A significant proportion of medical activity is based on custom and practice rather than on sound evidence of effectiveness. It is a matter of concern worldwide to understand how best to encourage doctors to stop doing things that don't work. The paper by Black and Hutchings¹ in this issue of QSHC offers a fascinating account of the time trends in rates of surgery for glue ear and the possible impact of the 1992 Effective Health Care bulletin which outlined the reasons to doubt the effectiveness of this common surgical procedure.²

Rates of glue ear surgery in NHS hospitals, which were already falling slowly before the bulletin appeared, fell from 120 per 10 000 in 1992/3 to 68 per 10 000 in 1997/8. The fall in intervention rate remains significant—though less dramatic—when private hospital activity is taken into account. Several contextual factors are identified to explain why the message of the Effective Health Care bulletin did not fall on deaf ears: a downward trend had already begun as both otolaryngologists and GPs were beginning to doubt the value of grommets, media reports of these doubts were influencing parents' readiness to request referral, and the purchaser/provider split encouraged closer scrutiny of the effectiveness of interventions. A further contextual factor not brought out by the authors is the relative freedom of the British healthcare system from fee for service incentives for medical activity of dubious worth.

Experience from another study suggests that the rate of referral by general practitioners for glue ear fell by 50% during this period, but practitioners in both primary care and hospital interviewed in a separate qualitative study reported little awareness of or influence by the Effective Health Care bulletin.³ We cannot always easily identify what makes us change what we do, and when we do change our practice we are sometimes reluctant to acknowledge, even to ourselves, that we have done so. Many different factors contribute to a climate of opinion, but in this case the Effective Health Care bulletin looks to have made a real impact—or at least to have benefited from an extraordinary serendipity of timing.

The ancient Greeks had a word—kairos—for doing things at the right time. The Effective Health Care bulletin was pushing at an opening door. But in a highly debatable conclusion the authors of this study suggest that producers of guidelines should “focus on those topics for which the environment is likely to be conducive to change” and “avoid expending effort in areas where change is unlikely”. We should remember that powerful forces are at work to shape the environment in which clinical evidence is presented—including commercial interests and episodes of spectacular unwisdom from factions in politics and the media such as we are currently witnessing over MMR immunisation. Guideline production, in common with other forms of scientific activity, should not confine itself to picking winners. What do you think? Whether you are a writer or a reader, or just a weary recipient of guidelines, the journal would like to hear your views.