Compliance Check of NHS Digital’s Mental Health Act Statistics

Dear Chris


As you are aware, we have recently conducted our review of the Mental Health Act Statistics – Annual Figures, produced by NHS Digital, against the Code of Practice for Statistics. Mental health awareness has grown considerably in recent years, as has the need to better understand the level of support and quality of services being provided to those most at need. An independent review of the Mental Health Act 1983 was commissioned by the Government for 2018, examining legislation, how the Act is used and how its practices can be improved. Much of the focus of the review was on the quality of existing data concerning the Mental Health Act and the need for greater harmonisation of multiple data sets – providing a clearer picture to policymakers, local decision-makers and the public of how the Act is used and in what context.

I can confirm that the Mental Health Act statistics can remain designated as National Statistics. In reviewing these statistics, we found a number of good examples of relevance to users, clarity and insight and innovation and improvement. These good examples include:

  • The producer team has been transparent regarding the level of data returns from providers of care under the Mental Health Act, whilst acknowledging that completeness of returns for the Mental Health Services Data Set (MHSDS) has been an issue following the transition to the new data collection series. The new data collection series re-uses operational data from service providers, with the previous collection an aggregate data series. The team has been clear that as a result, comparisons of detention rates between the two separate data collection series are not valid at a national level.
  • The publication is written in a simple and informative manner. The accompanying figures are clear and the inclusion of more-detailed breakdowns of patients detained under the Mental Health Act by gender, age and ethnicity allow for greater insight.
  • Providing an easy-read version of the full publication demonstrates that the producer team has considered the needs of a wide range of users.
  • The presentation of the statistics within the Mental Health Data Hub allows users to refine the data based on individual requirements, allowing the statistical bulletin to be presented in a more condensed format.

We have identified some areas where improvements could be made to ensure that the highest standards of National Statistics continue to be met:

  • The producer team has set out a series of plans to improve the statistics. This includes working collaboratively with national bodies to develop a comprehensive data quality improvement plan for the Mental Health Services Data Set, and the establishment of a dedicated data quality team – responsible for identifying and communicating with providers who have either failed to submit or have submitted largely varying data on a month to month basis. The team has also developed specific plans to improve Mental Health Act data within the wider data set. We encourage the team to review the plans on a regular basis to understand their effectiveness, and to amend their plans where necessary.
  • With the producer team set to implement the above changes, we recommend the team publish its plans so that users are informed of future developmentsDoing so will ensure that the team maintains transparency regarding the improvements to the statistics that the users can expect to see.
  • We further encourage the team to integrate statistics on discharge rates following detentions under the Mental Health Act. We understand that this analysis is possible, but that the most appropriate method needs to be determined. We recommend the team works towards identifying the best approach to producing these figures, in line with the findings of the independent review.

We appreciate the team’s cooperation throughout the review process and welcome the plans that it has set out. We also encourage the team to maintain its close working ties with Department for Health and Social Care in supporting the recommendations of the independent review of the Mental Health Act.

Our Health and Social Care team will continue to engage with you and your team in the coming months to follow up on areas that have been highlighted for improvement.

I am copying this letter to Ramesh Notra (Principal Information Analyst, Community and Mental Health Statistics Team, NHS Digital).

Yours sincerely

Mark Pont

Assessment Programme Lead


Compliance check of NHS Digital’s Recorded Dementia Diagnoses Statistics

Dear Chris


We have recently conducted our review of the compliance of NHS Digital’s Recorded Dementia Diagnoses official statistics against the Code of Practice for Statistics. This letter highlights our findings and recommendations.

We recognise the importance of having high quality statistics about dementia diagnoses that serve the public good. NHS Digital’s statistics on Recorded Dementia Diagnoses demonstrate various examples of good practice. The statistics are timely, insightful and the producer team responds well when it identifies new user needs. For example, NHS Digital introduced statistics on patients with a dementia diagnosis and a prescription for anti-psychotic medication at a more granular level after a clear user need was identified.

