Session ID: 

Push Not Pull: Using Data Science to Improve OR Operations

1:30pm - 2:30pm Thursday, March 8
Las Vegas - Venetian Convention Center
Lando 4301


Data science and machine learning – typically regression or classification analysis – are powerful tools for healthcare. While much of the attention given to data science in healthcare focuses on improving clinical outcomes, using these tools to identify opportunities for improving operational metrics have gone largely overlooked. While most hospitals administrators have a dashboard for operational data, including room utilization, the number of cases, first case on-time starts, turnover metrics, length of stay, etc., most use these tools in a “pull” manner; they start with something in mind and then look for the data. This causes problems when administrators aren’t clear on what to look for or just get lost in the reams of data. The problem is not lack of data, it’s lack of the insight of what data to use. Data science and machine learning – especially trend and anomaly detection – combined with innovative mobile data delivery mechanisms can turn “pull” into “push”, and uncover details that dashboards and EHR reports can miss.

Learning Objectives: 

  • Identify what metrics mean the most when optimizing hospital operations and understand why dashboards and reports often fail to disclose opportunities for improvement
  • Discuss how predictive analytics, machine learning, and mobile technologies were used to improve OR operations
  • Explain why “push” is the preferred method of data gathering versus “pull”


Senior Financial Analyst,
President and CMO,


Management Engineer and Process Improvement Prof.
Senior IT Executive