The guidelines challenge in question is that of how to reconcile guidelines with the needs of clinicians when treating individual patients. Clinical guidelines are devised to treat subpopulations with common ailments, and so are pitched at a general level. Individual patients, however, often present with complex multi-morbidities and the question arises as to whether a particular patient is an exception to the general rule encapsulated in a guideline. There is often professional pressure on clinicians to follow a guideline, even when doubts arise in a particular case.
There are two main responses to this conflict. One is to make guidelines more nuanced and context-sensitive, and thereby eliminate exceptions. This is the approach of the AID Knowledge working group of the Guidelines International Network, for instance. Of course, building contextual factors into guidelines increases the complexity of those guidelines, and some unforeseen exceptional cases will inevitably remain.
A second response is to keep guidelines pitched at a general level, but to treat them as default rules and to be explicit about the kinds of exceptions which would warrant departing from a guideline. Clinicians are highly skilled at identifying idiosyncratic features of particular patients – such as mechanisms involved in multi-morbidities and social mechanisms that impact on health – and these clinicians are ideally placed to evaluate these mechanisms and weigh them against the general characteristics of the subpopulation covered by a guideline. Perhaps clinicians should be given more leeway in deciding whether and how a guideline should be implemented in a particular case.
Keynote speakers were Brian Broom, Nancy Cartwright, Trish Greenhalgh and Mike Kelly and there were really a lot of fascinating talks. Several of the talks highlighted the importance of mechanisms in addressing the guidelines challenge, including those of Mike Kelly, Elena Rocca, and Sarah Wieten. There’s an interesting problem here as to how evidence of mechanisms should influence the application of a generic causal claim to a specific individual. This is a question that the EBM+ handbook begins to tackle in one of its appendices.
Centre for Reasoning, University of Kent