Responding to COVID-19 with real-time general practice data in Australia.

INTRODUCTION: As SARS-CoV-2 spread around the world, Australia was no exception. Part of the Australian response was a robust primary care approach, involving changes to care models (including telehealth) and the widespread use of data to inform the changes. This paper outlines how a large primary care database responded to provide real-time data to inform policy and practice. Simply extracting the data is not sufficient. Understanding the data is. The POpulation Level Analysis and Reporting (POLAR) program is designed to use GP data for multiple objectives and is built on a pre-existing engagement framework established over a fifteen-year period. Initially developed to provide QA activities for general practices and population level data for General Practice support organisations, the POLAR platform has demonstrated the critical ability to design and deploy real-time data analytics solutions during the COVID-19 pandemic for a variety of stakeholders including state and federal government agencies. METHODS: The system extracts and processes data from over 1,300 general practices daily. Data is de-identified at the point of collection and encrypted before transfer. Data cleaning for analysis uses a variety of techniques, including Natural Language Processing and coding of free text information. The curated dataset is then distilled into several analytic solutions designed to address specific areas of investigation of interest to various stakeholders. One such analytic solution was a model we created that used multiple data inputs to rank patient geographic areas by the likelihood of a COVID-19 outbreak. The model utilised pathology ordering, COVID-19 related diagnoses, indication of COVID-19 related concern (via progress notes) and also incorporated state based actual confirmed case figures. RESULTS: Using the methods described, we were able to deliver real-time data feeds to practices, Primary Health Networks (PHN) and other agencies. In addition, we developed a COVID-19 geographic risk stratification based on local government areas (LGAs) to pro-actively inform the primary care response. Providing PHNs with a list of geographic priority hotspots allowed for better targeting and response of Personal Protective Equipment allocation and pop-up clinic placement. CONCLUSIONS: The program summarised here demonstrates the ability of a well-designed system underpinned by accurate and reliable data, to respond in real-time to a rapidly evolving public health emergency in a way which supports and enhances the health system response.

Coding and classifying GP data: the POLAR project.

BACKGROUND: Data, particularly 'big' data are increasingly being used for research in health. Using data from electronic medical records optimally requires coded data, but not all systems produce coded data. OBJECTIVE: To design a suitable, accurate method for converting large volumes of narrative diagnoses from Australian general practice records to codify them into SNOMED-CT-AU. Such codification will make them clinically useful for aggregation for population health and research purposes. METHOD: The developed method consisted of using natural language processing to automatically code the texts, followed by a manual process to correct codes and subsequent natural language processing re-computation. These steps were repeated for four iterations until 95% of the records were coded. The coded data were then aggregated into classes considered to be useful for population health analytics. RESULTS: Coding the data effectively covered 95% of the corpus. Problems with the use of SNOMED CT-AU were identified and protocols for creating consistent coding were created. These protocols can be used to guide further development of SNOMED CT-AU (SCT). The coded values will be immensely useful for the development of population health analytics for Australia, and the lessons learnt applicable elsewhere.

What a Comprehensive, Integrated Data Strategy Looks Like: The Population Level Analysis and Reporting (POLAR) Program.

In Australia, general practice (GP) acts as the gatekeeper to the rest of the healthcare system, and therefore the vast majority of the population have an electronic medical record. It follows that the largest database of the population is therefore on the distributed GP computers. Informed by a comprehensive system-wide data strategy, the Population Level Analysis and Reporting program extracts data from the GP electronic medical records and repurposes it for multiple uses. The program requires the data to be coded and then structured for multiple uses clinical care, clinical governance, research, and policy.

POLAR Diversion: Using General Practice Data to Calculate Risk of Emergency Department Presentation at the Time of Consultation.

OBJECTIVE: This project examined and produced a general practice (GP) based decision support tool (DST), namely POLAR Diversion, to predict a patient's risk of emergency department (ED) presentation. The tool was built using both GP/family practice and ED data, but is designed to operate on GP data alone. METHODS: GP data from 50 practices during a defined time frame were linked with three local EDs. Linked data and data mapping were used to develop a machine learning DST to determine a range of variables that, in combination, led to predictive patient ED presentation risk scores. Thirteen percent of the GP data was kept as a control group and used to validate the tool. RESULTS: The algorithm performed best in predicting the risk of attending ED within the 30-day time category, and also in the no ED attendance tests, suggesting few false positives. At 0 to 30 days the positive predictive value (PPV) was 74%, with a sensitivity/recall of 68%. Non-ED attendance had a PPV of 82% and sensitivity/recall of 96%. CONCLUSION: Findings indicate that the POLAR Diversion algorithm performed better than previously developed tools, particularly in the 0 to 30 day time category. Its utility increases because of it being based on the data within the GP system alone, with the ability to create real-time "in consultation" warnings. The tool will be deployed across GPs in Australia, allowing us to assess the clinical utility, and data quality needs in further iterations.