GP Matthew Lockyer explains how his Suffolk practice decides which guidelines to follow

I am not sure about the rest of the country but in our little corner of rural Suffolk we are still at the banging the rocks together stage rather than the cutting edge of guideline implementation. We are, though, beginning to understand some of the issues surrounding the use of guidelines in general practice.

A guideline is not a clinical protocol. Protocols tell the user exactly what to do in a given situation. They are very useful for resuscitation situations and to help paramedical and nursing staff to manage patient care, but apply them to a GP's daily work and a range war with the local LMC results. Why? Because protocols erode doctors' individual clinical freedom. Doctors must have the freedom to make clinical decisions – even if sometimes they are the wrong ones.

Guidelines provide a means of preserving this freedom while minimising the chances of a wrong decision being made. They remind you of what you should do most of the time, so that you think about and justify deviations from the accepted management. A reasonable definition of a clinical guideline would be advice from evidence or consensus on the optimum way of managing a clinical problem in the majority of cases.

Most of the pithy one-liners that GP registrars absorb as marker beacons of reasonable practice such as, 'Third visit and no better? Admit' may now be dignified as guidelines. So too is the practice formulary. Want to prescribe a non-formulary drug? No problem, but you need to ask yourself why. There are also, of course, the formal disease management guidelines which are usually complicated.

GPs are being deluged with guidelines from many different sources. Each clinical problem may have international, national, regional, local and practice guidelines attached to it and all are subtly different. Some guidelines are definitely more useful than others and we are beginning to apply criteria to those that we try to implement or adhere to.

The researchers and experts who digest the evidence to give us best practice guidelines are often academically based. We therefore favour some robust debate before admitting management guidelines to the dirty coalface of general practice. As an example let us take the management of left ventricular failure.

The first guidelines which came to us on left ventricular failure told us how difficult it was to diagnose. An echocardiogram was mandatory in all clinically definite cases but also in those with non-specific symptoms such as tiredness. Leaving aside the impossibility of sending every tired patient for an echo, when we received these guidelines an outpatient echocardiography service was not available locally.

Our first criterion, therefore, is to reject guidelines which require obviously inappropriate levels of referral, intervention, or the use of inaccessible or non-existent services.

Our next problem is applying guidelines from research populations to those found in community practice. Again, using heart failure as an example, many of the excellent landmark studies have been carried out in patients with fewer multiple problems and a lower age than our typical patients. Most of the patients GPs see are in their late seventies and eighties with multiple pathologies. There is nothing intrinsically wrong with the guidelines but they do need modification for use in primary care.

Mrs Frail in her early eighties hates hospital. Mildly short of breath on 80 mg of frusemide, with slightly abnormal electrolytes, on anti-parkinsonian medication and painkillers for her arthritis she is a not an infrequent clinical picture. However, she is difficult to fit into guidelines designed for an otherwise healthy 55-year-old post-MI patient.

Finally, we have to feel that the guidelines are based on solid and reasonably semi-permanent evidence. We feel that our criteria will allow us to assess critically which guidelines to adopt, which to modify and which to reject. This is, we feel, the way forward to help all primary healthcare doctors offer best current practice.

Guidelines in Practice, October 1998, Volume 1
© 1998 MGP Ltd
further information | subscribe