The Northern Specialist Group heard a presentation on 'connecting up healthcare - a view from the NHS Technology Office' by its chief technology officer Dr Paul Jones on 22 January. The group’s chair Professor Bernard Richards reports on the meeting, based on Tom Sharpe's notes as rapporteur.

Dr Paul Jones' appointment followed that of the chief information officer for the NHS in September 2008. A function was created called 'Technology and Standards', which led to the creation of the Technology Office for the NHS.

Paul's first slide showed a phlebotomist working in Salford and using a mobile clinical assistant (a type of mini-laptop) developed by Motion Computing. It is the first computer of its type designed specifically for use on the ward. It was developed following a meeting between Richard Granger, the former chief executive of NHS Connecting for Health, and representatives of Intel.

The two items he wanted Intel to help develop were a ward-specific computer and a device for collecting health data in the home. One challenge was to provide an eight-hour battery life. This was addressed by allowing a battery which was nearly flat to be swapped with a fully-charged one from a base station without loss of data.

An unforeseen problem that surfaced in testing was that health professionals used any convenient pointed object to operate the touch screen - including scalpels and syringes - thus destroying the screen within a couple of days. The solution to this problem was to make the screen respond only to an active stylus. Although the prototypes were expensive, prices are expected to fall when such devices are in common use.

The National Programme for IT (NPfIT)

Paul showed figures which gave a measure of the size of the NHS, including: 305 Trusts in 10 strategic health authorities; more than 8,000 GP surgeries; over 10,000 pharmacies; more than 400,000 nurses; nearly 50 million patients; and 360 million plus prescriptions per year.

Paul said that it is a common misconception that using a supercomputer would be the best way to handle all this complexity. There does have to be a high-level architecture and in that there will be both national and local systems. There has to be a single place to go to find a patient's unique NHS number. Facilities for providing data for research are also centralised. Some of the things that were done in the early days of NPfIT were over-centralised, but Paul claimed that this is no longer true.

There is also integration via the transaction messaging system (TMS) with many local systems including both new and legacy systems. Multiple vendors are producing applications - in the north of England examples of these are Lorenzo (iSOFT) and TPP. The full picture is very complex, as demonstrated by a slide containing roughly a hundred program modules.

The Electronic Prescription Service provides an example of interoperability. Release 2 of this is designed to support a paperless transaction from GP practice to patient to dispensing pharmacist. The link to the Prescription Pricing Authority carries some £6 billion worth of prescription details.

There are six main GP systems within the NHS, and some sixteen different systems within pharmacies. With suitable security measures in place, prescriptions generated in the GP systems cause messages to be sent across TMS to the Spine, and then onwards to the dispensing pharmacist.

The National Integration Centre has been set up to deliver this kind of interoperability. It has assured 15 national releases, 350 versions of 61 products outside the local service provider (LSP) contracts, released 121 Clinical Authority to Release certificates to suppliers outside the main LSP contracts, and assured 79 separate systems under the LSP contracts.

In summary, NPfIT uses a variety of bespoke and commercial off-the-shelf (COTS) products to deliver requirements. There are systems with more transactions (e.g. Google) and there are systems with greater data volumes (e.g. Ordnance Survey), but NHS systems combine both. The processing power in the Spine would place it in the top 100 of the World’s supercomputers.

In business terms, the original vision for NPfIT in 2002 consisted of the ‘Greek Temple’ model with three columns corresponding to the provision of a prescription service, a booking service and a life-long health record service, together with a base corresponding to the national electronic infrastructure (N3 and Spine).

This has now grown to an edifice with perhaps eleven columns - adding patient choice, digital imaging, secondary-user services (SUS), email, GP-to-GP record transfer, Quality Management Analysis System (QMAS), NHS Numbers for Babies and Bowel Cancer Screening. All of these systems are up and running. In addition, commissioning, payment-by-results, and plurality of provision are supported.

Other programs are delivered by other bodies, like the Staff Salaries System commissioned by the Department of Health’s Workforce Directorate. They also commissioned MTAS, the system for processing applications for junior doctor posts which suffered from adverse publicity in 2008. Paul suggested that there may be advantages in keeping everything under one organisational roof.

In a typical week (figures from May 2007), 1,250 new users registered for access to the NHS Care Record Service and 50,000 unique users accessed the main NCRS database. A total of 650,000 prescriptions were transmitted electronically, 6.5m patient digital images were captured and stored, and 7m queries recorded on the Patient Demographics Service (PDS). Electronic bookings made via Choose and Book (C&B) totalled 90,000. About 50 per cent of GP-referred first outpatient bookings to the acute sector are now made this way.

Paul suggested that there are several myths associated with NPfIT. The National Audit Office confirms that it is not over budget in terms of its core contracts, these remaining at just over a forecast outturn cost of £6bn. NPfIT is not a single programme so press statements of the form ‘The Programme is delayed’ cannot be totally accurate.

Some projects (like N3 and PACS) have come in early and under budget, but it is true that some projects are well behind schedule. There has been plenty of clinical engagement, as Paul himself can testify. There have been policy changes since the contracts were made, but this is hardly surprising and the architecture referred to earlier is designed to support change. Paul argued that QMAS has improved patient care, and PACS has reduced diagnostic waiting times and shown payback (in one typical example) within six months.

Electronic prescriptions have reduced prescribing errors, C&B has reduced ‘did not attend’ occurrences, and the PDS has reduced the number of patient letters sent by hospitals to the wrong address. The often chaotic condition of paper records makes it difficult to believe that things were better before NPfIT.

March 2009