NHS Spine 2: Q&A with HSCIC

Sooraj Shah speaks to Spine programme head Andrew Meyer and solution architect Martin Sumner about the changes required at HSCIC for the switch to Spine 2

The NHS's Spine is a part of the national infrastructure that stores patient information and enables electronic messaging. The original Spine went live more than 10 years ago. As the NHS's contracts with the suppliers who delivered it were coming to an end, the Health and Social Care Information Centre (HSCIC), a subsidiary of the NHS, investigated what it should do next in order to upgrade the system.

Computing spoke recently to many of the suppliers involved in or affected by the new Spine, dubbed Spine 2, to understand how successful the transition had been. The general consensus was that the transition had gone smoothly.

In this Q&A, Computing speaks to the Spine programme head, Andrew Meyer, and one of the main solution architects, Martin Sumner, who had been on the project from the start.

Computing: Tell me a bit more about the changes you've had to make from Spine 1 to Spine 2?

Martin: Three years ago when we started the project, it was an undertaking that was very unconventional for the organisation - to take on something of this size... it was a total change in the way it formed relationships with suppliers. It meant working with smaller suppliers, doing more things in-house and getting more choices with technology. To see it come to fruition as a live service that is able to support the users out there and also to enable broader benefits to be achieved by the NHS, giving us flexibility going forwards, is very pleasing.

Some of the suppliers we spoke to were pleased that they were able to stay online with Spine 1 while the transition to Spine 2 went ahead. How did you manage to do that?

Martin: Our aim was to make sure we maintained read access to the service throughout the transition period. We had a lot of rehearsals around data migration. There were about half a billion records which we had to migrate from Spine 1 to Spine 2 so we tried to get a large bulk of those moved across.

The final bit was to shut down Spine 1 from an update perspective, move the final delta across and leave Spine 1 open from a read perspective. [A delta migration allows you to perform an initial migration and then follow up with a background pass]. Then we could open read on to Spine 2 and with the final delta we could open up updates. This all meant that we could keep read active throughout the transition period although we had a short period where it wasn't possible to do updates. Overall, this was the most challenging part given the volume of records that had to be moved across, and it was the most complex area, where we had the most issues to contend with.

How many people do you have working on the IT team for Spine 2?

Martin: The core technical team that worked on Spine 2's core started with just two people, and at its peak we had up to 25 people working on it. Currently, there are between 15 and 20 people and our long-term target is to have about 15 people working on that particular project. We have a number of other projects that we need a similar amount of people on - we like having smaller teams as we make savings that way and communicate better as well.

Andrew: Keeping the team small was done by design.

Martin: Initially, it was a small team that had already worked in HSCIC in different roles and in other areas of the business. Then we got [consultancy] BJSS on board. We wanted to bring them in to flesh out our experience around the development and delivery process, to bring in more skills, but also to improve the skills that we had in-house. Then we had to gradually build up the team so we didn't suddenly need 25 people all at once.

What skills do you look for in your in-house team?

Martin: We've got a really tight-knit team so we'd rather get people with a broad set of technical skills. We're not interested in people who can do pure development. We want people who have skills in operational support, in trouble-shooting, in design, in testing and can actually do all of those things as an individual rather than having to break up responsibilities into different teams. We wanted, as far as possible, to maintain an agile team working for us on these projects going forwards.

Have you found that requiring people to learn on the job is the best way to ensure you have the right staff? And isn't it hard to compete with the salaries that the private sector offers?

Martin: We do have a lot of people that are genuinely interested in working for the NHS because of the outcomes we can achieve: we can save money and put money towards front-line care, in terms of patients, and and that does help balance the potential differences in pay with the private sector. It has been a successful approach so far, but we have grown the team fairly slowly. Going forward, we're going to have to look at whether we continue to expand at the rate we need.

Andrew: I think the technology we're using now is something people look for in a job. They don't just want to go into an organisation using old technology, they want to use something new and get their hands on these technologies as well.

Martin: There are lessons of scale you can learn on a project like this where you have a distributed system dealing with a large volume of messages. You perhaps couldn't learn this in some private-sector organisations because the problems we take on in terms of scale don't necessarily exist outside the big ISPs and the big banks.

There has been much criticism of the NHS's past IT projects, particularly under the old National Programme for IT (NPfIT). Does that hamper or worry you as people who work for the NHS?

Martin: We look to the future and what we can achieve going forwards. There have been well-documented problems in the past but we are trying a very different approach and we are seeing benefits coming out of that approach. There are a number of projects within HSCIC where we're trying to work with different suppliers in different ways using different techniques.

Going back two to three years when we started this project, if we had gone out to recruit then, with this project being the first, there would have been a lot of scepticism about where it would go and whether it would get abandoned, but it hasn't happened and I think that sets out our stall going forwards: to show we are here not just for one project, but others as well.