ISQUA16-2984COMPUTER AIDED QUALITY ASSURANCE IN THE PLANNING OF THE SURGICAL PATHWAY

Abstract

Objectives: The surgical pathway is a complex, cost-intensive activity within a hospital that should be managed efficiently. The aim of this study is to automatically manage the scheduling of the operating theatre through real-time monitoring of all critical steps in the perioperative process and to present this information in a structured and understandable fashion to the personnel. Methods: We optimize operating room (OR) planning and scheduling by predicting OR time using the information that is available in the preoperative and perioperative phase. This includes both patient characteristics (e.g., age, gender, co-morbidity), and information that can be extracted from the OR environment in realtime, like the use and status of instruments and equipment. To first identify distinguishing phases in the procedure we monitor the activation pattern of specific OR equipment, such as the electrosurgi-cal device. Activations of the electrosurgical device were detected with a current sensor placed between the power plug of the device and the socket in the OR. The activation pattern of the electrosur-gical device was measured during 57 laparoscopic cholecystecto-mies. These patterns were used to train a support vector machine classifier that detects whether procedures are shorter or longer than a certain amount of time. During 21 laparoscopic cholecystectomies the activation of the electrosurgical device was recorded and processed in real-time using pattern recognition methods. The remaining surgical procedure duration was estimated and the optimal timing to start preparing the next patient for surgery was communicated to the OR staff. A tablet was used to communicate the advice to the OR staff as well as to gather feedback about the usability of the estimated time. Results: The mean absolute error in estimated procedure time was smaller for the real-time system (14 minutes) than for the OR staff (19 minutes). The system's estimates were more reliable for procedures with average or long duration than for the ones with short duration. For procedures longer than 40 minutes, the mean absolute error was 9 minutes. For these procedures, the system's estimates outperformed the OR staff (mean absolute estimation error of 29 minutes). Conclusion: Computer aided perioperative planning of subsequent procedures on the basis of automated phase identification can support the decision making of the OR team members and of the and managerial staff, thereby improving the quality and throughput of the procedures. Crucial in this respect is that the registered information is consistent with the actual situation and fits the requirements, needs and wishes of the users.