Mayo Clinic

Mayo Clinic's cardiac surgery schedulers faced the challenge of accurately estimating procedure durations for complex, multi-intervention heart surgeries. Traditional estimation methods were systematically conservative — overestimating durations to avoid cascading delays and unplanned overtime. This overcautious approach limited how many surgeries could be scheduled, creating access problems for patients requiring urgent, life-saving cardiac care.

Opmed.ai developed a custom AI model that predicts case durations for cardiac procedures at Mayo Clinic, accounting for procedure type, caregiver expertise, and patient clinical history. A user-friendly web application was built specifically for cardiac surgery schedulers, integrating predictive tools into existing scheduling and decision-making workflows. The model also estimates daily workloads for individual cardiac care providers.

In controlled testing, the AI model reduced Mean Absolute Error in case duration prediction from 60 minutes per case to just 34 minutes — a 43% improvement over traditional hospital estimates. This increased scheduling precision saves over 200 OR hours annually per operating room, boosting capacity for additional life-saving cardiac surgeries. The model's daily prediction accuracy enables schedulers to identify and fill gaps where traditional methods would have left OR time unused.

• Conservative human estimates systematically underutilize OR capacity; AI that accounts for provider-specific patterns can recover hundreds of hours annually without adding resources.
• Accuracy gains are most impactful for the top 10 most complex case types, where traditional hospital predictions historically struggle most.
• Precise case-length prediction is a prerequisite for downstream scheduling optimization — it unlocks the ability to sequence cases and allocate staff and equipment more effectively.