Dear Lewis,


Following our written evidence submitted to your Committee in February, I wish to offer further consideration and view on follow-up questions suggested to us for answer.

1. Measuring individuals’ outcomes, and outcomes associated with the integration of health and social care in Scotland. How can outcomes evaluation and measurement be implemented so that it is statistically sound and useful?

All public bodies who are involved in the production of official statistics should adhere to the principles set out Code of Practice for Statistics (the Code). Compliance with the Code ensures that statistics are of public value, are of high quality and are produced by departments and public bodies that can be trusted.

We note that the 31 Integration Authorities are not official statistics producers, and therefore there is no statutory requirement for them to comply with the Code. Scottish Government and Public Health Scotland (PHS) are official statistics producers. This means that, whilst the data sources for social care statistics are provided by bodies and organisations such as health and social care partnerships, local authorities and third sector organisations, statistics based on this data should comply with the Code.

In our February 2020 report, Adult Social Care Statistics in Scotland, we highlighted that, at the time of writing, fundamental gaps existed in social care statistics in Scotland which meant that they were not currently providing the range and depth of information needed to fully serve the public good. We cited as an example that a lack of information about outcomes for people who use social care was one of the most common frustrations we heard whilst conducting our report research. Users of social care statistics told us that too much emphasis is placed on counting system outputs, such as the number of hours of care delivered, costs of services and numbers of staff, and would instead like to know more about the outcomes achieved for the people using those services.

For outcome evaluation and measurement data to be published as official statistics, this will require that they are developed by an official statistic producing body in line with the Code. In our view, this will likely require some form of standardised data collection across the 31 Integration Authorities. Recent examples of such standardised data collection include those developed for the Carers Census and the new PHS adult social care data collection system.

In our report we made the following recommendation: ‘All social care statistics producers need to work together, in consultation with health and social care partnerships and statistics users, to identify and prioritise actions to address social care data gaps – including by making better use of existing data – and meet users’ information needs.’

In addition to the development of social care statistics, we recognise there may be a need for organisations delivering social care services to develop operational management information on service outcomes to inform commissioning and delivery of social care services. As these organisations are not official statistics producers, this type of information is out of scope of our formal regulatory remit. Notwithstanding this, in 2018 we introduced the ability for organisations who are not official statistics producers to sign up to voluntary application of the Code. This option is available to any producer of data, statistics and analysis which are not official statistics, whether inside government or beyond, to help them produce analytical outputs that are high quality, useful for supporting decisions, and well respected. A commitment to the Code pillars of Trustworthiness, Quality and Value offers the opportunity for an organisation to:

  • Compare its processes, methods and outputs against the recognised standards that the Code requires of official statistics.
  • Demonstrate to the public its commitment to trustworthiness, quality and public value.

Outcomes evaluation and measurement is not a challenge unique to Scotland. In our January 2020 report, Adult Social Care Statistics in England, we outlined that reliable and comprehensive evidence is vital for evaluating delivery and informing policy decisions which can lead to improved outcomes and support individual choice.

There are two aspects to this challenge.

  • Definitions: In the health context, there is a well-developed approach to comparing different interventions based on their impact on the life of the patient – known as the Quality Adjusted Life Years (QALYs). There is not yet a comprehensive framework for thinking about and defining outcomes for social care interventions.
  • Cost effectiveness: In our England report, we highlighted that unlike health, where the effectiveness of interventions is a priority research area, in social care there is very little understanding of the most cost-effective intervention and what the impact of each intervention is. We strongly encouraged the implementation of joined up data across health and social care in England to understand how the two systems interact and what drives the best outcomes.

2. Professor David Bell, University of Stirling, highlighted in his evidence to the Committee the lack of data collection in Scotland in comparison with other parts of the UK. Currently, Scottish researchers rely on English statistics used for projecting demand. Do you have a view on this, particularly in relation to the policy divergence between the two health and care systems?

With regards to the specific issue that Scottish researchers currently rely on English statistics used for projecting demand, we do not have sufficient evidence to make a judgement on this.

However, in our February 2020 report, Adult Social Care Statistics in Scotland, we noted that many researchers are keen to make more use of health and social care data. We welcomed the creation of Research Data Scotland, which we hope will help address the data access issues that researchers currently face when seeking health, social care and other data, and support greater joining up of these data. In addition, in our report we recommended that PHS and Scottish Government should convene a social care data user summit in 2020 to help inform Research Data Scotland’s development and PHS’s plans for making more use of linked health and social care data.

Unfortunately, due to the restrictions imposed due to the pandemic, this summit has not yet been convened. We are in regular discussion with both PHS and the Scottish Government and are keen for this recommendation to be realised.

3. What approach(es) to data collection do you think need to be considered and what data do you feel is required?

As part of our research for our February 2020 report, Adult Social Care Statistics in Scotland, we spoke to organisations who are data providers for the current social care statistics. We highlighted the following important issues that require careful consideration as part of any new or amended official statistics data collection system.

  • Resourcing issues beset all aspects of social care data collection and statistics production. This includes the availability of staff to collect data and return it, the need for investment to improve its quality, the need for entirely new forms of data to be collected to better meet user needs, and the availability and capability of staff to use the data themselves to inform service development locally. We recognise that the resource implications associated with building new data systems are far greater than those associated with improving existing statistics. The drivers and funding to do this will also be largely beyond the reach of statistics producers alone.
  • Building data collection systems that deliver value to staff and users of social care is also difficult to do without imposing unreasonable administrative burdens. A significant amount of social care activity takes place beyond the scope of public sector service settings and the majority of the social care workforce (around 70%) is employed by private and third-sector providers. This makes the task of building routine data collection systems significantly harder.

It is our view that Scotland is not alone in facing these challenges. Our work in this sector as a UK wide regulator has identified similar difficulties with collecting data in disparate settings about human experiences (as opposed to flows of money or service provision).

Whilst we cannot be specific around what data is required in this case, we do expect that, in line with the Code, users of statistics and data should be at the centre of statistical production. Understanding user needs and seeking the views of users is important and should be used to direct what data is required. Official statistic producers should establish an ongoing dialogue with users to ensure that statistics continue to meet changing user needs and demand.

The COVID-19 pandemic has emphasised the importance of responding to user need and has brought attention to existing gaps in adult social care statistics.

I hope this is useful to the Committee.

Yours sincerely
Ed Humpherson
Director General for Regulation