Q As there is no 'Guidelines Ombudsman', what is the best way for a jobbing GP to evaluate the credibility and validity of a given set of recommendations?

A There are perhaps two different situations to consider: one is where you need to judge whether a guideline is sufficiently sound to implement in your practice; the other is when a patient comes into the surgery, clutching print-outs of guidelines he/she has found on the internet.

In the first situation, I suggest that since there are so many guidelines one might implement, you choose only those that have the strongest endorsement from the most respected institutions, e.g. guidelines on important aspects of clinical care, where experts from government-funded agencies or professional bodies consider that the quality of evidence is good and there is general agreement about the best course of action.

In these circumstances, you can safely leave judgments about the quality of the guideline to others, and concentrate on making things happen in your practice.

In the second situation, you may have to deal with any one of a large number of ever-changing guidelines being developed and published all over the world. There is no simple answer to judging whether any given guideline is good or bad.

At one end of the spectrum, you can dismiss a guideline if its source is unknown or obviously biased, or if it describes very different circumstances from those applying to you and your patient.

At the other end, if it is a guideline from a respected independent source you will probably feel comfortable in taking the recommendations seriously.

Unfortunately there are many shades of grey along this spectrum, and no easy answer to your question. A rule of thumb might be that if the recommendations look reasonable, accept them; if you are not sure, refer to a specialist colleague; and if they look unreasonable, explain your reasons for thinking this to the patient .

Q GPs often feel a sense of foreboding in putting guidelines into practice. With target setting and National Service Frameworks, people often feel they must be reinventing the wheel. Are there any central repositories with 'how to do' recipes for busy GPs?

A There are lots of different sources of advice. For example, there are many books and articles on implementing guidelines: if you locate one you can follow the references from there. Three books (all from Radcliffe Medical Press, Abingdon) you might start with are:

  • Clinical Guidelines – From Conception to Use. 2000.Edited by Martin Eccles and Jeremy Grimshaw
  • Guidelines in Clinical Practice. 1999. Edited by Allen Hutchinson and Richard Baker
  • Implementing Clinical Guidelines – A Practical Guide. 1999. Edited by Debra Humphris and Peter Littlejohns.

A MEDLINE search may also be useful, and can be done through your local PGMC library. The librarian will be able to help you identify other resources.

Other useful sources of information include: CRAG in Scotland; NICE; The North of England Evidence Based Guideline Development Project; Centre for Health Services Research in Newcastle; University of Newcastle upon Tyne; the Health Care Evaluation Unit at St George's Hospital Medical School, London; the Clinical Governance Research and Development Unit at Leicester; and the NHS Clinical Governance Support Team.

Local resources include: the local primary care audit group (PCAG); other PCAGs (e.g. Suffolk); various regional and subregional groups (e.g. CHAIN in London); and colleagues who have already implemented the same or a similar guideline.

You can also get help and advice from the RCGP and other Royal Colleges by contacting their guidelines/audit unit.

Finally, you can rely on your own resources. None of the above is likely to give you a recipe in the sense of a 'colouring by numbers' solution to the problem. The principles of guideline implementation are the same as those of implementing any project: be clear about the objectives; take participants along with you; have a project plan; and identify an individual as having responsibility for coordinating activity and taking things forward.

The skills required for successful implementation of guidelines are not highly specialised, but are common sense. Anyone with determination and reasonable experience of life can make a guideline work. Just don't make it too complicated – for yourself or others.

Q It is sometimes difficult to compare numbers needed to treat (NNTs) between different trials because they often relate to different time frames. Is it statistically valid to divide by the relevant denominator, and are there any plans afoot to agree a convention?

A NNTs are defined as the number of patients who need to be treated to prevent one adverse outcome.

In order to calculate, or to understand an NNT, you need to specify the treatment, its duration, and the adverse outcome being prevented.

You can convert this to an NNT for a specific patient by estimating that patient's susceptibility relative to the average control patient in the trial report.

Iû am not aware of any plans to agree a convention, and you should therefore use only NNTs derived in a situation that is the same as, or similar to, the one in which you are making treatment decisions.

If you have a question or a problem that you would like to put to our editorial board, please contact Guidelines in Practice:
by email to corinne@mgp.ltd.uk
via the feedback page on the website
by fax 01442 877100
or by post to Guidelines in Practice, The Chapel, Park View Road, Berkhamsted, Hertfordshire, HP4 3EY

Guidelines in Practice, December 2000, Volume 3
© 2000 MGP Ltd
further information | subscribe