โ† Back to audit log

Agent invocation ยท 5/9/2026, 7:54:54 AM

Summarize survey risk for the facility.

Latency

17.3s

Tools

1

Tokens

4,851

Model

claude-opus-4-7

Final response

CaraDay Demo Facility โ€” Survey Risk Summary

Open survey-risk recommendations: 50 total

  • ๐Ÿ”ด Critical: 13
  • ๐ŸŸ  High: 37

Every open recommendation in the queue right now is categorized as survey_risk โ€” there are no other recommendation types competing for attention. That means the entire backlog (50 items across 50 census / 31 active residents) is survey-exposure work.

Context that shapes the risk picture

  • Census: 50 (31 active, 9 discharged)
  • Payer mix is Medicare-heavy: 27 Medicare + 13 Managed Medicare = 80% Medicare-line residents. Survey scrutiny on PDPM documentation, skilled need, and care planning is correspondingly high.
  • Acuity signals from average HIPPS letters:
    • PT M / OT M โ€” mid-to-high function-score therapy categories
    • SLP F โ€” high (acoustic/cognitive/swallowing comorbidities present)
    • Nursing I โ€” Special Care High / Clinically Complex range
    • NTA B โ€” very high comorbidity burden (12+ NTA points typical)
  • PT distribution skew: 21 residents at PTโ€‘O and 17 at PTโ€‘K โ€” the highest function-score tiers. Surveyors often probe whether ADL/Section GG documentation supports these levels.

What this likely implies (to verify case-by-case)

  • With NTA averaging B and Nursing averaging I, the 13 critical items are most likely tied to comorbidity capture, skilled-service justification, or care-plan/MDS alignment โ€” the same areas that drive both reimbursement integrity and F-tag exposure (F636, F655, F656, F684, F697, F812-adjacent documentation).
  • Only 1 resident is sub-benchmark on PT, so therapy intensity itself isn't the systemic risk โ€” documentation is.

Recommended next step

Pull the 13 critical survey-risk residents first so we can triage them by F-tag theme and assign owners. Want me to run that list now (highest-risk residents with open critical recommendations) and group them by likely F-tag driver?

Tool chain

  1. 01Facility summary
    485ms
    input
    {}
    result preview
    {"facilityId":"2ff057db-734f-4d27-824f-d4cbc864cac8","census":50,"activeResidents":31,"discharged":9,"payerMix":{"medicare":27,"managed_medicare":13,"ltc_insurance":5,"private":5},"hippsAvgLetters":{"pt":"M","ot":"M","slp":"F","nursing":"I","nta":"B"},"ptDistribution":[{"letter":"B","count":1},{"letter":"F","count":1},{"letter":"K","count":17},{"letter":"L","count":3},{"letter":"N","count":5},{"letter":"O","count":21}],"subBenchmarkPtCount":1,"recommendationsBySeverity":{"high":37,"critical":13},"recommendationsBreakdown":[{"severity":"critical","type":"survey_risk","n":13},{"severity":"high","type":"survey_risk","n":37}]}

Audit ID

dae079f5-4686-4948-a6a7-db7ce9902b49

Session

demo-session