Is 'continuity of care' an outmoded concept? While we still extol its virtues, the current reforms seem to threaten the whole idea. Whether the discussion is about GPs' out-of-hours co-operatives, walk-in centres, junior doctors and their shift systems, booked admissions or NHS Direct, all have an impact on continuity. Can they be reconciled, or is continuity just one more NHS shibboleth about to be laid to rest?

Continuity of care means different things to different people. The expression has been used mainly in British general practice, where continuity of carer reflected the single-handed nature of most GP practices in the NHS's early days.

Most GPs in the 1950s and 1960s spent their working lives with a relatively fixed group of registered patients, developing ties with individuals and their families, and learning as much about the nonphysical aspects of their patients' illnesses as about their physiology.

Where it worked well, such a philosophy allowed doctors to develop a more holistic view of their population and a contextual approach to general practice that ensured treatments matched needs more closely than from a purely 'medical' viewpoint.

Perhaps more importantly, continuity extended to the way in which patients went through the secondary-care sector. Apart from providing primary medical care itself, general practice's key role has always been to organise and co-ordinate access to more specialised providers, so-called 'gatekeeping'.

By insisting that virtually all access to such specialists went through a single channel, the NHS created a dual-function control. It linked supply to demand in a relatively coherent fashion (and so eventually gave rise to the notion of NHS 'purchasers' procuring services from other NHS 'providers'), and it ensured that all such specialist care could be brought together at a single locus, so that whoever sat at that point could provide the continuity, the case management, that helped minimise the overlap between different parts of the NHS system.

It almost didn't matter what treatments GPs chose to administer themselves, as long as they were in a position to co-ordinate the services others were providing to their patients.

As the 1960s progressed, the GPs' charter was introduced in an effort to strengthen general practice. One effect was to encourage GPs to combine into larger group practices, a movement which continues today, with the average practice now consisting of 2.5 full-time equivalent GPs.

This meant that the original concept of continuity of carer began to be questioned. While some group practices still acted like a series of single-handed practitioners, most began to look more like associations, sharing their staff and patients, and introducing the notion of continuity of care, with the actual person providing the care no longer being so closely specified. Communications within practices had to be improved to maintain the approach to care and its consistency, and so the focus of continuity began to move away from the provider of care towards the idea of consistent information.

The changes of the past few years, particularly the emergence of out-ofhours centres, have increased the emphasis on information links.

Patients seen overnight have the details of their contact with the NHS transmitted to their own practice the next day to maintain the practice's case-management role.

The latest challenge to the concept of continuity comes from services such as NHS Direct and the so called 'walk-in' centres. Is it important that they are integrated into the continuity of care, and if so, how might this be done?

The short answer to the first question is: it is very important, since a large part of the NHS's efficiency lies in the fact that (when the system works properly) duplication of service is rare, with good coordination of care to individuals.

Users are usually guided through the system, with notes and results arriving at the right place at the right time, and with little overlap between specialists. As in other good case management models, the care is person-centred rather than disease centred, an important distinction in these consumer-sensitive days, and one which probably has implications for the system's effectiveness.

In which case, how may the increasingly dispersed models of care delivery be co-ordinated? As the individual's first point of access to the NHS, primary care is ideally placed to establish the problem, determine with the client the pathways required, set them up, supply the appropriate information to the client and monitor progress.

The fact that the organisation is also a provider of services is subsidiary, since primary care has always been used to 'sub-contracting' work that cannot be carried out locally, and so the notion of 'doing what you properly can, and buying in the rest' is not a new one - merely one whose time has come.

Second, the centrality of the primary care organisation in case management needs to be acknowledged and strengthened. All NHS activities need to be co-ordinated, including general practice, accident and emergency, NHS Direct and routine surgery.

Family planning and sexually transmitted diseases have traditionally been excluded as exceptional cases in which special clauses for confidentiality exist, but to extend such special cases to the NHS more widely - as seems to be the case with NHS Direct - makes no sense, attacking the service's very heart.

And if co-ordination is strengthened in primary care, it makes sense to base the infromation technology in the same milieu. As the central NHS 'server' to the 'web' of services, let's base the whole of the NHS information strategy in primary care, and make continuity of information the leitmotif of 21st century healthcare.