Guidance

Quality management approach in official statistics

Updated 1 April 2026

Summary

When we assess the quality of our official statistics products, we:

• consider how fit for purpose they are
• use the 5 dimensions of quality to consider the quality of our official statistics in a way which is relevant to user needs
• aim to cultivate a ‘quality assurance (QA) mindset’ in the production of our official statistics, ensuring that QA is considered throughout the production process, not only at the end
• aim to document the quality of all our official statistics products in QA logs
• aim to publish quality and methodology information (QMI) reports alongside all of our official statistics products
• have self-assessed that we are on a journey to meet all of these standards, but not there yet

Purpose and scope

This document sets out UKHSA’s approach to the quality of our official statistics, one of the 3 core principles of the Code of Practice for Statistics.

You can find a full list of our official statistics products on GOV.UK. While all producers of our statistics take care that their work is high-quality, this specific document aims to reflect the level of transparency and data quality which we work towards in our analytical products which have attained official statistics status.

This approach is aspirational and reflects the standards we strive for in the production of our official statistics.

Our approach to quality

When we refer to high-quality analysis, we mean analysis which is fit for the purpose it was commissioned to meet. This means statistical outputs which are unlikely to be misleading as they are accurate, assured, evidenced, proportionate, adequately communicated, and documented.

We consider the quality of our official statistics in accordance with the European Statistical System’s 5 dimensions of quality, as follows:

1. Relevance, meaning the degree to which the statistics meet user needs in both coverage and content.
2. Accuracy and reliability, meaning the proximity between an estimate and an unknown true value.
3. Timeliness and punctuality, meaning the time gap between publication and the reference period.
4. Accessibility and clarity, meaning the ease with which users can access the data.
5. Coherence and comparability, meaning the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar.

While these 5 dimensions overlap, you cannot achieve all of them at their highest level at the same time. For example, it takes time to produce analysis which is as accurate and coherent as possible, so there will have to be a trade-off with timeliness in order to achieve this. When making trade-offs, we consider user needs and prioritise the data quality dimensions that align with those.

We aim to cultivate a culture of quality statistics in UKHSA, through promoting communities of practice and ensuring that statistics producers have access to opportunities for training and for refreshing their knowledge on the latest developments in best practice.

We work closely with other quality assurance stakeholders including the team that produce UKHSA’s risk-based analytical quality assurance framework, which also signposts our quality resources.

We also work to our reproducible analytical pipelines (RAP) implementation plan to improve uptake of RAP in UKHSA, improving the quality and trustworthiness of our official statistics.

Quality assurance

Analytical quality assurance (QA) is the process and set of practices that ensure analysis is fit for purpose, and it plays an essential part in ensuring that our analysis meets high standards of quality. Effective QA includes checking code and outputs but is much more than this too – projects must have a credible end-to-end approach to demonstrate outputs are fit for purpose, and ensure remaining uncertainty and risks are communicated.

The nature and extent of the quality assurance carried out for an official statistics product depend on what is considered appropriate and proportionate for each individual product.

We operate QA throughout the entire data journey, as it is not something which can be simply added at the end of production. This means that we consider QA at each stage of statistics production, from planning through to producing final outputs and communicating results. We consider data sources and their appropriateness, whether our methods are sound, and if the interpretation of results is robust, communicating any caveats that need sharing with users. This is known as having a ‘QA mindset’ and this is the approach we aim to cultivate in UKHSA, alongside a culture of continuous improvement.

To help statistical production teams to record and sign-off on quality, we advise producers of statistics, especially official statistics, to use QA logs, a standard practice across government statistics which helps to ensure that a QA mindset is present when producing analysis. We encourage our official statistics producers to complete a QA log alongside every publication. All of our official statistics producers either currently use a QA log or are working to develop a QA log specific to their production pipeline.

Quality and methodology information reports

Quality and methodology information (QMI) reports provide detailed information about the data sources and methods used to produce statistics and the quality of those statistics. For all of our official statistics products, we either already publish an accompanying QMI report or are working to develop one.

Providing QMI reports ensures that we are fully transparent about how our statistics are produced with all of our users, the general public, other experts, and the Office for Statistics Regulation. It not only enables greater trust in our statistics but also ensures all relevant details are available to serve user needs.

Our main official statistics publications all include brief information about data, quality, and methods, but only to the extent necessary to understand the statistics, to prevent these main documents from becoming too technical or confusing. The separate QMI reports are then available for any users who would find the additional information useful or interesting. For example, here is the QMI report for our healthcare-associated infections (HCAI) statistics.

Our QMI reports cover:

1. Data sources, including their:

   • quality
   • strengths and limitations
   • accuracy
   • completeness
   • uniqueness
   • consistency
   • timeliness
   • validity

2. Best available methods.

3. Quality summary, including:

   • confidentiality and disclosure control
   • geography
   • the 5 dimensions of quality; relevance, accuracy and reliability, timeliness and punctuality, accessibility and clarity, and coherence and comparability
   • any description of how the statistics were produced and quality assured

4. Uses and users.

5. Related statistics.