In surgery, it is accepted that there may be a ten to twenty-year learning curve to reach mastery for certain procedures. We believe this timeline can and should be shortened to improve patient care. Our short-term goal is to make a major contribution to the emerging field of Surgical Data Science by building a database of mastery level surgical performance and generating a roadmap for multimodal data collection and analysis procedures. Sharing our process, procedures and results broadly, will help to change measurement culture in healthcare. Through our newly developed partnership (October 2019) with the American College of Surgeons, we have already experienced early success in starting the conversation through the ?Surgical Metrics Project? (www.facs.org/education/surgical-metrics). Using a standardized data collection platform (a mastery-level hernia simulation), we will deploy and synchronize multiple data capture approaches (motion tracking, video, audio and validated surgical performance checklists) to build our database. Data analysis will quantify surgical mastery and consist of new applications and discoveries in machine learning. Hypothesis: Using multiple, synchronized data capture approaches and machine learning, it is possible to create a database of mastery level surgical strategies that can be translated into a value-added, surgical navigation tool for surgeons. To test this hypothesis, we will empirically investigate the following paraphrased aims:
SPECIFIC AIM 1 : Quantify surgical mastery (cognitive and technical) during a simulated laparoscopic ventral hernia (LVH) repair by using a post-procedure analysis of multi-modal performance metrics captured from hospital credentialed surgeons (N~125).
SPECIFIC AIM 2 : Establish validity evidence for surgical mastery metrics by comparing simulation-based LVH performance with operating room LVH performance from the same surgeons (N~60).
SPECIFIC AIM 3 : Empirically investigate the best implementation strategy for utilization of a surgical navigation tool designed to deliver value-added information regarding mastery-level surgical performance strategies to a new group of hospital credentialed surgeons (N~125).
The significance of this study is the opportunity to make a major contribution to the emerging field of Surgical Data Science by: 1) Generating a large multimodal dataset that can be de-identified and shared and 2) Providing a roadmap for predicting and supporting operative mastery with an artificial intelligence powered decision tool.
