Modeling and improving chemotherapy patient flow.

Abstract

175 Background: Chemotherapy administration scheduling is dependent on infusion room hours of operation, availability of oncologist, capacity for treatment, pharmacy hood time, nursing and pharmacy staffing, and physical space limitations. In 2013, the main infusion center at MGH Cancer Center had 45% room/chair utilization reported by the Ambulatory Patient Tracking System and 33% exam room utilization in the clinics. However our infusion center experienced extremely high volume during peak hours, 10am to 2pm, but was underutilized before 10am and after 2pm, making it difficult to add on additional patients. Methods: In November 2013, MGH Cancer Center began collaboration with MIT Operations Management experts to ultimately improve patient flow through the Cancer Center by flattening the bottleneck at peak hours of operation and improving utilization of the infusion area and clinics. In phase I of the project, data was reviewed and refined. Key role groups of schedulers, prescribers, pharmacists and infusion nurses were shadowed by collaborators to gain insight into complex scheduling practices. A multidisciplinary working group met weekly to discuss the progress and suggest areas for further investigation. In phase II, optimization models testing the impact of alternative scheduling practices, physical space changes and alternative clinic configurations will be created. Finally, in phase III, change implementation and measurement will take place. Results: Process flow maps of patient movement through the cancer center were created. Patient tracking data was manipulated to understand key operational metrics. Several insights include an overall same-day chemotherapy cancellation rate of 10.7%, with the majority of cancellations from thoracic, GI and GU disease centers. Our mean scheduled infusion treatment length was 2.13 hours and 25% of appointments booked into infusion are not linked with a same-day clinic appointment. Conclusions: Understanding and refining incomplete or problematic data was a key part of understanding the issues contributing to the middle of the day bottleneck in the infusion area. Future work on this project will include optimization modeling and change implementation.