Rebuttal: Should primary care guidelines be written by family physicians?

Guidelines should be of high quality and guidelines without specialists will miss their required expertise.

I think we all agree that guidelines should be of high quality. However, previous guidelines have sadly not met the standards laid out by Dr Bourbeau.¹

• Transparency: For 69% of Canadian guidelines, no information is provided on conflicts of interest.²

• Conflicts of interest: Up to two-thirds of guideline contributors have conflicts of interest.³

• Group composition: Specialists outnumber all other health care providers combined² (even though they provide ≤ 22% of the care⁴).

• Evidence foundation: About 50% of guideline recommendations are based on nothing more than expert opinion.^5,6

Clearly, we have long way to go. Further, my recommendation is not to exclude specialists from primary care guidelines. Rather, as Dr Bourbeau suggested, we must attain an appropriately balanced group composition.

There is no evidence that guidelines by family physicians will be better.

This is correct. However, limited evidence for an alternative approach should not mandate that we continue doing what we know is wrong. Based on the evidence listed previously,⁷ it is time for primary care providers to manage their own guidelines.

The most important concern is that practice is changed in the right direction.

I agree wholeheartedly. However, on multiple occasions guidelines have directed care in the wrong way. Some were unavoidable but some were avoidable, driven by spectrum bias, idolatry of surrogates, and conflicts of interest. Putting guidelines in the right hands might help with these issues.

I would like to offer 2 final points. First, I thank Dr Bourbeau for the wine steward analogy. I see specialist-run guidelines like a wine steward who knows rare wine from one specific region, while primary care clinicians have experience with many types of wine, especially those that most people like to drink. Second, Dr Bourbeau defines evidence-based medicine around how guidelines should be developed. I prefer the original definition, which the primary care lipid guideline⁸ modeled: the best evidence combined with (primary care) clinical expertise and patient values. Perhaps that is why our primary care guideline is recognized as a step forward.⁹