The Medicines Control Agency has just begun to map out the future of the GP Research Database, a project with a long and chequered history for which it is due to assume full control in October.
The database is the largest of its kind in the world, holding the anonymised records of some 4 million patients. But the MCA is only the latest in a series of homes for the database, created in the mid-1970s as a by-product of one of the first ever GP computer systems, the Integrated General Practice system developed by Essex GP Alan Dean. Dr Dean wanted to computerise his records and become 'paperless' - a highly futuristic vision at the time.
He employed a computer analyst to develop a system for recording and printing medical notes, spanning the four branches of the practice he shared with six partners. He set up a company - Value Added Medical Products, or VAMP - to market the systems.
He sold the first in 1979, using one of the first ranges of desk-top microcomputers. Business boomed as VAMP became a pioneer of off-the-shelf multi-user computer systems. Doctors paid about£25,000 for systems, encouraged by the prospect of saving time with computerised notes and prescriptions, and side-benefits such as automatic reminders to monitor repeat prescriptions and do preventive tests.
The research database was born out of the growing business. GPs using the IGP system were not obliged to supply data, but if they did they were paid enough to cover the computer lease fees. This meant they were getting a free system, provided the data met the agreed specification.
The database holds information ranging from drug prescription details, clinical notes and referrals, to demography and family links between patients.
At the peak of VAMP's popularity, with 6,000 doctors using 2,000 systems and around a third of them supplying information to the database, disaster struck. A consortium of City firms, which had agreed a£25m investment, pulled out, leaving a hole in the company's finances from which it never recovered.
Dr Dean was forced to sell the company to Reuters in 1993. Reuters was interested in the computer systems but not the research database. It considered shutting it down, but decided instead to offer it to the Department of Health as an asset 'for the public good'.
That created a new problem. Many of the GPs supplying data were afraid that the DoH would use the data to monitor their activity and link it to funding. A compromise was reached when the DoH assigned day-to-day management of the database to the Office for National Statistics, ensuring the data was used only for agreed purposes, such as drug safety work.
The data was also leased back to a new company, EPIC, set up by Dr Dean, which pays towards running costs and sells on data - modified to protect patient confidentiality - to pharmaceutical companies.
But sales slumped. Doctors deserted the system, and Reuters has sold the business again to In Practice Systems, which hopes to make a fresh start. But the MCA deal could be the last chance for the database.
The agency is seen as the database's natural home for two reasons. First, the historical emphasis on the use of data to monitor drug prescription fits well with the MCA's role of ensuring the safety and efficacy of UK medicines. Second, as an executive agency of the DoH the MCA has the funds to invest and can retain generated profits.
The agency has pledged to inject£3m to stabilise and develop the collection of data, which it hopes will bring back VAMP users and allow collection and integration of data from other GP computer systems.
Some observers are pessimistic. Dr Paul Cundy, chair of the British Medical Association's GP information management and technology sub-committee, says the contents of the database are 'not proven to be particularly useful information', and suggests this is the real reason why it has been passed from one custodian to another over the years. He says the data centres on prescription information which is not reliably linked to the clinical information, resulting in 'lots of records of penicillin being issued, but little else'.
Dr Dean vehemently denies the charge. 'The clinical data on the system is recorded in great detail, and its value is only just coming to light,' he says.
'The new National Institute for Clinical Excellence will send directives to primary care groups on care and the use of drugs, and PCGs will have funds from which they have to manage patients according to best practice. But a practice cannot know what its performance is without data.'
The National VAMP User Group, which represents data-contributing GP practices, says it is 'delighted that the value of this data is finally being recognised by both the government and by academic researchers'.
But it says it will continue to monitor the MCA to ensure that it observes the agreed prohibitions on the use of the data for performance management or discipline of GPs.
After a three-month consultation, the MCA seems ready to commit itself to an ambitious plan to expand the data rapidly, bringing it back up to the peak levels of the late 1980s, and eventually extending its scope to all PCGs.
Detailed planning will continue until 1 October, when around 20 staff will transfer to the agency from the ONS. By early next year it should become clear whether the long and tortuous saga of VAMP is likely to be brought to a happy conclusion.