Clinician Article

Comparative effectiveness of pharmacologic interventions for knee osteoarthritis: a systematic review and network meta-analysis.

  • Bannuru RR
  • Schmid CH
  • Kent DM
  • Vaysbrot EE
  • Wong JB
  • McAlindon TE
Ann Intern Med. 2015 Jan 6;162(1):46-54. doi: 10.7326/M14-1231. (Review)
PMID: 25560713
Read abstract Read evidence summary Read full text
  • Special Interest - Pain -- Physician
    Relevance - 6/7
    Newsworthiness - 6/7
  • Internal Medicine
    Relevance - 6/7
    Newsworthiness - 5/7
  • Rheumatology
    Relevance - 6/7
    Newsworthiness - 5/7
  • Family Medicine (FM)/General Practice (GP)
    Relevance - 6/7
    Newsworthiness - 4/7
  • General Internal Medicine-Primary Care(US)
    Relevance - 6/7
    Newsworthiness - 4/7
  • Surgery - Orthopaedics
    Relevance - 6/7
    Newsworthiness - 4/7
  • Geriatrics
    Relevance - 5/7
    Newsworthiness - 3/7


BACKGROUND: The relative efficacy of available treatments of knee osteoarthritis (OA) must be determined for rational treatment algorithms to be formulated.

PURPOSE: To examine the efficacy of treatments of primary knee OA using a network meta-analysis design, which estimates relative effects of all treatments against each other.

DATA SOURCES: MEDLINE, EMBASE, Web of Science, Google Scholar, Cochrane Central Register of Controlled Trials from inception through 15 August 2014, and unpublished data.

STUDY SELECTION: Randomized trials of adults with knee OA comparing 2 or more of the following: acetaminophen, diclofenac, ibuprofen, naproxen, celecoxib, intra-articular (IA) corticosteroids, IA hyaluronic acid, oral placebo, and IA placebo.

DATA EXTRACTION: Two reviewers independently abstracted study data and assessed study quality. Standardized mean differences were calculated for pain, function, and stiffness at 3-month follow-up.

DATA SYNTHESIS: Network meta-analysis was performed using a Bayesian random-effects model; 137 studies comprising 33,243 participants were identified. For pain, all interventions significantly outperformed oral placebo, with effect sizes from 0.63 (95% credible interval [CrI], 0.39 to 0.88) for the most efficacious treatment (hyaluronic acid) to 0.18 (CrI, 0.04 to 0.33) for the least efficacious treatment (acetaminophen). For function, all interventions except IA corticosteroids were significantly superior to oral placebo. For stiffness, most of the treatments did not significantly differ from one another.

LIMITATION: Lack of long-term data, inadequate reporting of safety data, possible publication bias, and few head-to-head comparisons.

CONCLUSION: This method allowed comparison of common treatments of knee OA according to their relative efficacy. Intra-articular treatments were superior to nonsteroidal anti-inflammatory drugs, possibly because of the integrated IA placebo effect. Small but robust differences were observed between active treatments. All treatments except acetaminophen showed clinically significant improvement from baseline pain. This information, along with the safety profiles and relative costs of included treatments, will be helpful for individualized patient care decisions.

PRIMARY FUNDING SOURCE: Agency for Healthcare Research and Quality.

Clinical Comments

Family Medicine (FM)/General Practice (GP)

Does hyaluronic acid really work this well? Not what other reviews have found.

Family Medicine (FM)/General Practice (GP)

Of course, network meta-analysis must be taken with a grain of salt substitute. Experience has shown that high-quality direct comparison randomized trials often have results that are different from direct comparisons. Most importantly in managing osteoarthritis like any pain problem, we must keep in mind Dr. Moore et al`s brilliant observation published in BMJ. When managing pain, expect failure but strive for success. Every treatment studied was better than oral placebo, so a series of trial-and-error treatments in patients with osteoarthritis is necessary.


A network meta-analysis is an impressive methodologic undertaking that is said to discern effects that are missed in standard meta-analyses. However, it does not avoid any of the pitfalls that plague meta-analyses. Assumptions upon assumptions about the quality, consistency, and biases in individual studies are unavoidable. Perhaps that`s why standard meta-analyses lead to the conclusion that hyaluran has minimal specific efficacy, whereas this study concludes it has the most efficacy even in light of the well known "placebo" effect of the IA act. I remain unconvinced that the efficacy is an effect of the biologic rather than a result of the analysis.

Special Interest - Pain -- Physician

This is a well done systematic review and meta-analysis of the comparative effectiveness of pharmacologic interventions for osteoarthritis. One of the primary values of this article is to point out to the practitioner how complex the data are, i.e., there are relatively few direct head-to-head comparisons of active treatments. There are different durations of treatment and different outcomes - making comparisons even more complex and requiring the use of standardized effect sizes. Figure 1 and Table 1 are especially informative. As a geriatrician who takes care of many older adults with knee osteoarthritis, the take aways for me are that all active treatments (oral NSAIDS, and intra-articular steroids or hyaluronic acid) are better than placebo in improving pain at 3 months. Acetaminophen was not helpful. In addition, there's nothing to choose among the NSAIDS (celecoxib, ibuprofen, diclofenac) and intra-articular hyaluronic acid is better than oral NSAIDS.

Surgery - Orthopaedics

This review of medical therapy for OA is a useful summary, but limited by the quality of the studies that are available; therefore, the summary effect estimates must be interpreted with caution.

Surgery - Orthopaedics

The authors state that "intra-articular treatments were superior to non-steroidal anti-inflammatory drugs, possibly because of the integrated IA placebo effect." I believe that the placebo effect is not only possible but rather the most likely explanation for the `superiority` of IA treatments. As the IA placebo effect on pain is -22 (Supplement Figure 1) and both IA treatments show an effect of -29 (confidence intervals overlapping the placebo effect), this leaves little doubt about the placebo nature of this `treatment effect`.

Register for free access to all Professional content

Want the latest in aging research? Sign up for our email alerts.

Support for the Portal is largely provided by the Labarge Optimal Aging Initiative. AGE-WELL is a contributing partner. Help us to continue to provide direct and easy access to evidence-based information on health and social conditions to help you stay healthy, active and engaged as you grow older. Donate Today.

© 2012 - 2020 McMaster University | 1280 Main Street West | Hamilton, Ontario L8S4L8 | +1 905-525-9140 | Terms Of Use