To the Editor:

When I read the FPIN's Clinical Inquiry summarizing the effectiveness of hyaluronic acid injections, the most commonly used form of viscosupplementation for adults with knee osteoarthritis, I was surprised by the authors' conclusion.¹ The authors accurately reported the data from the sources they cited; however, they did not mention the large amount of data from randomized trials. Randomized trials represent the least biased source of effectiveness data, overcoming most of the biases found in observational studies, yet they were not included. Instead, the authors reported data on outcomes from multiple retrospective and prospective cohort studies and from cost-effectiveness analyses based on observational studies.

A 2022 meta-analysis reported outcomes from 24 large placebo-controlled randomized trials (N = 8,997; 24 weeks median follow-up) evaluating viscosupplementation for knee osteoarthritis in adults.² Viscosupplementation was statistically more effective at relieving pain than placebo; however, the difference was not clinically meaningful (the equivalent of a 2-mm decrease on a 100-mm pain scale). Similarly, it found small, clinically meaningless improvements in function. Although serious adverse events were infrequent, they occurred more often in the adults who received viscosupplementation (3.7% vs 2.5%; number needed to harm = 84; 95% CI, 49–319).

Randomized trial data should always be included in a comprehensive assessment of the best available evidence for a therapeutic intervention. Based on the best evidence, the correct conclusion is that viscosupplementation provides little or no benefit and should not be recommended to our patients.

In Reply:

I agree that if all else is equal, randomized controlled trials are the preferred type of evidence to guide clinical decision-making. However, the quality of randomized controlled trials varies, and a meta-analysis is only as useful as the studies it includes. Pereira, et al., note that there is “extreme between trial heterogeneity” in the studies they included, making their conclusions less certain.¹ Furthermore, of the 25 studies they used for their analysis, the five (20%) that were described as unpublished were less favorable to hyaluronic acid than the 20 published studies. They also state that their “findings represent summary estimates and do not exclude the possibility that selected osteoarthritis patient populations could benefit from viscosupplementation.”

The specific question addressed by the FPIN's Clinical Inquiry was not how well hyaluronic acid treats pain and function, but whether the treatment successfully delays expensive knee replacements and thus decreases costs.² Knee osteoarthritis is an extremely common diagnosis with significant morbidity and limited treatment choices. Physical therapy and exercise are often not tolerated by, or effective for, these patients. When conservative management fails, options include steroids and hyaluronic acid injections, which have mixed evidence of utility; orthobiologics; “regenerative medicine,” which is exorbitantly expensive and currently has very little supporting evidence; and surgery. Many patients want to postpone surgery as long as possible, so we sought to determine if hyaluronic acid can successfully do that.

The heterogeneity of the studies reviewed by Pereira, et al., suggests that some patients may experience significant benefit from hyaluronic acid injections and others may not.¹ This is where it is important to remember that patients are individuals, not statistical averages. Hyaluronic acid should not be recommended to all patients, but it can be a treatment option for select patients. These patients include those with comorbidities that prohibit surgery, older patients who could potentially delay a knee replacement for the rest of their lives, and younger patients who could delay a knee replacement long enough that when they have surgery, it is more likely to last for the rest of their lives without needing a revision surgery.