In the first of a new regular column from NHS Connecting for Health’s clinical leads, Professor Mike Pringle, joint GP National Clinical Lead, sets out progress and remaining challenges on GP2GP records transfers.
What’s the problem to which GP2GP is a solution?
The problem has been well understood for a long time. A patient with a complicated past and a plethora of current problems and drugs joins the practice. There is no choice but to start from scratch, hoping that the patient’s memory is reliable. Finally the fat paper folders arrive with a printout from the previous practice stuffed inside. A fraction of the clinical material gets entered.
This system compromises patient care and makes the work of primary care clinicians much more difficult. Yet all the information needed is stored electronically on the previous practice’s computer. Alongside the absurd printing out and then scanning of letters from hospital, the problem of information transfer when people change practice – and millions do every year – is unacceptable and should be easily solved.
What is the vision?
In a few years all practices will be able to transfer electronic records – GP2GP as it is called – to follow the patient as they register with a new practice. The process will start with the new practice. When the receptionist identifies the patient on the Personal Demographic Service the old practice will be identified.
If the previous practice is enabled for GP2GP, a message will be sent asking for the electronic records and the patient’s record and all related files such as letters, will be copied to the new practice.
At the new practice someone will need to review the record and integrate it into their clinical record. In particular the reviewer will need to set review dates and check the prescribing file for acceptability. From then on any clinician seeing the patient has access to their old records which are clearly flagged to distinguish them from the new practice’s record.
So what’s so difficult about that?
Before NHS Connecting for Health started to crack the problem there were two other attempts. In these it became clear that what seems like a simple problem is much more complicated.
The first is that we don’t want a text file, but one that represents the richness of the original record, including codes. If all practices used the same software and versions of Read codes this would be straightforward – but we don’t and it isn’t. Some of the fine detail will be lost because different systems record the relationship between data items differently, but the art is to minimise loss.
Because the new practice and the old practice might have any combination of GP systems, the message has to be standardised and then tested in all combinations. Not all pairs of systems have yet been tested in both directions.
Some systems use third party software to handle attachments and the complex referencing they use makes extracting attachments from the old practice difficult. The maximum file size was set by the network (N3) at 5 megabytes without realising that many records will exceed that size. These two problems are still being resolved, but should be sorted out soon.
How is it going?
However, despite these problems the trials of GP2GP are going well. They started with EMIS to EMIS in Gateshead and In-Practice Systems (INPS) to INPS in the Isle of Wight. Trials of EMIS to INPS and INPS to EMIS are beginning in Croydon.
iSoft have not yet started implementation. TPP have done a version of GP2GP from TPP to TPP and should be able to receive records from other systems. But TPP practices will find it difficult to transfer to another system since translating from Clinical Terms Version 3 back to Read 2 is challenging; TPP will therefore not be compliant with GP2GP for some time.
As of the end of January 2007, GP2GP had been used for 1,917 medical record transfers and 368 GP practice systems had access to GP2GP. A further 300 practices were being recruited in the first quarter of 2007.
Will it live up to expectations?
The message from the early trials is that it works and works well. It is safer for patients and it’s easier for GPs and practice nurses. As a practice manager reported: “I did my first, second and third yesterday!! It was very exciting and I was amazed at how easy it is just to hit the send button and have them float through the air to the other surgery. I love it!!”
Dr Rakesh Chopra of Cowes is an early adopter and he says that GP2GP is “the best innovation for medical records since we started using computers for patient care.”
It seems that there is no reason technically why GP2GP will not allow the accurate and effective transfer of patient records to follow the patient from practice to practice. Yet there are two difficult issues.
The first is the quality of the record. The content, accuracy and coherence of our electronic records vary widely. If we transmit poor records they will be of limited value to the receiving practice. The IM&T Detailed Enhanced Services (DES) is one way to encourage data quality. PRIMIS+ facilitators are available in many localities to support quality recording. But practices will need to make this a priority.
The second difficult issue is the transfer of meaning. Some practices use terms in idiosyncratic ways. This doesn’t matter until the record is shared. Some terms however have different meanings according to the source or author. As we share our records more, we will all need to involve patients to ensure that accuracy and meaning are clear; the process of generating Summary Care Records will be ideal to encourage this. And the roles of summarisers will evolve to supporting the practice in maintaining the completeness of its patients’ records.
So will it deliver on its promise?
The GP2GP Programme is progressing well. Although the obstacles are formidable, all those involved in it believe that within two years it will be an established part of everyday general practice. It will make the working lives of GPs and practice nurses much easier. Most importantly it will make the NHS a more efficient and safer place for patients.