We have identified three areas where we consider that the team should enhance the value of the statistics:

  • NHS Digital engages regularly with NHS England and the Department of Health and Social Care, for whom this data collection was established. The team also engages regularly with the media to ensure that data are reported accurately. Wider user engagement is limited, however. As you are aware, as part of our review of these statistics we consulted with stakeholders to gain an understanding of who they are used by, how they are used and their overall value. Users from the charity sector identified improvements that they would like to see to the statistics and told us about some unmet needs. NHS Digital should take steps to understand the needs of the broadest possible range of users, including charities, to maximise the public value of these statistics.
  • Our review also highlighted that methodological information such as the calculations for estimated rates of dementia is not always easy to find in supporting documentation, nor is it easy for a non-expert user to understand. We welcome your plans to improve this accessibility and encourage you to seek user feedback on any developments made to ensure that they ease the use of these statistics.
  • Users informed us that they require more-granular data, for example, on subtypes of dementia. NHS Digital should take steps to implement the views received from users, where practicable, to ensure that the statistics and data are as relevant as possible. We encourage NHS Digital to keep an open dialogue with commissioners of the data and push for changes to data collection where a clear user need is identified.

We welcome your team’s motivation to improve these statistics and would like to thank both you and your team for your positive engagement throughout this review process.

Our Health and Social Care domain looks forward to continuing to engage with you and the team on these and other health statistics.

I am copying this letter to Robert Danks (Principal Information Analyst, Primary Care Domain).

Yours sincerely

Mark Pont

Assessment Programme Lead


Response on General Ophthalmic Workforce statistics

Dear Chris

Thank you for your letter of 19 February setting out the situation regarding your General Ophthalmic Workforce statistics. I am grateful to you for publicly consulting with your users and for reflecting on the meaning of National Statistics status in the light of the questions over data completeness, coverage and the forthcoming changes to the source of data for these statistics.

Your letter raised the issue of whether it would be better for these statistics to be published as official statistics, without the National Statistics designation. Designation of official statistics as National Statistics shows that the statistics meet the highest standards of trustworthiness, quality and value. I, therefore, agree with you that the General Ophthalmic Workforce statistics should now be regarded as official statistics.

I welcome your plans to highlight the data quality issues to users when you next publish the statistics, and to continue to work closely with data suppliers and stakeholders to improve the quality and reliability of future publications. I would encourage you to also clearly explain the rationale for the decision to present the statistics as official statistics.

Our domain team for health and social care statistics will keep in touch as the work develops. Please let me know if there is any aspect of this letter that you wish to discuss.

Yours sincerely

Ed Humpherson
Director General for Regulation


Related Links:

Chris Roebuck to Ed Humpherson (February 2020)

Chris Roebuck to Ed Humpherson (March 2017)

General Ophthalmic Workforce statistics

Dear Ed

NHS Digital remains committed to producing high quality Official Statistics, management information and other releases to support the health and social care sectors, and each year we publish around 300 statistical releases across almost 100 series of Official Statistics.

Our long-standing General Ophthalmic statistics relate to Activity and Workforce and we have published these as two distinct reports for several years. The statistics were re-assessed in 2012 and retained the National Statistics designation at that stage.

However, we wrote to you in 2017 about the General Ophthalmic Activity statistics and you supported our proposal that these be released solely as management information rather than as National Statistics.

We have recently undertaken a public consultation to review our General Ophthalmic Statistics and while we recognise that there is a very small user base, nonetheless we believe that these statistics cater for specific and specialist needs that cannot be met elsewhere. As a result, we will continue to produce General Ophthalmic Workforce Statistics for England.

However, we have identified some significant concerns about the data completeness and coverage, which undermine the overall quality of these statistics and will affect the next release. In view of the fact that the National Statistics designation provides assurance to users of the quality, trustworthiness and value of the release, and given that we do not believe the data quality supports such assurance, we would be grateful if you would support our request to remove the National Statistics status from this release.

