Using consensus methods in developing clinical guidelines for exercise in managing persistent low back pain.

Objectives: To generate expert consensus evidence for the purpose of developing more complete guidelines for people with persistent low back pain than is possible using current research evidence alone. Gaps in research evidence lead to incomplete practice recommendations unless a scientific process can provide supplementary consensus evidence that is a basis for additional recommendations.

Design: A modified Nominal Group Technique (NGT). This followed a systematic review indicating incomplete research evidence.

Setting: UK-wide coordinated by the Chartered Society of Physiotherapy.

Participants: Twenty-three individuals selected for their expertise as clinicians, researchers, managers and patients.

Methods: Three stages: a first-round questionnaire of clinical questions unanswered by the systematic review; an electronic conference for outstanding questions unanswered by the first questionnaire; and a second-round questionnaire for these outstanding questions. All three stages were carried out electronically.

Results: Of 17 clinical questions unanswered by the systematic review, consensus evidence was generated for 14 questions by the modified NGT and this led to 14 recommendations for practice. Consensus was not reached for the remaining three questions.

Conclusions: The modified NGT was a practical and cost-effective way of generating consensus evidence from a UK-wide group. The consensus evidence was the basis of appropriately graded recommendations for effective care of people with persistent low back pain. Consensus methods have been little used in physiotherapy to date but are likely to be valuable in developing clinically useful, evidence-based tools for future practice.