MeadowCare

MeadowCare's MDS coordinators were spending 3-4 hours per assessment manually abstracting charts from the EHR, scanned paper records, wound photographs, and therapy notes. Clinical severity was under-documented because relevant data was scattered across systems — resulting in an estimated $2,840/day gap between documented acuity and actual PDPM reimbursement. CMS survey preparation consumed 2-3 weeks of dedicated staff time per facility, and the chain averaged 4.2 F-tag citations per annual survey.

Mixpeek's clinical documentation pipeline ingests all clinical content — EHR exports via FHIR, scanned charts through OCR, wound photographs through image analysis, and therapy notes through text extraction. A taxonomy classifier maps the extracted data to MDS 3.0 sections, and a hybrid retriever surfaces relevant documentation on demand for coding, billing, and audit preparation. MDS coordinators review and sign off on auto-populated sections instead of performing manual abstraction.

MeadowCare deployed the pipeline at a three-facility pilot group over six weeks. FHIR integration with their EHR provided structured data, while a parallel batch pipeline processed scanned documents and photographs. After validating MDS accuracy against manual assessments for 200 residents, the system was rolled out chain-wide over the following eight weeks. Total implementation from kickoff to full deployment was 14 weeks.