Over the coming year, the underlying data source for the General Ophthalmic Statistics is transferring to new provider and the entire collection will be undergoing significant change. Although the structure and content of the new data set has not yet been confirmed, we expect to be able to provide richer and more useful statistics once the new collection is properly established.

In the meantime, we will publish General Ophthalmic Workforce Statistics for the 2019 calendar year as a final release using the original data source, highlighting the data quality issues as necessary, and continue to work closely with our data suppliers and stakeholders to improve the quality and reliability of future publications.

Yours sincerely

Chris Roebuck
Chief Statistician


Related Links:

Ed Humpherson to Chris Roebuck (March 2020)

Chris Roebuck to Ed Humpherson (March 2017)

Response from NHS Digital on Adult Social Care in England

Dear Mary,

Thank you for your letter of 16 January. I welcome the observations and findings of your report and share your eagerness to augment the power and influence of data and statistics in Adult Social Care.

You rightly focus on three pertinent themes in your review, which I would like to offer initial responses to in turn:

Leadership and Collaboration. We acknowledge the Adult Social Care system is devolved and that several organisations have data and analysis roles within this system. I am encouraged by the collaborative culture that is developing in our Data Delivery Action Group (DDAG), which brings together a cross-section of such specialists from the Department of Health and Social care (DHSC), NHS Digital, the Association of Directors of Adult Social Services (ADASS), The Local Government Authority (LGA) and the Care and Quality Commission (CQC). ONS colleagues have recently joined this group where we track progress towards statutory obligations and initiatives driven by the English Health Statistics Steering Group (EHSSG) and our Four Nations Group. We remain keen to bolster our links with other relevant peer organisations such as the Ministry of Housing Communities and Local Government (MHCLG), NHS England, Public Health England and NHSX, with the ultimate shared vision of a stronger, representative delivery team to prioritise and address gaps in the provision of data and insight.

Data Gaps. You acknowledge in your report that Key Performance Indicators for Adult Social Care are notoriously difficult to summarise, given the informal and untracked nature of vast areas of care by family and friends, for example. Through our engagement with colleagues in Local Authorities (through our Annual National Workshops – the most recent of which we held in January) we understand the sizeable burden their teams are already under, just to collect currently mandated data in a way that adequately embraces our professional commitments to Quality, Trustworthiness and Value. We remain dedicated to exploring opportunities to augment and, where appropriate, refresh the suite of data collected and analysed in England. We strive to build multi-disciplinary teams at NHS Digital, where we treat collections and publications as ‘customer-facing products’, reducing and / or justifying burden and demonstrating real-world value. When our stakeholders tell us that value is diminishing, we will act swiftly to redirect effort to unmet needs.

We acknowledge the compelling case to keep pace with changes to delivery across the system. We will produce a new Adult Social Care Summary Report this Spring, in response to demand from our stakeholders; a step toward providing proportionate, insightful, cross-cutting summary insights that decision-makers can act on. We also plan to collaborate with other organisations in extending the Client Level Data initiative. By applying the processing power of our new Data Processing Services infrastructure at NHS Digital, we hope to enable users to enhance their ability to link data across health and care settings in a safe, robustly governed way, and to better tailor analytical products and insights to their own local needs.

Existing Statistics. In your letter you assert the need for greater investment in social care data and analysis. In addition, we all have a duty to seek out opportunities for incremental and continuous improvements in our existing products and operations. Not only are we collaborating across organisational boundaries in Adult Social Care, but also between all our experienced Analysis Teams within NHS Digital, to level up our statistical products. Accessibility, Quality and Granularity should feel consistent, regardless of the subject matter the user comes to NHS Digital for. By improving the flexibility and agility of our analysts, we hope to dynamically deploy people where and when they are needed, using more automated processing and publishing technologies to align publication Timeliness with user expectation. Through cross-system collaboration we are committed to aligning Insight to the highest priority needs of the user.

