How UC San Diego Health cut procedure no-shows by 40% with AI agents
Building on proven results in gastrointestinal procedure preparation, UC San Diego Health is expanding AI-powered patient readiness and care navigation to improve safety, quality, and access to care.
reduction in no-shows and same-day cancellations, from ~10% to as low as 6%
lowest no-show and same-day cancellation rate among patients completing AI conversations.
with clinical oversight and defined escalation protocols
For patients, receiving care often requires more than showing up for an appointment. Medication instructions, preparation requirements, symptoms, and transportation can all affect whether care can safely proceed.
At UC San Diego Health, nurses and schedulers spent significant time calling patients before gastrointestinal (GI) procedures like colonoscopies to review preparation instructions and identify potential concerns. Many calls reached voicemail, while some patients arrived without completing necessary preparation, resulting in avoidable cancellations and delays.
With no-show and same-day cancellation rates around 10%, UC San Diego Health saw an opportunity to improve patient readiness, protect procedural capacity, and give nurses more time to focus on patients requiring their clinical expertise.
UC San Diego Health partnered with Notable to deploy AI agents that proactively engage patients via Voice AI before GI procedures, using Epic data and health system-approved instructions to guide preparation, identify barriers, and surface clinical concerns.
In one conversation, a patient reported that insurance would not cover their prescribed bowel preparation and that the pharmacy was seeking an alternative. The AI agent identified the issue and flagged it for nursing follow-up before procedure day.
Other conversations surfaced missing medications, reported symptoms, and questions about post-anesthesia transportation.
Unlike standalone voice automation, Notable connects these interactions directly to clinical workflows. Call transcripts and summaries are documented in Epic, while concerns requiring intervention are routed into appropriate work queues. Nurses can review what patients were told, identify who needs assistance, and take action within their existing workflows.
This approach is especially important for clinical safety, where medication adherence, preparation requirements, and potential complications demand consistent protocols and appropriate escalation. Clinical decisions remain with UC San Diego Health’s care teams.
The initial workflow was deployed in weeks, with UC San Diego Health and Notable jointly refining instructions, clinical exclusions, and escalation criteria through ongoing review.
UC San Diego Health reported a reduction in gastrointestinal procedure no-show and same-day cancellation rates from approximately 10% before deployment to as low as 6% across all call outcomes, a 40% relative improvement. Among patients who successfully completed AI conversations, rates were as low as 2%.
Beyond procedure slot utilization, the approach helps clinical teams identify preparation barriers earlier and focus nursing attention on patients requiring intervention.
Building on its successful GI results, UC San Diego Health has expanded Notable into inbound Urgent and Express Care scheduling, using health system-defined symptom-routing protocols to support appropriate care navigation.
The health system is extending this model across additional workflows, establishing a more proactive approach to patient readiness that creates clinical capacity, supports safety and quality, and helps more patients receive timely care.
Get in touch to learn how you can use Notable’s Platform to provide safer, more timely care for your patients.
Frequently asked questions
How can AI reduce patient no-shows and same-day cancellations?
AI agents reach patients proactively and consistently, which manual call programs often can’t do at scale. They walk patients through preparation, answer common questions, and surface issues like missing prescriptions or insurance problems in advance. Care teams can then resolve problems early or reschedule, so open slots can be refilled instead of lost.