UK Hospital System Charts its Own Way to EMR

In the UK, the national effort to build an electronic health record (EHR) system has taken the form of the National Programme for IT (NPfIT). That program, set up in 2002, has been marked by controversy and delays; and in September, the British Department of Health issued a press release that announced the centralized approach under NPfIT has concluded, giving way to a more “locally led plural system of procurement.”

Nearly two years prior to that announcement, Rotherham NHS Foundation Trust, a hospital system about 300 miles north of London, made a decision that it would opt for alternative solutions to NPfIT, and explore best-of-breed solutions, one of the first hospital systems in Britain to do so. It selected an electronic patient record (EPR) system in March 2009. David Brown, Rotherham’s head of information and communication technology, and Ken Dobson, systems and infrastructure manager, recently spoke with HCI managing editor John DeGaspari on the hospital’s decision to go it alone.

Healthcare Informatics: Explain what Rotherham NHS Foundation Trust has done to implement its electronic patient record (EPR) system.

David Brown: In the UK, we have the NPfIT [National Programme for IT] program, the national program for IT, which was supposed to deliver a National Health Service EPR, and didn’t. So our chief executive took the view that an EPR was so important to the success or continued success to the organization, that we should go outside the national program and procure our own, which led us through the tendering process, to [the Westwood, Mass.-based] Meditech.

HCI: So you went with the Meditech system, but it was customized for your own purposes?

Brown: It’s been ”Anglicized” to take into account the NHS [National Health Service], the English way of dealing with healthcare, and dealing with some of the critical elements that a U.S. system needs before it will get sanctioned [for use in the UK]. It’s been going on for nearly two years; we are going live June of next year. Our teams have done a lot of Anglicization and we are introducing SNOMED [Systemized Nomenclature of Medicine—Clinical Terms] into the system, which is new to Meditech and quite new to the NHS. So that is being embedded in the product.

There is a lot of user acceptance testing around the Anglicized [product], and that is in the process of concluding this side of Christmas. And what they are doing currently is beginning to configure it as per our process groups, we are obviously process mapping everything and working out what the future stage should look like, and creating that within the configuration.

HCI: How do you expect it to conform to clinicians’ workflow?

Brown: I suspect that is something that will become more apparent next year. We are trying to configure the system to the point where the impact on the clinician is minimal. They are looking at all sorts of things, like voice to text, so voice recognition, handwriting recognition, the use of tablet devices, etc. And we’re trying to tailor it very closely to the processes that the clinical workflow in this country.

I think that there is a very big piece of work yet to come about exposing the clinicians to the system, [which] will be done in field trials. The desire here is to go very much as paperless as possible from day one [in] the [inpatient] and outpatient clinics. We don’t want to be carrying on with the paper records, paper case notes, for very much after go-live. So it is a very aggressive proposal, and has the potential to be very impactful on clinicians. So there is a very big piece work to be done, and constantly we are worried about clinical risk and those sorts of things. So we have a whole team delivering risk impact and analysis of the system.

Ken Dobson: It’s been a great opportunity for us as well to induce standards across the trust, to use elements within the system to make sure that doctors are following the same guidelines and procedures to benefit. It’s to make sure that standards are consistent as possible across the trust and are used appropriately. [We are] matching the workflow that standard procedures are followed as much as possible, and where there is variation, it can be audited and assessed whether it needs adjusting.

Brown: We’ve got probably very clinically rich, but not joined up, subsystems: PACS, cardiology systems, endoscopy systems, laboratory, so, quite a few separate clinical systems. Nothing is really sitting above, within a core administrative system. What we wanted to do is sweep that aside and replace it with a single integrated electronic patient record, to manage the whole, and provide potentially clinical decision support.

HCI: What other software components will comprise the system for data backup and disaster recovery?

Brown: There are only a couple of vendors that can deal with the Meditech data or really understand the underlying database. The most mature, most widely used product was the Bridgehead [HT Backup] product [selected by the hospital in October 2009]. This is integrated service backup and disaster recovery. The key thing with the disaster recovery element is we have got two server rooms, and using the power that is available to us through Bridgehead, if we lost one server room, we can get the database back, to a maximum of four hours. And so it is a very powerful tool. We actually did some disaster recovery tests last weekend, and it was successful. We did some tests, and we found that we could get the recovery time down to about an hour.

HCI: The move to electronic patient records will have a huge amount of patient data to deal with and store, and there will be a great deal of expense to create the needed backup and storage. How are you dealing with that challenge?

Brown: Yes, it is a challenge. When we went into partnership with the likes of Dell Perot and Meditech, as part of the contract, we said we wanted enough storage for clinical data for five years. The sum that we’ve got will hold, without adding extra disks, is 80 terabytes of data. These sums scale up to 480 terabytes. The BridgeHead product has all sorts of de-duplication facilities within it. And using de-duplication for, we’ve got their FileStore [EHR] product, which de-duplicates stored archive records, so it’s not, doesn’t keep us awake at night yet.

And because MailStore [e-mail archiving and management software] builds on FileStore, that was also a big problem for us. We’d just moved to [Microsoft] Exchange 2010, and we wouldn’t have been able to do it without the power associated with MailStore. Our clinicians and our administrators keep e-mails forever. And when you consider that on expensive, fast spinning disk, it doesn’t bode well, if you don’t archive it to cheaper disk, which is what we’ve done.

HCI: Are there other hospital systems in the UK that have taken a similar approach, taking their own route to EHR outside the national system? How much resistance did you encounter?

Brown: A lot of resistance. A lot of political pressure was placed on our chief exec. Obviously, we had a loan [of] a significant amount of money from the NHS bank in order to pay for this system. There have been a lot of high level discussions about opting out of the national program in order to. But we have always treated this in the way that the chief executive has tackled it: that this is an interim solution for the Trust to deal with real clinical and operational difficulties. If the national program ever delivers a clinically rich system, we will move to it. It’s just so important for health organizations in this country to have access to good clinical data. And, in both cases, is the ability to work out what your income is. And we have gotten ourselves ahead of the curve, and there are a number of other hospitals beginning to tender outside the national program, in order to deal with their problems.

If they are not doing that, what they are doing is trying to create their own electronic patient record by stitching it together with clinical portals. So that is the other approach, to take it here and say, before this national product hits us, we will take all of the clinical systems that we have and we will stitch them together in a nice front ended clinical portal. Which obviously takes a large amount of development work, or you are dealing with third party companies, which are also quite expensive.