Dear Scott
CARE HOME CENSUS FOR ADULTS IN SCOTLAND STATISTICS
As you are aware we recently carried out a short review of compliance with the Code of Practice for Statistics of the Care Home Census for Adults in Scotland statistics as part of our ongoing programme of compliance checks and to support our systemic review of adult social care statistics in the UK.
We met with your team who shared their aspirations to improve these statistics to ensure they continue to meet the standards expected of their designation as National Statistics. Here are our main findings of our review, based around the three pillars of the Code.
Trustworthiness:
- We welcome the fact that the privacy impact assessment (PIA) for previous waves of the Care Home Census is easily accessible from ISD’s website. The easy read version for care home residents is also a clear example of good practice, demonstrating how information about data privacy can be tailored to meet different populations’ needs. We understand that your plans for providing this kind of information have changed following the introduction of GDPR. As outlined in our regulatory guidance on data governance, we expect statistics producers to provide transparent information about data safeguarding to help build trustworthiness. We think the Five Safes are a helpful way to present this information.
Quality:
- The data undergo a series of appropriate validation checks, both automated and manual, and these are detailed in the quality statement which is readily accessible from the ISD website. However, there is limited information about the response rate, levels of missing data, and any associated implications for the statistics and interpretations. This information would help users understand how well the statistics represent the care home population, and could be used by ISD to help target non-responders.
Value:
- The timeliness of these statistics is not acceptable. The statistics for financial year 2016/17 were published in October 2018, over a year later than scheduled, while the 2017/8 data face a similar length delay. We know from conversations with users that this time-lag does not meet their needs. We understand that the delay was caused by the need for new information governance arrangements to be implemented following the introduction of GDPR. This required agreement between ISD, Scottish Government and the Care Commission and staffing resources were stretched. We understand that a new data sharing agreement (DSA) will be in place shortly that will enable you to return to producing the statistics in a more timely way. We have suggested Scottish Government review this process to help identify any useful learning. Given these delays, we also suggest you consider publishing the 2017/18 and 2018/19 data in the same release this autumn, rather than producing two in quick succession.
- We share your aspirations to provide greater insights for readers. More commentary explaining what the statistics mean, what might help explain any trends or patterns in the data, and information about the wider social care policy context would be a valuable addition to the next release. We know from speaking to users that the information about funding is valuable and could be enhanced. We have shared additional feedback about the publication with your team and we encourage them to consider the suggestions.
- We agree with you that it is time to work with users to review the contents of the census and how its results are presented. Users’ views about the Care Home Census were last canvassed in detail in 2012. Many changes to social care policy and provision have taken place since then, and a broader transformation of the social care statistics landscape is currently underway. We expect that any changes resulting from this process to be effective from the 2020 census date.
We are content for these statistics to retain their National Statistics designation, provided two conditions are met. Firstly, the DSA needs to be signed in time for the 2018/19 statistics to be published this autumn (alongside the delayed 2017/18 statistics). Secondly, the short-term improvements your team has already identified, including a greater focus on providing useful insights, need to be implemented for the next publication. Our Health and Social Care co-lead, Catherine Bromley, has agreed with your team that they will continue to engage with each other on progress over the coming months.
In the longer-term, we would like to see producers of social care statistics in Scotland working collectively with users to identify more innovative, informative and coherent ways to present statistics. The production of multiple outputs across different producers, each with a slightly different focus on an aspect of social care, does not maximise the value of these statistics and meet user needs. We will continue to develop this recommendation as part of our ongoing systemic review of adult social care statistics. We look forward to working with you on this. I am copying this letter to Roger Halliday, Chief Statistician and Data Officer, and Angela Campbell, Deputy Director, Health and Social Care Analysis, Scottish Government.
Yours sincerely,
Mark Pont
Assessment Programme Lead
Related Links:
‘What you told us about the statistics on adult social care across the UK’, 5 February 2019