In spring we wrote a letter—backed by 24 charities (large and small)—to the Secretary of State, Jeremy Hunt, outlining the need for a Health Data Lab.

Using our Data Labs model, a Health Data Lab would analyse a charity’s impact on various health related outcomes—like A&E, elective and non-elective admissions and lengths of stay—by using government data.

The analysis would be entirely confidential, with the charity receiving an aggregate anonymised report of the results, like with the Ministry of Justice’s Justice Data Lab.

We recently received a letter from the Secretary of State in response to the request. It was positive—showing a good understanding of the concept and benefits of a Health Data Lab—but it did not make a commitment to making sure one is developed.

The letter suggested we continue the dialogue between NHS Digital, local commissioners and providers. Yet we have discussed with charities whether a data lab could work at a local level, and we are convinced that a national level data lab should be the priority for three reasons:

  1. National datasets are needed to ensure a robust comparison group.
  2. It would be inefficient to set up several regional data labs.
  3. Access to data, skills, capacity and most importantly relationships between the charity sector and local commissioners vary considerably across the country and are, in some places, non existent. A national level data lab would help get around this disparity.

We know that a national Health Data Lab is feasible, as a similar service is being run by NHS England and the Health Foundation to evaluate the New Care Models: Vanguards programme. And the need for the service has not disappeared, as a recent blog from Nesta highlights. So we’re not giving up.

Where do we go from here, then? We have several ideas:

  • We have responded to the Secretary of State’s letter making a request to speak with senior leaders at the Department of Health and NHS Digital.
  • We are taking steps to engage with more government departments.
  • We are exploring how we could operate a data lab ourselves, by making requests for de-identified data from NHS Digital and undertaking the analysis in-house or in partnership.

We’re also thinking about how our data labs work fits with the findings of the VCSE Review, which recommended that a well-being data lab should be established. A well-being data lab would have a different format to the health data lab. No person-level administrative dataset exists for well-being so we would not be able to track individuals over time. However, survey data is available, often to small geographies, which would enable useful benchmarking of well-being. We have always been committed to ensuring that a health data lab encompasses a wide range of outcomes, so are keen to explore with the What Works Centre for Well Being how this model could be developed.

Fundamentally, though, we think that traction is likely to be made by piloting a national data lab using established datasets (ie, admissions, lengths of stay) and then broadening metrics as the service becomes more embedded. So that’s why we’ll keep pushing for a Health Data Lab.

As always, we are keen to hear (and respond!) to your views, so do get in contact if you would like to discuss this further.

Footer