1. Introduction to the Quality Assurance Framework for Data Linkage (QAFDL)

1.1 Purpose of QAFDL

Data linkage is not a by-product of analysis, but a critical modelling step requiring an appropriate level of quality assurance. The objective of understanding and consolidating best practices in data linkage led to the creation of the Quality Assurance Framework.

The Quality Assurance Framework for Data Linkage acts as a guide designed to provide a structured approach to managing and assuring the quality of data linkage processes in NHS England, but we welcome collaboration from other organisations.

The QAFDL framework offers support for all stages of data linkage, from discovery to delivery. It is advised that the framework is consulted regularly to ensure that the quality assurance principles it contains are included in the project from inception. We view the QAFDL as a 'living' framework which we will iterate over time as it is implemented into our buisness processes.

The broad aims are:

• To equip and empower practitioners with the means to decide the appropriate levels of quality assurance.

• Function as a tracking instrument, enabling comprehensive evaluation of all dimensions described by the framework and facilitating the assessment of data linkage projects.

The QAFDL is based on three principles, linkage quality, transparency and safety. It can help boosting users' trust in the validity of the linked data for their analysis or research. It guides linkers and users on managing and interpreting linked data effectively, providing insights into the limitations of their linked data assets, enabling well-informed decisions and robust analyses. Moreover, the QAFDL promotes ethical and responsible use of linked data by suggesting best practices for data privacy and governance.

1.2 The 5W's of the QAFDL

1.3 Overview of the Quality Assurance Framework dimensions

Over 4 key dimensions, the QAFDL outlines best practices, principles, and strategies to facilitate high-quality data linkage.

1.4 Definition of data linkage

Data linkage refers to the process of connecting different datasets to create a more comprehensive view of an entity. An entity can be a patient, events, organisations, etc.

Data linkage is essential for healthcare research, statistics, and direct care, as it enables seamless integration of information and greatly impacting the quality of healthcare outcomes.

1.5 Benefits of data linkage

The NHS, as stated on the Long Term Plan, aims to use data more effectively across health and social care services, for both individual care and for planning and research purposes and this can be done through data linkage. Other benefits are:

• Enhanced Insights: By combining data from different data sources, more in-depth information can be learnt about specific people, groups of people, or occurrences. Linkage also helps policy makers in the development, and evaluation of policies and programmes.

• Longitudinal Analysis and research opportunities: Data linkage makes it possible to follow people over time, enabling longitudinal research that can more accurately highlight trends and patterns. Also, opens new possibilities for research that would not be feasible with isolated datasets.

• Risk Management: By spotting trends and correlations that might otherwise go unnoticed in isolated datasets, linked data can help organisations foresee and address possible problems.

• Reduced burden in providers: Data linkage reduces the need to collect the same information multiple times, thereby reducing the burden on providers. Data linkage can also reduce costs associated with duplicate efforts in data collection.

• Improved Data Quality: By facilitating the comparison of data across many sources, linkage enables the identification and correction of discrepancies or errors.

• Reducing Bias: When data from different sources is linked, it can help reduce selection bias in research studies, leading to more reliable findings.

• Service Delivery: In healthcare and social services, data linkage can improve service delivery by providing a more holistic view of a person's needs and interactions with different services.

To find out more about what the team is up to, visit our NHS England data science website.