Houston Methodist

More than 70% of surgeries industry-wide don't start on time, and Houston Methodist's ORs relied on manual, staff-recollected documentation that was frequently inaccurate — recorded start times often didn't match reality. These small discrepancies compounded into systemic inefficiencies: cascading delays, unplanned overtime, and unused OR blocks with no reliable data trail to diagnose root causes.

Houston Methodist partnered with San Francisco-based Apella to deploy an ambient intelligence system using four ceiling-mounted cameras and audio sensors per OR, integrated with EHR data. AI-driven computer vision automatically detects key perioperative events — staff arrival, room readiness, anesthesia start and end — and sends real-time text alerts to surgeons. The system also generates retrospective workflow analytics to identify bottlenecks such as slow room turnover, long anesthesia setup, or extended cleaning cycles. Piloted in orthopedic and cardiovascular thoracic ORs in early 2023, it was rolled out systemwide in early 2024.

The pilot program delivered a 43% improvement in on-time first-case starts, a 20% decrease in turnover times, and a 15% increase in surgical cases handled without adding staff or physical space. Beyond efficiency, the system improved safety by enabling rapid review of adverse events — identifying the cause of a close call involving a malfunctioning laser and diagnosing recurring equipment failures that previously required weeks of investigation.

• Automating perioperative event capture with computer vision eliminates the manual documentation errors that silently erode OR efficiency at scale.
• Real-time alerts that keep surgeons informed of case status (patient arrival, drape up, room ready) smooth communication and reduce idle time without adding administrative burden.
• Retrospective workflow analytics — not surveillance of the surgery itself — are the primary value driver, enabling leaders to pinpoint exactly where time is lost between cases.