What Works for Knee Pain? Good Medicine, Bad Medicine, and Wasted Money

Knee pain is a very common problem, affecting young and old alike.  Among athletes, knees take a beating, particularly in sports that involve running and twisting.  Among older people, the cartilage that provides a cushion in our knees often just wears out, producing chronic stiffness and pain.

I've had knee problems myself, so I was interested to see an article posted on CNN, with content from Harvard Medical School called "Alternative Treatments for Knee Pain."  Would it describe "alternative" medicine, as in "complementary and alternative medicine," or would it present real medicine for knee pain?

A little of both, as it turns out.  But it does more: it inadvertently illustrates one of the major flaws with the U.S. health system.  We don't discriminate between effective and ineffective treatments, and some doctors seem content to let patients try anything, regardless of efficacy or cost.

The article describes several treatments for knee pain, and for each one it turns to Eric Berkson, M.D., for an opinion on how well the treatment works.  Berkson is an orthopaedic surgeon at Massachusetts General Hospital and an Instructor in Orthopaedic Surgery at Harvard Medical School.

Despite these excellent credentials, Dr. Berkson doesn't seem to have a firm command of science-based medicine. The article starts out strong, recommending the most effective therapies, but then wanders into unproven therapies, and concludes with Berkson making an unscientific recommendation for the most wildly ineffective treatment of all: acupuncture.  Let's look quickly at each "alternative."

The most effective treatments, according to Dr. Berkson, are physical therapy and weight loss.  Berkson is on solid scientific ground here: physical therapy strengthens the muscles around the knee, relieving some stress on the joint, and weight loss does the same thing by reducing the load directly.  If you stopped reading at this point, you'd be in great shape.

Next is a discussion of chondroitin and glucosamine, two very popular supplements that are often used to treat joint pain.  After many years of competing claims about these supplements, NIH conducted a large, $12.5 million study called GAIT to test whether these supplements worked for knee pain. The researchers looked at glucosamine and chondroitin separately and together, and compared them to a standard pain reliever (celecoxib, brand name Celebrex®, an NSAID) and to a placebo.

The conclusion: celecoxib worked, but the supplements did not.  In a small subset of patients with moderate to severe pain, there was a hint of a benefit for glucosamine-condroitin, but it wasn't significant.  For patients with mild pain, though, only celecoxib was better than a placebo.

Despite this large and compelling study, Berkson says

"there's a lot of evidence that they [glucosamine and chondroitin] work, but only in a certain percentage of patients." 

Perhaps he's referring to patients with moderate to severe pain from the GAIT study, but the article doesn't explain if that's what he means.  And then Berkson recommends them anyway, saying "there's little downside to trying the supplements."  Especially if it makes the patient think the doctor is helping.

The article then describes three different treatments that involve injections of various substances into the knee.  To keep this brief, I'll just consider the first: viscosupplementation, a fancy name which means that hyaluronic acid is injected into the knee. It's supposed to help lubricate the joint, which sounds plausible: after all, we lubricate machines to keep them running smoothly.  Shouldn't some "grease" help our knees?

For hyalonuric acid, the answer is simply no.  A large review just last year, involving over 12,000 patients (Rutjes et al. 2012) shows that hyaluronic acid has little or no benefit, and possibly causes harm.  Their conclusion:

"In patients with knee osteoarthritis, viscosupplementation is associated with a small and clinically irrelevant benefit and an increased risk for serious adverse events."

Berkson's own colleagues, the American Academy of Orthopaedic Surgeons, just issued a 1200-page report on treatments for knee osteoarthritis, and they found strong evidence against the use of hyalonuric acid.  To be specific, the AAOS says "We cannot recommend using hyaluronic acid for patients with osteoarthritis of the knee."  They label this a "strong" recommendation.

Despite the the advice of his own professional organization, Dr. Berkson says

"I have some patients who think [hyalonuric acid is] a miracle drug that keeps their symptoms at bay.  Given that the risks of injection are very low, it makes sense to try it if insurance will pay."

Oh my.  Here we have a medical expert recommending an expensive treatment - he estimates the cost is about $1500 - despite the evidence, apparently because he has patients who think it works.  Paying for expensive, ineffective procedures is a poor way to treat patients and a great way to run up costs.

(Aside: I might seem to be picking on Dr. Berkson here, but I think his responses are probably typical of many practicing doctors.  Not to excuse him, but it's easy to follow your instincts and to rely on anecdotes from patients.)

After briefly discussing two other injection therapies, the CNN/Harvard article addresses one last alternative: acupuncture.  It quotes Berkson again, this time saying:

"Acupuncture works in some patients, although I don't think science knows exactly why it works. It's worth trying if other things aren't working for you."

Despite Dr. Berkson's comment, the science doesn't support him.  Acupuncture for knee pain, and for other pain, simply doesn't work.  (See this article for a lengthy discussion, with citations.)  Just to pretend for a second: in order for acupuncture to work, there would have to be lines of force (called "meridians") through which a mystical, apparently unmeasurable life force ("qi") flows, and furthermore you'd have to be able to manipulate this life force by plunging needles into the skin.  There is no support whatsoever for these ideas in human biology and physiology.

The American Academy of Orthopaedic Surgeon's report addresses acupuncture too, stating that "We cannot recommend using acupuncture in patients with symptomatic osteoarthritis of the knee."

Dr. Berkson's recommendation is akin to a shrug: while admitting that he doesn't know why it might work, he just says to give it a try.  His own professional organization, after an extensive review of the scientific evidence, disagrees.

The CNN Health article, with "content provided by the faculty of the Harvard Medical School," demonstrates in a microcosm how health care dollars are wasted in the U.S.  First it recommends the most effective and least expensive treatments: weight loss and exercise.  Both of these involve lifestyle changes that are difficult for many people.  The article then suggests alternative treatments for knee pain that include:

1. Glucosamine and chondroitin supplements - ineffective but cheap. 
2. Knee injections of hyalonuric acid - ineffective and very expensive.
3. Acupuncture - ineffective, wildly implausible, and sort of cheap.

Despite the evidence against them, the doctor quoted throughout the article recommends trying all three, especially if (as is common) the patient doesn't have to pay the bills.

Science only supports three treatments for knee pain, as physician Harriet Hall summarized in her recent blog post: exercise, weight loss, and NSAIDs.  The "alternatives" described by CNN, no matter how appealing they sound, are a waste of time and money.