for the current assessment, violates that requirement.

• How many of these errors can pre-visit automation prevent?

Seven of the ten on this list have an upstream automation component. Automation can pre-populate fields from referral data and prior episodes, enforce OASIS-E1 skip patterns in real time, flag M0090 date discrepancies, and surface prior data as a draft rather than a final record. The three errors that require bedside clinical judgment — homebound documentation, wound assessment, and medication reconciliation — need training and QA oversight regardless of the automation platform used.

Where to Start

Run a targeted chart audit before you change anything else. Pull ten SOC records from the last 30 days and check for the five highest-frequency errors: OASIS-narrative contradictions, GG scoring without documented observation, vague homebound language, M0090 date accuracy, and prior episode data carried forward without field-level confirmation.

If you find three or more of those patterns in a ten-record sample, the issue is systemic, and training alone will not fix it. The workflow that produces the error is the thing that needs to change.

For agencies evaluating what that workflow change looks like in practice, the OASIS documentation automation guide walks through how pre-visit automation changes the sequence, and the OASIS-E1 guide covers the specific rule changes that affect skip pattern errors and carry-forward documentation.