Mr Philip Rawson-Harris, Denise Delrosario-Kelly1, Megan Gritt1, Jaimie Yamada1, Shirley Leong1, Pauline Bass1, Dr Andrew Stewardson1,2
1Alfred Health, Australia, 2Monash University, Melbourne, Australia
Biography:
Philip heads the epidemiology and data management service at Alfred Health for Infectious Diseases. He has a background in pharmaceuticals and clinical trials, and a Masters in Control of Infectious Diseases. He has an interest in improving health through data visualisation and audit. Outside work, he loves sport and rugby.
Abstract:
Introduction
In 2021, Infection Prevention and the Epidemiology and Data Management Service presented the then unnamed TRACET or Transmission Risk Area Classification and Exposure Tracker tool at ACIPC. This tool was an interactive Gantt chart with the ability to visualise potential crossover events on wards, filter based on genotyping and display further information about the specimen and patient. The tool, however, was locally stored, required excel data loads and significant setup and was therefore limited in its use.
Methods
A custom Power BI visualisation was generated using the code from the R tool. SQL scripts were used to pull the specimen data from the Carbapenamase Producing Organism (CPO) database and ward movement from the data warehouse. The tool takes only 6 fields as input to feed the custom visual. The structured storage of the CPO database also allowed previously filterable attributes including resistance type, gene and cluster.
Results
The migration of the tool onto the central data infrastructure, automated scripts and a standardised CPO database structure has created a deployable tool with daily refresh, accessible to all (with appropriate permissions) and removal of the EDMS from support. The simplified tool allows for deployment elsewhere where a data warehouse exists and can be hosted on a microsoft dashboard system.
Conclusion
The movement of the tool to an R powered custom Power BI visualisation with simple input and automated SQL scripts has resulted in an interactive Gantt tool that removes hefty data setup and has more potential applications than ward crossover alone.