Identifying primary care datasets and perspectives on their secondary use: a survey of Australian data users and custodians

Abstract

Abstract Background Most people receive most of their health care in in Australia in primary care, yet researchers and policymakers have limited access to resulting clinical data. Widening access to primary care data and linking it with hospital or other data can contribute to research informing policy and provision of services and care; however, limitations of primary care data and barriers to access curtail its use. The Australian Health Research Alliance (AHRA) is seeking to build capacity in data-driven healthcare improvement; this study formed part of its workplan. Methods The study aimed to build capacity for data driven healthcare improvement through identifying primary care datasets in Australia available for secondary use and understand data quality frameworks being applied to them, and factors affecting national capacity for secondary use of primary care data from the perspectives of data custodians and users. Purposive and snowball sampling were used to disseminate a questionnaire and respondents were invited to contribute additional information via semi-structured interviews. Results Sixty-two respondents collectively named 106 datasets from eclectic sources, indicating a broad conceptualisation of what a primary care dataset available for secondary use is. The datasets were generated from multiple clinical software systems, using different data extraction tools, resulting in non-standardised data structures. Use of non-standard data quality frameworks were described by two-thirds of data custodians. Building trust between citizens, clinicians, third party data custodians and data end-users was considered by many to be a key enabler to improve primary care data quality and efficiencies related to secondary use. Trust building qualities included meaningful stakeholder engagement, transparency, strong leadership, shared vision, robust data security and data privacy protection. Resources to improve capacity for primary care data access and use were sought for data collection tool improvements, workforce upskilling and education, incentivising data collection and making data access more affordable. Conclusions The large number of identified Australian primary care related datasets suggests duplication of labour related to data collection, preparation and utilisation. Benefits of secondary use of primary care data were many, and strong national leadership is required to reach consensus on how to address limitations and barriers, for example accreditation of EMR clinical software systems and the adoption of agreed data and quality standards at all stages of the clinical and research data-use lifecycle. The study informed the workplan of AHRA’s Transformational Data Collaboration to improve partner engagement and use of clinical data for research.

Keywords

• Australia
• Data curation
• Data linkage
• Data management
• Data quality
• Primary health care data