Strong CIHR Team presence in upcoming CORS Annual Conference

Former and current CIHR Team members will be presenting at the CORS Annual Conference in Vancouver, May 27-29, 2013.

A Mixed Integer Programming Approach to Improve Operational Scheduling of Radiation Therapists
Vincent Chow, Pablo Santibáñez, Travis Nordin, John French, Martin L. Puterman, Scott Tyldesley
Abstract: Radiation therapists are integral to cancer care treatment delivery and need to be scheduled efficiently. A large staff base and a wide range of assignments make scheduling complex and time consuming. To address this, our team developed metrics to quantify favorable RT schedules and integrated these metrics into a mixed integer programming model to generate new schedules. The model was further packaged into a customized web-based scheduling platform. Currently, this tool is being used on a daily basis by managers at 3 regional cancer centres within British Columbia. Preliminary results show an 83% reduction in schedule creation time, 41% reduction in the time spent managing schedules, and up to 20% improvement in schedule quality. In addition, the tool have enabled managers to forecast future staffing needs, changing the way they approach staff scheduling. Plans to add additional features and rollout to other cancer centres are underway.

A Simulation-based Approach for Dynamic Multi-priority Patient Scheduling
Antoine Sauré, Jonathan Patrick, Martin L. Puterman
Abstract: We present a method to dynamically schedule patients with different priorities to a health care facility in advance of the service date. We model the patient scheduling problem as a Markov decision process and describe a simulation-based algorithm for solving it. The algorithm is based on the use of a policy iteration framework, a post-decision state variable and a non-linear value function approximation. We provide insights regarding the policies obtained through the proposed approach and compare their performance to that of four other policies.

Allocating Home Care Nurses
Claire Ma
Abstract: This paper describes the development and pilot implementation of a computerized staff planning tool for a Home Care Nursing program at a major regional health authority. With in-depth understanding of the challenges in the operations and careful optimization modeling, a user-friendly, Excel-based tool was developed to support the daily Home Care nurse assignment and replace the previous manual assigning. The tool generates reliable assignments by considering factors that are essential to HCN with the ability to reduce the time spent in daily assignment planning by at least 90% at the pilot office. The tool also provides a platform where the health authority could use to standardize the Home Care nurse deployment, which is now governed by individual home health office. Furthermore, it enables systematic data collection and storage for future analysis regarding HCN services delivery. The tool is now running stably at three home health offices.

Planning the Location and Capacity of Healthcare Diagnostic Centres
Emma Liu, Chong Chen, Yue Zhang, Liping Liang, Derek Atkins
Abstract: This paper describes a case study for a medical diagnostic laboratory service provider to model the behavior of patients when choosing a patient service center for their medical tests and to estimate future demand volume. The methodology incorporates long term trending in demand for diagnostic services and identifies the most significant service center attractors. A tool developed based on our model allows the management of the diagnostic services to experiment with locations and capacities for locating or relocating service centers. The management can test the robustness of their decisions by running the model with future demand distribution and competition landscape. Our model is validated and shows good predictive capability. This case study is used to draw a number of lessons for applying these types of models to other similar services in order to assist other applications.

Strategic Location of Radiotherapy Centres in British Columbia
Marc Gaudet, Pablo Santibáñez, Emma Liu, John French, Scott Tyldesley
Abstract: We present a quantitative framework to evaluate current and potential locations for radiotherapy (RT) centres in British Columbia based on geographic access to the population. Record-level data for over 11,000 RT treatments started in 2011 is used to determine geographic access based on driving time to RT centres. An integer programming model is used to find optimal locations and required capacity under different scenarios varying the total number of centres, individual capacity and other clinical and operational considerations. Results show the potential to reduce baseline average travel time by 48% to 64%, and increase baseline coverage within 90 minutes by 13% to 23%. Future work will incorporate other clinical considerations such as volumes by specialty, adjacencies between services and minimum volumes.

Strategic Planning of Health and Related Services in Vancouver's Downtown Eastside
Greg Werker, Michael Krausz, Martin L. Puterman
Abstract: We study a marginalized population for which myriad organizations provide healthcare and other services in the absence of system-level quantitative planning. Using a queueing network we model clients with complex concurrent disorders consuming services in Vancouver’s Downtown Eastside. To analyze this network we present a novel approximation technique—called a linearized closed queueing network (LCQN)—for solving closed queueing networks. This approach uses an open queueing network with the fixed population mean approach as well as an approximation for representing capacitated stations to create a network representation that is solved with a linear program. We derive the approximation ratio between this approximation and the exact solution for a small network, and use simulation to show that this gap is of no practical significance for our model.

The Impact of Modifying Radiation Therapy Appoitment Times on Throughput and Resource Use
John French, Travis Nordin, Vincent Chow, Martin L. Puterman, Scott Tyldesley
Abstract: A simulation model to assess the impact of changing treatment appointment time was developed using data derived from a performance metrics database. The primary metrics were patient wait times for daily treatment, treatment unit utilization and treatment unit finish times. Model results suggested that a change in treatment slot duration from 12 to 10 minutes would result in an earlier unit finish time, an increase in utilization with minimal impact on patient wait times for daily treatment. Based on the modelling the treatment time duration on two treatment units was changed from 12 to 10 minutes and the impact evaluated using data collected from the performance metrics database. Preliminary data shows that the simulation model has been relatively accurate in predicting the impact of system changes.