Managing kidney care at scale demands precision and time. For a nationwide healthcare provider serving over 200,000 patients, clinical documentation had become a major obstacle to efficiency. The most time-intensive artifact was the comprehensive health evaluation (CHE), a multi-section summary consolidating lab results, medications, diagnoses, comorbidities, and care plans into a single longitudinal view.
Preparing one CHE required nearly four hours of detective work across multiple PDFs, EHR screens, and scanned faxes. With 8,000 CHEs due monthly, this translates into 32,000 clinician hours. That time is equivalent to 40 full-time nurse practitioners focused solely on paperwork. Beyond the $800,000 monthly labor cost, the hidden toll was diminishing patient interaction and declining job satisfaction.
The goal was to reclaim time for caregivers without compromising clinical rigor. The approach centered on integrating a large language model (LLM) into the workflow to handle repetitive extraction and drafting tasks while keeping final judgment in human hands. This strategy cut CHE preparation time by one hour per case, a 25% improvement that leads to significant savings and improves patient engagement.
