Real Clinical Outcomes

Challenge

Great Lakes had a high 30-day readmission rate for CHF patients discharged to home. Their care management team was spending 6 hours per day manually reviewing device readings from a fragmented fleet of 4 different RPM vendors, each with proprietary data formats incompatible with their Epic EHR.

Solution

Deployed the HealthCloud RPM Blueprint on the FHIR R4 connector platform, normalizing data from all 4 device vendors into FHIR Observation resources. Integrated a rule-based alert engine with ML risk scoring that filtered raw device events into a prioritized care team worklist inside Epic via CDS Hooks.

Outcome

Over 14 months across 2,800 enrolled CHF patients, the program reduced all-cause 30-day readmissions and enabled a 60% reduction in care manager manual review time, freeing capacity for higher-acuity patients.

Challenge

As a Federally Qualified Health Center network in rural Appalachia, Blue Ridge lacked the specialist access and lab infrastructure to accurately diagnose atypical infections and rare inflammatory conditions. Clinicians relied on empiric treatment, leading to high rates of unnecessary antibiotic prescribing and missed diagnoses.

Solution

Deployed SafeCDx, a point-of-care AI diagnostic assistant built on the HealthCloud platform. SafeCDx integrates with LabCorp FHIR APIs to pull lab results, applies a 40-condition differential diagnosis NLP model, and generates a structured DiagnosticReport FHIR resource surfaced as a clinical decision support card in the local EHR.

Outcome

Across a 6-month pilot at 4 clinics with 1,200 eligible patient encounters, SafeCDx demonstrated strong diagnostic concordance with specialist review and a measurable reduction in empiric antibiotic use.

Challenge

Meridian's physicians were spending an average of 2.4 hours per day on clinical documentation — more than any other non-clinical activity. Burnout scores were at a 5-year high. Prior ambient AI tools had failed adoption due to hallucination concerns and lack of FHIR integration for structured data population.

Solution

Deployed the HealthCloud AI Clinical Copilot across all inpatient services. The copilot uses ambient voice capture, FHIR R4 context from the patient record, and a fine-tuned clinical language model to draft SOAP notes, problem list updates, and discharge summaries. All outputs route through a physician attestation workflow before committing to Epic.

Outcome

Within 90 days of system-wide rollout across 3,200 physicians, nurse practitioners, and PAs, Meridian documented measurable improvements in documentation time, clinician satisfaction, and note completeness scores.