What Happens When Conflict-of-Interest is Stripped Out of Guidelines?

Nearly 20 years after hitting the prime-time following publication of the findings of the NINDS rt-PA Study Group, tPA use has become widespread.

However, there yet remain many vocal opponents. To say there is sometimes a hearty debate over the use of tPA in acute stroke would be a dramatic understatement. Indeed, even as select groups of stroke neurologists push tPA use beyond the limits of current guidelines, other groups remain steadfast in opposition.

This post is neither pro- nor con- in this ongoing controversy. Rather, this simply draws attention to a relatively interesting development over on the U.S. side of the pond. In 2013, the American College of Emergency Physicians published a new Clinical Policy regarding the use of tPA in acute ischemic stroke – giving tPA under 3 hours a “Level A” recommendation, and 3-4.5 hours a “Level B” recommendation. The uproar that ensued, however, was not entirely based on clinical grounds – it was regarding the composition of the guideline panel, co-written with the American Academy of Neurology, and stacked with experts with professional and financial conflicts-of-interest with the manufacturer of tPA. This gave rise to a BMJ investigative report, delving into the corrupting effect of COI on guideline writing. This further spun off another BMJ consensus publication regarding the evaluation of “red flags” for guidelines, dovetailing nicely with previously published Institute of Medicine recommendations.

The story, however, does not end there. ACEP’s council voted to reconsider the tPA guideline and implement a due, COI-free process, with a focus on methodology rather than content expertise, and an open review policy. The product of the revised process recently became available in draft format – and the difference is striking. The administration of tPA is no longer a “Level A” recommendation – both time windows are “Level B” in the current revision. However, there is a new “Level A” – a mandate to discuss the 7% average observed incidence of harmful intracranial hemorrhage prior to administration of tPA. Finally, a new “Level C” recommendation states, as consensus, to involve patients in shared decision-making during the process.

Whether one agrees with such changes is almost certainly tied to their view regarding the benefits and harms of tPA. It is, however, quite interesting to see how attempts to mitigate bias and COI change how the evidence is graded and the ultimate recommendations. Perhaps this demonstrates, at the least, hope is not necessarily lost regarding purifying the practice of medicine – and guidelines may yet regain the trust of physicians and patients.

(Visited 24 times, 1 visits today)