Government data, science and evidence: health and social care statistics
By Jessica Foster, CSaP Communications Coordinator
As part of the CSaP Policy Fellow seminar series on government’s use of data, science and evidence, Ed Humpherson, Director General at UK Statistics Authority, discussed the Office for Statistics Regulation report: Improving health and social care statistics: lessons learned from the COVID-19 pandemic (Oct 2021).
The report concludes that:
• The COVID-19 pandemic resulted in a huge public appetite for data and statistics.
• It is vital that health and social care statistics command public confidence and enhance public understanding.
• Building on the achievements of the pandemic and overcoming existing challenges will require strong leadership; a commitment to transparency; and sufficient investment, for example in data sharing and linking, data infrastructures, and analytical resource.
“The pandemic reinforced the need for public health data.”
Humpherson began the seminar by explaining how the pandemic demonstrated the importance for public health data. He said the most useful information was a household infection survey by the Office for National Statistics (ONS), which randomly selected households and the level of COVID-19 infection across society. He compared the results to separate data collected by NHS Test and Trace, which he said was an administrative system. As such it was very useful for operational and management purposes, but it was not necessarily the best at understanding a public health phenomenon, due to biases involved, such as the type of people willing to come forward for testing.
Another lesson outlined in the report was the pandemic exposed pre-existing data gaps in the social care system, such as information available from care homes, which Humpherson described as a “neglected area of public data” that has since been corrected. One participant noted that the pandemic reinforced the lack of available data on autism and called for improved methods of data collection across the UK government.
The power of linked data was another lesson discussed during the seminar. For example, how algorithms were able to identify vulnerable people who needed to shield and that different datasets helped to understand the impact of COVID-19 on different ethnicities, to learn about patterns of vaccine hesitancy. There are however many problems with linked data. Humpherson explained that not everyone supports its power and potential; some individuals are not comfortable with GP records being shared and one should never assume public consent is fixed – instead it is context specific. A participant suggested that linked data must be better across Whitehall, in ways that are both useful and impactful in helping to address any data gaps.
“Transparency is king.”
Humpherson’s final remark was the first, and arguably, most crucial lesson outlined in the report from the pandemic - that transparency is key. Statistics and data are not produced simply for the benefit of policy makers; he explained that regardless of policy, data should be accessible to all to help build public confidence, inform behavioural decisions, and avoid public scepticism. Humpherson argued that “the wrong lesson” was learnt from the pandemic; policy makers should not just look at the data and tell everyone what the answer is. However, another potential problem for policy makers, which many participants raised, is who decides what is a public good?
Participants then further discussed the current challenges of health and social care statistics. One suggestion was that a person’s lived experience cannot be reduced to a single data point, therefore qualitative data needs to also be considered. It was recognised that data gathering can be labour intensive and further automation tools could, perhaps, be introduced to make the process easier. There were also calls to support developing countries as data collection is dependent on a country’s infrastructure and capacities, which can ultimately lead to delays and data gaps. For example, it was raised that in Lebanon, COVID-19 data was based on the origin of a person, rather than their place of residence, which meant they could not zone or track levels of infection. One solution raised was to invest in daily methods of data collection, by working with global social media giants such as Google or Apple, who collect huge amount of data every day.
Ultimately, the COVID-19 pandemic highlighted the crucial role the public play in health and social care statistics. While there are strong arguments for data to be readily available to the public as well as to policy makers, actions need to be taken to ensure there are no data gaps and public consent for linked information is kept up to date. One participant concluded that it was “striking” we needed a pandemic to learn lessons on data, and suggested efforts should now be focussed on improving the data itself.