Methods of the Evidence Review and Recommendation Development

Guideline Development Process

Although not a life threatening disease, back pain is very common and the cause of considerable suffering, disability, and costs to patients, the NHS and Industry. In July 1995 the NHSE commissioned the Quality Improvement Group of the Royal College of General Practitioners to develop evidence-based clinical guidelines for primary care, based on the latest available international evidence. The remit was to update the evidence review in the U.S. guidelines (AHCPR 1994) and build on the Clinical Standards Advisory Group report on Low Back Pain (1994 a+b).

The objectives of these guidelines are:

1. To provide evidence-based recommendations on the management of low back pain to the range of clinicians involved in first contact care.
2. To ensure a multi-disciplinary approach to back pain management through the development and review process and through local implementation.

The U.S. Agency for Health Care Policy and Research provided a comprehensive review of the evidence up to 1992, which formed the basis of the U.S. guidelines (AHCPR 1994) and the U.K. report on back pain (CSAG 1994a & b). The CSAG report was mainly concerned with NHS services for back pain. It included management guidelines, but these were not fully evidence linked and the complete evidence base was not published.

The present guidelines were developed by a multi-disciplinary group facilitated by the Quality Improvement Group of the Royal College of General Practitioners. The advisory group was drawn from the RCGP, the Chartered Society of Physiotherapy, the British Chiropractic Association, the Osteopathic Association of Great Britain, the CSAG sub-committee on low back pain and patient representative from the National Back Pain Association. Observers attended from the Department of Health and National Health Service Executive.

The group started from the AHCPR review and undertook a systematic review of further evidence published from January 1993 to April 1996, using standard search techniques and review methods. Additional material, including work in press, was obtained from a number of sources, particularly from non-UK members of the Cochrane Collaboration on back pain and from other local UK guidelines.

Members of the development group carried out systematic reviews of four key areas of management:
bed rest, advice on staying active, manipulation and exercise. The latter two reviews were cross checked against independent reviews of manipulation by Koes et al (1996b) and of exercise by Faas et al (1996). Other areas of management are covered in a number of systematic reviews published since 1992 (Table 1). Much of this evidence was presented and discussed at the International Forum for Primary Care Research on Low Back Pain in Seattle in October 1995. There are also two recent mega-analysis of this complete field by Evans & Richards (1996) and van Tulder et al (1996b).

Recommendations have been derived from the evidence in two stages. First, we produced evidence-linked statements for each area of care. Secondly, small multi-professional groups prepared the recommendations in clinically important areas. These were then discussed and agreed by the whole group. Where only consensus or very limited evidence exists, we generally reproduced the original recommendations from AHCPR (1994) and the Royal College of Radiologists. Because both evidence and recommendations came from diverse sources and methodologies, we considered that a process of structured consensus was not appropriate.

The guideline review process was extensive, with 17 formal reviews of drafts of the evidence review and guidelines by a wide range of professionals and organisations (in addition to those involved in the advisory group, from Rheumatology, Management of Pain and Radiology). International opinions have been sought from recognised experts in Sweden, the Netherlands and the USA. Additionally, the guidelines have been field tested by a small group of general practitioners for their practical usability.

The Royal College of General Practitioners Quality Improvement Group will initiate a review of the current guidelines in April 1998, with a view to re-development in late 1998 if the evidence changes are sufficient to warrant action.

This review presents the guideline development group’s assessment of the evidence as it currently stands, and identifies areas where more evidence is required. In future, the Cochrane Collaboration should provide a complete and constantly up-dated systematic review of the evidence, but it is likely to be several years before this becomes available.