• Quote: “Inappropriate” by any other name is …

    The American College of Cardiology is changing its guidelines for when implanting coronary stents is appropriate — by banishing the term “inappropriate.”

    With 700,000 procedures in the U.S. annually — at an estimated cost of $14 billion — appropriate use of coronary stents has become one of the hottest issues in cardiology. Eight studies have found stents are no better than drug-based therapy in preventing heart attacks and death in patients with stable heart disease.

    Next year, the main U.S. heart-doctor group will remove the word it has used since 2009 to describe cases where people don’t need the metal-mesh tubes in their blood vessels. The label has become a liability in treatment disputes with insurers and regulators, said Robert Hendel, who led the effort that updated the wording.

    “The term ‘inappropriate’ caused such a visceral response,” said Hendel, a cardiologist at the University of Miami. “A lot of regulators and payers were saying, ‘If it’s inappropriate, why should we pay for it, and why should it be done at all?’”

    The cardiology group replaced the “Inappropriate” label with “Rarely Appropriate.” Another category — cases in which there’s medical doubt — will switch from “Uncertain” to “May be Appropriate.”

    Peter Waldman, Bloomberg


    • This has to be one area that I remain worried about the translation of randomized trial results to clinical practice. In most of the randomized medication trial work I have done, we managed to get adherence > 85% for the overwhelming majority of participants. When I analyze prescription claims data, the adherence rates always seem to be less.

      This creates the possibility of effect measure modification by population level medication adherence rates. What is nice about stents is that they have a high adherence rate by design (to oversimplify, they act a lot like a point exposure).

      I have been thinking about whether this could be investigated in ways where the assumptions would not be overwhelming but I haven’t had the right idea yet. [and I worry about purely observational work, as the selection effects might be pretty large making it hard to draw firm conclusions]

      I wonder if simulation work on adherence rates and plausible changes in outcome rates would be useful here? Because the adherence intuition could be dead wrong, and if it is then there is a lot of unnecessary procedures occurring — causing both a waste of resources as well as a fair amount of pain and incapacity among patients who receive surgery.

    • Did the randomized controlled trials measure the impact on quality of life? Stents significantly improved the angina my Dad suffered from and did so for many of the other patients who received the therapy. Combined with the fact that drug therapy has such low compliance rates, stents could still be seen as beneficial in many scenarios.

      That said, I don’t know how one could reasonably assess the “appropriateness” of these criteria