Study offers alternative explanation to placebo effect in assessing pain in OA
Click Here to Manage Email Alerts
The placebo effect on pain scores in osteoarthritis may be attributed to a phenomenon called “regression to the mean,” according to a speaker at the 2021 OARSI World Congress.
“The results of randomized controlled trials of potential disease-modifying osteoarthritis drugs (DMOADs) often include a precipitous decline in the mean pain score for the active treatment arm and the placebo arm during follow-up,” Erin Ashbeck, MPH, MS, assistant scientific investigator in the Arthritis/Musculoskeletal Population Health Program at the University of Arizona Health Sciences, said in her presentation. “The reduction in pain scores within the placebo arm is often attributed to the placebo effect.”
In the current study, the group aimed to offer an “alternative explanation,” largely in the form of so-called “regression to the mean,” according to Ashbeck.
“When participants are selected based on high (low) values of a longitudinal outcome measure that naturally fluctuates over time, the mean can be expected to decrease (increase),” Ashbeck and colleagues wrote. “This statistical phenomenon is known as regression-to-the-mean.”
Eligible participants were culled from the Osteoarthritis Initiative, a longitudinal observational study of participants with or at risk for symptomatic knee OA. After completing the Western Ontario and McMaster Universities Osteoarthritis Index questionnaire, participants underwent imaging analysis using bilateral posteroanterior fixed-flexion weight-bearing knee radiography with joint space width measurement.
The analysis included 1,837 patients with Kellgren-Lawrence (KL) grade 2 or 3 OA in the right knee who were grouped according to baseline WOMAC pain score. Groups were WOMAC 0, 1, 2-4, 5-7 or 8-20.
Results showed that individuals with WOMAC pain scores of eight or higher at baseline experienced lower average pain scores at the first year. “The average pain score dropped by about three points at the next annual clinic visit,” Ashbeck said.
Conversely, patients who had a WOMAC pain score of zero at baseline showed an increase in pain score at the 1-year visit, according to Ashbeck.
Out of possible concern that the baseline visit might have skewed the findings, the clinicians eliminated that pain score from the analysis. The trend persisted, both with pain scores higher than eight at 1 year decreasing and those that were zero at 1 year increasing. Moreover, the trend continued to persist when both the baseline and 1-year scores were eliminated. “Again, we see regression to the mean for knees with extreme pain scores, high or low,” Ashbeck said.
When the researchers assessed the same cohort for medial fixed joint space width (fJSW), no such regression to the mean phenomenon was reported.
In closing, Ashbeck offered some insight for researchers designing clinical trials assessing pain scores in OA, and for the rheumatologists who interpret the results.
“In randomized controlled trials of potential DMOADs, eligibility often requires a pain score greater than some value, like WOMAC pain score of eight or higher,” she said. “And we know that pain naturally fluctuates over time. These are the two ingredients to reliably produce regression to the mean.”
Collapse
Ashbeck E. The “Placebo Effect” in Osteoarthritis Clinical Trials: Challenging the Narrative. Presented at: OARSI 2021 World Congress on Osteoarthritis; April 29-May 1. (virtual meeting).
Read more about
Click Here to Manage Email Alerts