22nd January 2019

Session leader Katya Bozukova who works at the Lincolnshire Open Research and Innovation Centre, a university-based team that works with charities and SMEs on data driven innovation, and on projects that address the challenges posed by an ageing population.

Her organisation wants to publish open data around its work: “but we now need to do it in an ethical way.” In the meantime, she would like to know what data sets are already out there, that her organisation might use. What, she asked, were other session members hoping to get out of the session?

One participant, who volunteers with the NHS in Glasgow, said he was interested in inequality. Marius from Bristol City Council said it was very interested in health and social care and was planning to publish an inventory of open data sets; he was interested to link-up and see what is going on.

Jerry said he interested because he used to work with EU on health informatics policy and has COPD and so uses health services. A participant called Liz from Oxford University said she looks after students with health and social problems said the data it used tended to be closed data, but she was interested in what people are doing in the social space.

Nick from Aberdeen City Council said he was interested in using open data to model risk in society. Bill from an ODI said he was interested in finding out about open data sets for things that are not health but about health, such as open spaces.; and also about some of the management data that NHS Digital holds that could be put to wider use.

Owen said health and social care seemed like a sector in which opening data might be harder than other areas and he wanted to know how publishers dealt with consent and privacy. Magda from the ONS said it already helps the NHS with location data, so it could determine where to site services. She also suggested that ONS data could support health benefits, since it covered issues such as greenspace.

Sources of information - think laterally

The session moved onto what data sets would be useful to support these ideas; and what challenges there might be in making this data open data. Liz said that one of the things that she needed was baseline data, because she needed to know the prevalence of the problems presented to her.

Katya suggested that baseline data could sometimes be derived from other data sets; for example, people with mental health problems don’t tend to make dental appointments, so they have poor teeth. Dental decay records might give an indication of the prevalence of mental health issues in a population. Of course, as Magda pointed out, while derived data could be useful, people had to know what data sets would be relevant to them.

This raised a wider issue: that data that could help to address health problems might not be health data. For example, if somebody needs to go to hospital for treatment, and the bus service doesn’t make that easy, then improving the bus service might be a priority. And the first step to doing that might be to find some transport data. Similarly, several participants said care was often disrupted by a lack of hospital transport services or equipment; so what they needed was information about how it was organised or located.

Marius asked whether there were good sources of health information. Katya said she had used local datasets, including those produced by the Lincolnshire Health Observatory – but “Dr Google and nurse Wikipedia” could be a great source. Information about the organisation of health services and hospital services is available through NHS Digital, which has an open data portal and also publishes large data sets onto other government websites.