Modelling “Busyness” - Using discrete event simulation and soft systems methodology to create a multimethod decision support tool for radiology
Date
2022-01-13Author
Conlon, Mary
Metadata
Show full item recordUsage
This item's downloads: 271 (view details)
Abstract
In line with an increase in the incidence of chronic disease and an aging demographic, demand for radiology services is growing year on year. Radiology staff combine patient care and clinical and technical skills to meet this demand. Radiology workload has historically been measured in terms of the number of examinations completed, without regard to the patient characteristics or staff experience of workload. The motivation behind the work was to create a model which measured staff workload inclusive of the patient population profile. The novelty of the work lies in the richness of outputs obtained using hard and soft modelling OR methods, while contributing to the literature on staff and radiology workload.
This work describes the design and application of a framework for modelling aspects of a Computed Tomography (CT) service. The dual propose of the work is to provide insights into the service and staff workload by capturing process metric and to provide decision support to address the problem of an increasing waiting list. The framework was designed to facilitate high stakeholder involvement using soft systems tools to identify the components of, and scenarios for use of a discrete event simulation (DES) model.
The action research framework was validated by application in a CT scanner radiology department. A conceptual group model building approach was taken using System Dynamics (SD) notation which identified the patient characteristics of age, infection status, mobility and examination type as desirable model components. Using soft systems methodology (SSM), a rich picture of the service was created, and three scenarios identified by local decision makers. The outputs for each scenario in terms of waiting list evolution and process metrics such as resource utilisation, process delays and reliance on flexible staff were obtained from the validated model.
Using DES, it was demonstrated that mixing inpatient and outpatient services results in significant variation in demand and utilisation of resources. Radiology workload in terms of staff time and process perturbations were shown to be greater for inpatients. In the case of non-contrast exams inpatients were found to consumed 127% more staff time than outpatients. A simulation of an outpatient only service demonstrated that radiographer utilisation was less despite a greater average number of patients being scanned. A recommendation was made to separate the services to increase outpatient capacity. Future radiology workload metrics should include inpatient and outpatient population characteristics. The framework was endorsed by the case study department and future applications in other areas such as ultrasound were identified. Action research changes resulted from the work.