How do Dubious Norms Get Established?

A look at an old trial might explain how the norms of cardiology came to be despite a dubious evidence base

One area continues to dominate my field of cardiology. It is the search for and treatment of ischemic heart disease. Coronary heart disease dominates the minds of patients and doctors alike.

Find blockages so they can be fixed before a heart attack or death occurs—goes the thinking.

Coronary artery scans are the newest tool. But stress labs run days and weekends. Stress tests are like fuel for cath labs. Positive stress tests lead to coronary angiograms, which then lead to stents and bypass surgery. Patients feel fixed. Doctors and hospitals make money. All is in balance.

Here is the problem. Trial after trial after trial finds that fixing stable coronary artery disease does not improve outcomes over basic medications and lifestyle changes.

Seriously, this is not hyperbole. Go look at the ISCHEMIA trial. Millions of dollars of government funding went into a trial comparing early invasive coronary angiography vs a conservative strategy in patients with high-risk positive stress tests. (The kind of patient who today goes from the stress lab to the cath lab). The trial found no difference in outcomes. Patients in the conservative arm got the same results despite having about 70% fewer coronary angiograms.

The ISCHEMIA trial is not the only such trial. There is COURAGE, BARI-2D, and most of the early CABG trials all found no benefit to “fixing” coronary lesions vs basic medical therapy.

---

How did this “clogged pipe” norm get established in the absence of empirical evidence?

History helps explain it. You have to go back and look at trials.

Andrew Foy, Mohammed Ruzieh and me are chronicling all of the seminal trials in cardiology. Ultimately, we hope to make it into a book, wherein a clinician can quickly read about an important trial. For now, we are writing this book on the Cardiology Trials Substack.

Our most recent entry is the RITA-2 trial. It is one of most instructive cases of spin I have ever seen. Please go look at our post.

Before I tell you about the trial, stop and think about two treatments.

Let’s say one treatment led to a rate of death and MI of 6.3%. The other treatment led to a death or MI 3.0% of the time. You would have no trouble making a conclusion. You don’t need statistics.

Ok. Here is RITA-2. The trial randomized about 1000 patients to either angioplasty or medicines. These were patients with serious coronary lesions but not in the midst of a heart attack. They had “stable coronary artery disease.”

• In the group that had their artery opened, 6.3% of patients had a primary outcome (death or heart attack).

• In the group that had medical therapy, 3.0% of the patients had a primary outcome (death or heart attack).

This image shows what a doubling of the risk of death or heart attack looks like.

Patients in the angioplasty arm had better relief of chest pain symptoms. But that difference was diminished at 2 years and gone by 3 years. And, of course, this was an unblinded trial. So one group got a procedure and the other did not. There is surely placebo effect as well.

Here is that the authors concluded in the prominent journal Lancet in 1997—the dawn of coronary stenting.

In patients with coronary artery disease considered suitable for either PTCA or medical care, early intervention with PTCA was associated with greater symptomatic improvement, especially in patients with more severe angina. When managing individuals with angina, clinicians must balance these benefits against the small excess hazard associated with PTCA due to procedure-related complications.

You see the problem.

The trial showed a clear doubling of risk from opening the artery. Yet the authors essentially ignore it. They focus on angina relief—in an unblinded trial!

The Lancet publishes the trial.

Despite doubling the risk of death or MI, fixing blockages does not lose any steam.

Evidence is ignored. In the years to come, stents are developed. And the norms of searching for and fixing coronary blockages proceeds in earnest. To this day.

This is why Andrew and Mohammed and I believe it is valuable to revisit older trials.

RITA-2 shocks me. I can’t explain how it happened. But it helps explain why the long list of current trials showing no benefit (over medications) to fixing stable coronary disease has not changed the norms of cardiology.