The Power of Data Governance: Improving Healthcare with Quality Data
With the introduction of the Government Data Quality Framework in 2020, the commitment of these organisations to understand, assess and optimise their data has become more of a priority than ever. Many organisations have a specialist Data Governance team to assume responsibility for the security, privacy, accuracy and useability of data, but with the advent of data-driven technologies, the potential for transformative change within our health and care system is vast.
The challenge - where to start?
For public sector organisations dealing with high volumes of data across multiple systems, establishing a data quality assessment methodology is invaluable. To do so, a data quality tool that is highly adaptive, flexible, and quick in its analytic capabilities is essential.
By utilising online services such as G-Cloud, these organisations can engage in proof-of-concept activities with data quality software providers, allowing Data Governance teams to assess the capabilities of any new tooling before making a decision on whether to adopt.
Harnessing the power of Data Quality (DQ) tooling
Via G-Cloud, Datactics engaged with an arm’s length body of the Department of Health and Social Care to implement their proprietary data quality assessment tool. In collaboration with the Data Governance team, a diverse sample of data sources were selected, including transactional and highly confidential personal data, varying in size and purpose. Datactics’ tool empowered users to create customised rules for various data quality checks, without requiring coding or SQL expertise. This was a significant advantage to a team constrained by recruitment and salary budgets.
Profiling analysts from the Data Governance team tested the strength of their new DQ tool, ranging from standardisation of patient data formats, to more complex, data source-specific rules, such as matching against third party sources like the Royal Mail’s Postal Address File. This allowed them to validate addresses and ensure patient’s confidential documents were being sent to residential addresses, not businesses.
Rapid Data Quality Assessments
Once the pilot had proven successful, the Data Governance team began to tackle profiling the 70 critical data assets (CDAs) within their remit, performing rapid data quality assessments. To ensure all data assets are being measured effectively, the Data Governance team are working towards establishing a rinse-and-repeat process – a framework to allow users to pinpoint areas requiring data quality improvement and to identify the origins of data quality issues in other business areas, ultimately enhancing decision-making and operational performance across the organisation.
Leveraging Machine Learning capabilities
One standout feature of the tool that would be massively beneficial to the Data Governance team’s in establishing a rapid data quality assessment is Datactics’ innovative Rule Suggestion feature- a Natural Language Processing (NLP) model that reduces the manual effort of building rules. Defining data quality rules for a DQ project is often a time-consuming task. Typically, it takes around 10 hours for a data expert to scope, define, build, test and execute a set of DQ rules. Using NLP technology, which reads data headers and content to suggest DQ rules that should be implemented to a data source, as well as all of the metadata required to build such a rule, Rule Suggestion has the ability to vastly reduce the length of time required to build and implement rules. This feature can rapidly predict relevant basic rules, such as checking for completeness and accuracy of data, allowing the Data Governance team to focus their efforts on building more complex, business-specific rules during the data quality assessments.
The wider impact
A year on from the pilot, the Data Governance team continue to prioritise data profiling and target the most impactful opportunities to drive the value of their new service as they systematically work through the full range of CDAs. In the health and social care sector, the greatest benefits lie in those that carry broader implications for service delivery. Put simply in the government’s Data Quality Framework, “to make better decisions, we need better quality data”. Improved data governance and a deeper understanding of essential data enables more informed decision-making, optimised resource allocation, and more effective interventions. All these aspects, in turn, enhance outcomes for patient safety and incite trust in our public health organisations.
Authors
