Knee Injections Don't Really Work

Euflexxa, Gel-One, Hyalgan, Monovisc, Orthovisc, and Supartz are not worth the time, money, and risk of infection.

[Category : Well : evidence-based recommendations for better health]

Chronic knee pain from arthritis is an almost universal complaint as we get older. As cartilage degrades, and eventually leads to bone rubbing on bone, people are left with a lot of unsatisfactory options for the pain. Tylenol? Maybe helpful. Antinflammatories like naproxen? More helpful, but with side effects and risks. Physical therapy? Sometimes reduces pain. Cortisone/steroid shots? Usually help, temporarily, but may accelerate wear and tear in the long run. And so it is only natural that people, and doctors, desperate for another option, have been turning to what many call “artificial joint fluid” injections. The name brands are listed above. Do they work?

Usually these injections are made of hyaluronic acid or its derivatives, and are supposed to improve the quality, viscosity, and quantity of the fluid lubricating a major joint like the knee. Another term for these types of injections is “viscosupplementation.” At least 1 in 7 patients with chronic arthritis has received these sorts of shots. But a major recent meta-analysis looked at the combined results of 169 unique studies involving a total of 21,000 patients, and found that viscosupplementation did not work for pain any more than placebo.

Some individual studies did find some benefit, but once again this is a meta-analysis… a summation of individual studies coming together to make a stronger scientific case.

What’s worse is that despite not having any consistent benefit over placebo, these sorts of injections can have complications including knee joint infections. The pooled study found an overall rate of 1 out of every 25 injections causing an infection in the knee joint. [As an aside, this rate seems quite high to me, as I don’t recall anyone over my years of practice developing an infected knee joint from viscosupplementation. Then again, doing these types of shots was never my jam or follow up.]

Interestingly, the authors of this new pooled study from the British Medical Journal included 80 trials that had negative results, but were mostly not published in journals. Negative meaning the shots didn’t help. Maybe they remained unpublished because they did not show benefit and were considered less interesting? This exclusion of negative results would lead to bias.

To their credit, leading Orthopedic docs got together and published guidelines on this practice almost 10 years ago:

In a 2013 guideline, the American Academy of Orthopedic Surgeons reviewed the evidence on efficacy of hyaluronic acid (HA) injections for knee osteoarthritis and concluded, “We cannot recommend using HA for patients with symptomatic osteoarthritis of the knee.”

So… these 2013 guidelines then changed behavior among providers of arthritis care like orthopedic doctors, physician assistants, and nurse practitioners, right? The stats provide a different answer:

The overall number of clinicians who provided these injections annually (≈11,000 providers) stayed relatively constant during the 7-year period from 2012 to 2018. However, the number of injections per provider increased modestly (from 101 to 108 per provider annually). The number of orthopedic surgeons who provided HA injections declined over time, whereas the number of physician assistants and nurse practitioners who provided HA injections increased markedly over the 7-year period (from ≈1600 to ≈2900). Clinicians in other specialties were responsible for only a small minority of injections.

I truly believe it’s not just about money. People want to help other people, and by the time you are getting viscosupplementation you have been suffering for a while. I don’t perform these injections, and so I don’t know the cost, but a good reference stated that each injection costs about $300, and most patients get a series of 3 to 4 injections. Insurers have still been paying for their members to receive these shots. Maybe they calculate that paying for shots saves money in the short run, or that the going rate of 30% response to placebo is better than nothing?

Anyway, I thought this was important to share. You can come to your own conclusions, and perhaps make better informed decisions based on the pooled knowledge gained from these 169 trials.

It seems pretty clear to me at this point.

I'll plan a post in the future reviewing evidence-based options for alleviating chronic arthritic pain in the knees - tempered with realistic expectations, and lots of sympathy.

~