Ultrasound again, but this time with the smoking gun showing the poor state of the evidence

A few days ago I reported that I was pretty disappointed to find so little ultrasound research for me to study. A few days later, I presented you with the smoking gun: the evidence about the state of the evidence. I’ve summarized it all here, with some repetition of the previous post, but much more, and with the references. Quite a few are provided, and as always they are hotlinked and easy to check for yourself. Remember (because it’s easy to forget) that there’s a gigantic bibliography under the hood of this website.

Therapeutic ultrasound ignored by science but sold to millions of patients

There’s hardly any research about ultrasound at all! Every scientific paper about ultrasound starts by pointing out there is not enough research on this topic. There are practically more reviews of scientific papers than there are scientific papers to review. A major review of ultrasound for a common knee problem in 2001 comically found only a single worthwhile test of efficacy to report on!

That’s not a lot to go on.

I didn’t think it would be like this. For years now, I’ve been looking forward to delving deeper into this topic, assuming that there had to be a pile of science about it. We’re talking about ultrasound, here: one of the staples of physical therapy! It practically defines the experience of going for physical therapy. Everyone has had that cold gel slapped on an injury, and felt that tingling, penetrating … placebo?

The disconnect between the ubiquity of the service and the more or less total lack of (adequate) research is jarring. A handful of studies is a disturbing joke for a therapy that is worth literally billions in the marketplace! How can that much therapy be sold without a satisfactory body of evidence that it works? Bizarre! This is exactly what I mean by “pseudo-quackery” — popular treatments that aren’t necessarily junk, but are nevertheless sold with a confidence that is out of whack with reality.

This does not mean that ultrasound never works for anyone. It does mean that it has been prescribed and sold to patients for decades with unjustified confidence. Not cool.

The pathetic state of the art and science of ultrasound

“In most cases I consider ultrasound less than useless — that's 8-10 minutes wasted that could be used doing something that might actually help.”

Jason Silvernail, DPT, Board-Certified in Orthopedic Physical Therapy, in an internet forum discussion

Ultrasound is not a difficult therapy to test (it’s easy to fake it), and if it works reasonably well then the results should be pretty blatant: simply compare results in patients who received real ultrasound to patients who get a fake instead. To a shocking degree, these simple tests have simply not been done adequately. There should be hundreds of them in the archives. Instead there are dozens.

Between 1995 and 2008, the science that has been done was reviewed in ten main papers. (See Gam, Windt, Brosseau, Robertson, Welch, Baker, Buchbinder, Ho, Ho, Jamtvedt.) Eight of those were unambiguously negative, some of them strongly so. Authors had almost nothing good to say about ultrasound. Conclusions like this one are the rule:

As yet, there seems to be little evidence to support the use of ultrasound therapy in the treatment of musculoskeletal disorders. The large majority of 13 randomized placebo-controlled trials with adequate methods did not support the existence of clinically important or statistically significant differences in favour of ultrasound therapy.

Windt et al, Pain, 1999

I like that phrasing, “did not support the existence of clinically important differences.” Ouch. Ultrasound’s therapeutic effect has an existential crisis!

Most of these reviews give a nod to some reason for optimism about ultrasound used for a particular purpose, or in a particular way. For instance, the review I just quoted, despite its overwhelming negative conclusion, also notes that “findings for lateral epicondylitis [tennis elbow] may warrant further investigation.” Naturally, that optimism about tennis is contradicted by other studies, of course (Ho, Staples).

In short, it’s all just a discouraging mess, and a classic case (yet another one) of failing the impress me test. If ultrasound were generally effective, it certainly should have performed much better in the few studies that have been done.

Therapeutic ultrasound … has fallen out of favor as research has shown a lack of efficacy and a lack of scientific basis for proposed biophysical effects.

Baker et al, Physical Therapy, 2001

Except it hasn’t fallen out of favour — it’s still widely used. The only professionals it’s fallen out of favour with, I imagine, are a small minority of scientists and unusually alert clinicians.

Consider this marketing language from a Canadian company, Shockwave Alberta, specializing in delivering ESWT (a more intense and expensive form of ultrasound, rather popular lately):

Provided you are a candidate for this type of treatment, clinical studies suggest there is a 80–85% chance this technology will improve your condition.

from the Shockwave Alberta FAQ, as of Nov 30, 2009

Here we have an entire company devoted to delivery of therapeutic ultrasound, and selling it with the implication that it is not only proven to be effective, but that they know exactly how effective — to within 5%. Based on the available evidence, do you think it’s actually possible or meaningful to declare that ESWT is “80–85% effective”? Where are the scientific review papers confirming this marvellous triumph of ultrasound? Where is the data to support such a specific promise of therapeutic success?

More ideas and references about ultrasound coming soon.