Most recommendations annexed in your letter are already in train, transparently tracked and governed through DDAG. I’d warmly welcome you as a guest to one of their meetings to see how we’re getting on.

I would like to close by returning to Leadership and Collaboration. All organisations in scope of your review have a substantial, exciting and shared opportunity to make a bold difference to Adult Social Care by pushing data (and its utility) to the front of the agenda. I’m personally reassured we all share that vision, passion and urgency. The next sensible step is to coordinate, pool and utilise our expertise and resources for optimal effect and drive visible change, as well as setting out the opportunities that could be fulfilled with additional investment. The workshop you propose is an ideal way to initiate that.

Yours sincerely

Chris Roebuck
Chief Statistician

Related Links:

Mary Gregory to Chris Roebuck (January 2020)

Adult Social Care Statistics in England Report – Letter to Chris Roebuck, NHS Digital

Dear Chris

Today, we have published the findings from our review of Adult Social Care statistics in England. The need for good data to support delivery of adult social care should not be underestimated. While there is rightly a focus on delivery of social care, a scarcity of funding has led to under investment in data and analysis, making it harder for individuals and organisations to make informed decisions. This needs to be addressed if social care is going to evolve to support a changing society and meet the increasing demands expected over coming years. Data matters in solving problems, supporting efficiency and improving outcomes.

Our review identified important improvements needed covering: leadership and collaboration; data gaps; and existing official statistics. We would like to see stronger leadership and collaboration across government to support better data on adult social care and, as the body holding the data and publishing the majority of the statistics, we consider NHS Digital should be taking a leading role in implementing improvements.

As part of our review of existing official statistics, we considered the quality and value of official statistics about adult social care against the standards set out in the Code of Practice for Statistics. It highlighted improvements around accessibility, coherence, quality, timeliness and granularity of the data. The review included outputs published by NHS Digital, immediate actions related to these outputs are outlined in the Annex to this letter.

Improved statistics are essential to support policy makers who are developing proposals to reform the funding and delivery of adult social care as well as individuals who will be able to hold government to account and make better informed decisions about issues which impact the lives of themselves and their families.

We will continue to work with a range of organisations to make the case for improvements to social care statistics. Specifically, our health and social care lead will liaise with you regarding progress towards these recommendations.

I am copying this letter to Sandra Tudor, Head of Profession for Statistics at MHCLG.

Yours sincerely

Mary Gregory
Deputy Director for Regulation


Related Links

Report on Adult Social Care statistics in England (January 2020)

Response from NHS Digital on Adult Social Care in England (February 2020) 

Update on Review of NHS Digital’s Data Access and Sharing Processes

Dear Sarah


We have today published an update of our work looking at NHS Digital’s data access and sharing processes. We recognise that NHS Digital’s Data Access Request Service (DARS) shares our ambition to see data being used more effectively to provide important insights about the population’s health. DARS has introduced significant improvements to help support users to access data, with appropriate safeguards in place to maintain public confidence. As set out in our update, the challenge now is for NHS Digital to communicate these improvements effectively. It will also be important to involve users and other interested organisations in future changes to access arrangements, and to monitor their impact.

The next stage of this work will focus on identifying ways to ensure that the full potential of NHS Digital’s data can be realised. It will address two areas:

  • How can users with innovative or complicated proposals be supported to access health data in England?
  • How can data users and NHS Digital work together to improve data quality and documentation?

We are speaking to organisations with a common interest in achieving these objectives and will report our final recommendations in September.

The positive contribution of your staff throughout this work has been very much appreciated. We look forward to engaging with the DARS team about data access issues on an ongoing basis from now on.

Yours sincerely

Ed Humpherson
Director General for Regulation


Related Links:

NHS Digital data access review update – May 2019

Joining Up Data for Better Statistics, September 2018