Sunday, 19 October 2014

Surgery for high blood pressure

The most recent casualty of the sham surgery trial, adding to the list of operations that looked good and had good results until put to the leased biased test, is a procedure called renal denervation (cutting the nerves to the kidney). Years of good results showing that this procedure lowered blood pressure are now met with a blinded sham-controlled trial that showed no significant benefit over placebo.

The article is reported here in the New England Journal of Medicine and a nice editorial can be found here (paywall, sorry). You can read overviews and interviews in Medscape,

Renal denervation is a procedure that involves damaging nerves to the kidney. It is usually done in patients with high blood pressure who do not respond to medications. It is approved for use in many countries and thousands have been (and probably still are) performed worldwide.

Previous reports (without blinded control groups) showed that the blood pressure falls after using this device, dramatically. It should be noted that the blood pressure (on average) also fell quite a bit (14 mm Hg) in those receiving renal denervation in this sham study. My point, however, is that the blood pressure also fell in the sham group (12 mm Hg), either from natural history, other treatments or behaviours, regression to the mean or what is loosely referred to as the placebo effect. The fact that both groups improved by a similar degree (there was no significant difference) means that the surgery had no specific effect (of its own) and this, above all else, is why you need properly controlled, blinded, placebo studies. This is how we find out what really works, not what we think works. This is how we assign causation, instead of assuming it from association.

The bottom line

Once again we have a highly touted treatment, in this case one that was initially thought to be a “cure” for high blood pressure, having a true level of effectiveness much lower than we believed, and not coincidentally, a lot lower than we wanted to believe.


  1. yet, the marketing of this, yet another (surgical) miracle cure is unrelenting and not bothered by evidence that would question the effectiveness - or safety.

    "Renal denervation, which is now available at eight hospitals in Australia via GP referral, is minimally invasive and involves the insertion of a catheter through the femoral artery located near the groin. Once inserted, a tip at the end of the catheter emits a radio frequency to ‘silence' sympathetic nerves in the arteries that deliver blood supply to the kidneys. The procedure is expected to revolutionise treatment options for people with high blood pressure."

    "In 2011, Professor Murray Esler AM and Professor Markus Schlaich were awarded the prestigious Eureka Prize for Medical Research Translation for their pioneering research into renal denervation. Their work is an important example of Australian research that has translated into a revolutionary new treatment for a global health epidemic. "

  2. and here:
    Expert body endorses renal denervation
    2 May, 2013

    1. Thanks for these, very interesting.

  3. Do you recommend abandoning an "accepted" surgical procedure based on the results of a single sham surgery trial? Epistemically, if established practice is a tree, I see new knowledge as a rainforest vine, slowly enveloping, suffocating, and eventually replacing the tree. Can any single study instead act as a single axe blow, bringing down the tree?

    1. Interesting analogy. It all depends on the weight of evidence for and against (the weight of the vine and the strength of the tree?). It also depends on the communication and understanding of that evidence by end-users.
      There are examples, like in my following blog post on hormone replacement therapy, where practice was largely reversed based on the (very large) WHI trials.
      There are also examples, like knee arthroscopy, where the rates of usage have increased, despite multiple, high-quality trials showing no benefit from the procedure.
      To answer your question: it is possible for one trial to do this, if the existing evidence was poor, and the trial was good, but replication of such a trial would also be important.

  4. 2014 Dubious Innovations In Cardiology

    Dubious Innovative Device: Renal Denervation

    SYMPLICITY HTN-3 was the trial that finally ended the massive hype over renal denervation, the novel blood-pressure lowering technology. The trial followed in the wake of the publication and appearance of hundreds of articles and CME programs virtually guaranteeing that renal denervation would be a cure for treatment-resistant hypertension. Instead the trial showed, definitively and absolutely, that renal denervation, in its current form, and as it had been used in earlier less rigorous trials and in clinical practice in Europe, was no better than sham treatment.

    Medtronic, the trial’s sponsor, and the trial principal investigators, Deepak Bhatt and David Kandzari, deserve credit for supporting the trial and faithfully reporting the results. Credit is also due Darrel Francis and his colleagues for having the courage to publicly predict that the trial would fail to show a large benefit for the device. At the time they were a voice in the wilderness and have still not received nearly enough credit for their analysis of why the trial would likely fail. Prophets, alas, often go unheeded.

    Perhaps, after a lot more basic research and development, renal denervation will prove to be a useful tool. But for now there is simply no excuse for its still being available for clinical use in Europe outside of a research setting.