Guardoc Health processes clinical documentation using Amazon Nova models

Details

Guardoc Health has published deployment figures for its AI clinical documentation platform targeting long-term care providers. The system handles the messy reality of clinical paperwork: multi-page PDFs with handwritten physician annotations, prior authorisation forms, medication lists in free text or tables, and mixed-format intake forms. Amazon Textract handles extraction; a RAG pipeline then pulls evidence from each patient's own records before Amazon Nova models on AWS Bedrock run condition classification, with DynamoDB as the data layer. The reported results are significant: 46% reduction in documentation errors, 70% drop in audit fines, and over $400,000 annual ROI per facility. A quarterly pilot covering two facilities and 200 patients produced 847 documentation corrections, flagged 86 Patient-Driven Payment Model reimbursement issues, and was associated with a 74% reduction in hospital transfers per 100 admissions. A broader study across seven facilities and 1,618 residents identified 10,612 issues. The critical caveat: every figure comes from Guardoc itself, with no published baseline period, independent audit, or methodology disclosure. The 74% drop in hospital transfers is a very large effect for a documentation intervention and should be treated with caution. That said, the RAG + Textract + Amazon Nova stack is concrete and reproducible, and PDPM compliance flagging is a high-stakes real use case where even modest accuracy improvements translate directly to Medicare reimbursement recovery.