To the Editor:

We thank Riddle, Harris, and Christensen1 for commenting on our editorial2 related to the Good Life With Osteoarthritis in Denmark (GLA:D) exercise and education program and the article by King et al.3 We would like to address systematically the concerns they raised.

There was a call to encourage stronger consideration of the causal evidence that examines the efficacy and effectiveness of GLA:D as an intervention for patients with hip or knee osteoarthritis (OA). We agree that single-arm prognostic studies such as that of King et al3 are not designed to be the sole basis of causal inference. We appreciate and agree that randomized controlled trials (RCTs) are important in informing clinical decision making. However, it is also important to recognize the value of findings that emanate from observational studies. The study by King et al3 raised the possibility that difficulty walking should be considered an alternative outcome measure to the usual outcome measure of pain in evaluating the efficacy of the GLA:D program. Future studies are needed to validate this as an appropriate outcome measure for interventions for knee OA. However, until then, the findings from King et al’s observational study that among > 2000 patients with moderate/severe difficulty walking at baseline, 51% and 58% of the participants later reported no/slight difficulty walking at 3- and 12-month follow-ups, respectively, is indeed, an impressive finding.3 In our editorial, we indicated that this finding “highlights the benefits of the GLA:D program.”2 Perhaps this is what Riddle and colleagues found troubling. We are happy to revise this statement to say that this impressive finding supports the possible benefit of the GLA:D program.

Riddle et al1 next raised an issue regarding the discussion of the RCT of GLA:D.4 We mentioned this study in our editorial2 because it has generated negative publicity about the GLA:D program. It was not our original intent to focus on the details of that study. However, since this question has been raised, we would like to take this opportunity to remind the readers of the context of that study.4 This was an open-label, RCT of GLA:D, an 8-week exercise and education program, compared with 4 intraarticular (IA) saline injections over 8 weeks. We respectfully assert that comparing GLA:D to IA saline injections is an inappropriate control group for this intervention. A comparison to a saline injection is not just comparing GLA:D to the “contextual and nonspecific effects (not specific treatment effects) of a placebo injection treatment”.1 IA injections have a profound effect on pain. Many recommended oral therapies fail to surpass the benefits of IA placebo injections,5,6 prompting some investigators to challenge the notion of identifying IA saline injection as a placebo at all.7 There has been substantial controversy surrounding the concept of what is the appropriate placebo for exercise and education interventions like GLA:D. We acknowledge that it is challenging, if not impossible, to identify a trustworthy “placebo” for programs like GLA:D. This may necessitate comparative effectiveness trials or observational studies to help inform practice guidelines.

We encourage readers to view RCTs as a common standard for causal evidence but also to appreciate that observational data can yield novel insights. Yes, these insights will warrant further study with a more robust design and analytic approach, but we should not dismiss insights from real-world data that can inform future research and clinical practice guidelines. This is especially true when the interventions may be low risk and readily deployed—like GLA:D—and when RCT designs fail to offer a robust comparison to the intervention or may be unethical.

FUNDING

GHL’s salary is supported in part by the Biology of Inflammation Center at Baylor College of Medicine. This work is supported in part with resources at the Veterans Affairs Health Systems Research and Development Center for Innovations in Quality, Effectiveness and Safety (#CIN 13-413), at the Michael E. DeBakey VA Medical Center, Houston, Texas. The perspectives detailed in this article do not reflect the views of Baylor College of Medicine or the Department of Veterans Affairs.

COMPETING INTERESTS

The authors declare no conflicts of interest relevant to this article.