Exploring Cannulation Process in Chemotherapy through a Computer Simulation

The aim of this study is twofold. Firstly, to demonstrate how combining computer simulation, data from multiple data sources, and statistical methods, can extend the understanding of the issues associated with process modelling and analysis in healthcare environment, and therefore contribute to improvements in resource utilisation and safety in hospitals. Secondly, to provide simple re-useable methodology for cross-validation of multiple data-sources such as interviews, hospital IT data management systems and simulation results. The insights from this study are threefold. Firstly, the accuracy of the estimates of duration of cannulation obtained through the interviews with the nurses and the chemotherapy unit manager is very high. Secondly, although the duration estimates were precise, the process descriptions obtained through interviews with nurses were oversimplified or incomplete and therefore did not realistically reflect complexity of a medical process with a significant number of relatively rarely occurring exceptions. Thirdly, by combining multiple data-sources it is possible to reduce costs associated with observation as a most expensive data-capturing approach. A detailed exposure of the methodology including step-by-step description is provided to facilitate conducting similar research in hospitals in the future.

Chemotherapy is the use of anti-cancer drugs to destroy cancer cells. In Australia, chemotherapy is usually administered in chemotherapy treatment units (outpatient oncology clinics). In this paper we focus on intravenous (IV) cannulation, referred as cannulation in the rest of the paper, which is an important sub-process that is often a part of chemotherapy process. The cannulation is a technique in which a cannula is placed inside a vein to provide venous access. Venous access allows sampling of blood as well as administration of fluids, medications, parenteral nutrition, chemotherapy, and blood products.

Chemotherapy consumes significant resources, therefore an assessment of the utilisation and productivity of such resources is important. Here, we consider utilisation of resources associated with cannulation, because we identified cannulation as the source of significant variation in chemotherapy nursing resource utilisation. This study was undertaken as a part of a larger project that has objective to improve the performance of a chemotherapy treatment unit by increasing the throughput and reducing the average patient’s waiting time. Similar work undertaken in Canada and supported by CancerCare Manitoba, with a focus on development of a scheduling template for chemotherapy treatment, was described by Ahmed.

The Basic Treatment Equivalent (BTE) model was developed in radiotherapy to measure complexity and treatment duration differences between external beam radiotherapy treatments. It was derived from the time measurements of radiation treatment fractions and then developed from mathematical modelling of the data. It was shown to be a more sensitive productivity measure than fields per hour when tested in radiation oncology departments in New South Wales and other locations in Australia and New Zealand. Further enhancements of the model have subsequently occurred.

Methodology and Input data

The input data used in this study comes from four different sources; interviews with nurses and a chemotherapy treatment units manager, electronic sources such as ARIA (an oncology information system), samples collected by direct observations and the simulation itself. The data is processed in R, open source programming language and software environment for statistical computing, using the “data.table” and “ggplot2” libraries. The results are provided in accessible, easy to understand, visual format that allows for interactive exploration, comparison and validation.

The examined data from ARIA contained 19,937 rows, each representing a patient visits, of which 9,210 (46.2%) were rows without cannulation attempt data. We are unable to provide the snapshot of this table due to its size. The descriptive statistics for number of cannulation attempts is provided in Tables I and II, as a function of sex and year in which the cannulation occurred.

Process simulation in AnyLogic

The visual process definition as depicted in AnyLogic simulation software

Related posts