Return to 2001 Abstacts List

40. Arthroscopic Surgery versus Placebo Arthroscopy in the Treatment of Osteoarthritis of the Knee: Results of a Randomized, Double Blind, Placebo-Controlled Trial

NP Wray, Houston VA Medical Center; JB Moseley, Houston VA Medical Center; K Wristers, Houston VA Medical Center; N Petersen, Houston VA Medical Center; T Menke, Houston VA Medical Center; RB Giesler, Indiana University School of Nursing; B Brody, Baylor College of Medicine; CM Ashton, Houston VA Medical Center; AJ Greisinger, Houston VA Medical Center

Objectives: One third of Americans over age 65 have osteoarthritis (OA) of the knees. Multiple studies report that arthroscopy reduces pain and improves function in OA patients. However, none of these studies controlled for the placebo effect. There is broad evidence of a placebo effect in the treatment of pain. We questioned whether at least some of the improvement in knee pain previously reported to be due to arthroscopy might be explained by the placebo effect. We therefore undertook a randomized, double blind, placebo-controlled trial to determine whether arthroscopic surgery has any benefit greater than placebo arthroscopy in relieving knee pain due to osteoarthritis.

Methods: Patients with knee OA were recruited from the VA Medical Center in Houston, Texas. Patients met the American College of Rheumatology definition of knee OA, and had at least moderate knee pain. Patients were stratified into three groups of radiograph severity: mild, moderate, or severe. Using a stratified random process, patients were assigned to arthroscopic debridement plus lavage, arthroscopic lavage alone, or placebo arthroscopy.

Data were collected at eight occasions. Knee pain was assessed using the Wisconsin Brief Pain Questionnaire, Arthritis Impact Measurement Scale (AIMS 2) pain subscale, and SF-36 pain scale. Functioning was assessed with the Physical Functioning Scale, AIMS 2, and SF-36 Functioning Scale. Patient satisfaction was assessed with the Satisfaction with Knee Procedure Scale.

Results: Participants (n=180) were slightly younger (p=0.02), more likely to be white (p=0.03), and had more severe radiographic severity of OA (p<0.001) than non-participants (n=144). Among participants, there were no significant differences at baseline between the three treatment groups.

For pain, function, and satisfaction, there were no significant differences by treatment group at 6 weeks, 3 months, 6 months, 1 year, 18 months, and 2 years. However, two weeks after the procedure, placebo patients reported statistically less pain (p<.05), better functioning (p<.05), and greater satisfaction (p<.05) than patients who received arthroscopic surgery. In addition, longitudinal analyses demonstrated no significant differences in pain, function, and satisfaction by treatment over time. Compared to baseline, all three groups showed statistically significant improvements in pain, function, and satisfaction at 6 weeks, 3 months, 6 months, 1 year, 18 months, and 2 years.

Conclusions: All three groups showed improvement that was still apparent at 2 years. However, placebo had as great an effect as arthroscopic surgery in relieving pain and improving function. In this randomized, double blind, placebo-controlled trial, arthroscopic surgery had no benefit greater than placebo arthroscopy in relieving knee pain and improving function due to osteoarthritis.

Impact: Knee arthroscopy for OA of the knee is approximately a six billion-dollar per year industry. This study casts doubts on whether this procedure has greater value than a placebo. In addition, this study demonstrates the substantial placebo effect that can occur with surgical interventions and should lead to a debate on the appropriate role of placebo-controlled trials to investigate the efficacy of surgical procedures.