The Operation Everyone Believed In
The logic seemed airtight. An arthritic knee hurts because the joint is full of debris, frayed cartilage, loose fragments, a torn meniscus snagging with every step, and the standard fix is to slide a camera into the joint, wash it out with liters of saline, shave the rough cartilage smooth, trim the torn meniscus, and watch the pain ease. By 2002, American surgeons performed about 650,000 of these procedures a year for knee osteoarthritis, at roughly $5,000 each, and in uncontrolled reports about half the patients said they felt better.
The logic had a hole nobody had tested: no one had ever compared the operation against a fake version of itself. Every pill must beat a placebo before reaching the market, yet surgery had never faced that test: cutting someone open as a control was unthinkable. An orthopedic surgeon at Baylor College of Medicine decided to think it anyway.
Three Groups, One Secret
Between October 1995 and September 1998, J. Bruce Moseley and his colleagues recruited 180 patients with knee osteoarthritis at the Houston Veterans Affairs Medical Center, average age 52.3, and randomly assigned each to one of three groups. The first got arthroscopic débridement, the joint washed with at least 10 liters of fluid, rough cartilage shaved, loose debris removed, and torn meniscus trimmed, while the second got lavage alone, the washout and nothing else, and the third got placebo surgery, prepped and draped, given three 1-centimeter incisions, the surgeon asking for the instruments, manipulating the knee, and miming the procedure while no arthroscope ever entered the joint.
One surgeon performed every operation, and neither patients nor assessors knew which group anyone was in. Consent was blunt: each volunteer wrote in their chart that they understood they might receive only placebo surgery, which would not benefit their arthritis. Even so, 144 of 324 eligible patients, 44 percent, declined. Pain and function were measured repeatedly over two years, the primary outcome the Knee-Specific Pain Scale, a 100-point measure, at 24 months.
The Results That Embarrassed a Profession
At no point did either surgical group report less pain or better function than the placebo group. At one year, pain scores were 48.9 in the placebo group, 54.8 in the lavage group, and 51.7 in the débridement group, and at two years the numbers were 51.6, 53.7, and 51.4, differences that were not significant at any time point, with p-values of 0.14 and 0.51 at one year and 0.64 and 0.96 at two years.
All three groups improved. Pain fell by 6 to 12 points within the first two weeks, then eased a little more over two years, in the real groups and the fake one alike. The improvement patients had been crediting to the operation for decades was real, but it was not caused by the operation. Expectation, the ritual of surgery, the anesthesia, the incisions, and the natural fluctuation of pain together produced the relief the scalpel was supposed to deliver.
Now run the arithmetic: at 650,000 procedures a year and roughly $5,000 each, the annual bill comes to about $3.25 billion, and the oft-quoted figure that half of patients report relief matches the placebo arm's improvement almost exactly, which suggests the quoted success rate was the placebo rate. Nor was the trial risk-free theater: two patients had minor complications, and the placebo group also went under anesthesia, so even the fake carried real risk.
The response was swift, and within weeks the Department of Veterans Affairs directed that arthroscopic procedures for osteoarthritis not be done in the VA system without expert-panel review. Kirkley and colleagues randomized 178 patients in 2008 to arthroscopy plus physical therapy or therapy alone and found no benefit from adding the operation, while Sihvonen and colleagues ran a second sham-controlled trial in 2013, for degenerative meniscal tears, with 146 patients, and again found no difference between real and fake surgery. A 2015 BMJ meta-analysis found the benefit small and inconsequential, and the American Academy of Orthopaedic Surgeons recommended against the procedure in 2013.
The Strongest Counterargument
The strongest case against this result starts with who was studied: one surgeon, one hospital, one mostly male veteran population. Critics writing in the Baylor proceedings in 2003 argued that the trial's patient selection missed the point: success with arthroscopy depends on selectivity, and a surgeon choosing the right knees might get results a trial of unselected patients cannot see. The trial also excluded the most severe disease, and its pain scale was a 12-item instrument created for this study, not a standard measure.
A second objection concerns scope: Moseley tested débridement and lavage for osteoarthritis, not arthroscopy for a traumatic meniscal tear, a locked knee, or a young athlete's injury, and the finding is sometimes stretched beyond what it showed. A third objection is ethical: sham surgery exposes people to anesthesia and incisions for no benefit, and some ethicists say such trials should never run, even with written consent.
There is also a subtlety the headline version misses. The improvement in the placebo group cannot be cleanly separated into placebo effect and natural history, since arthritic pain fluctuates and people tend to seek treatment when pain is at its worst, so some of the gain would have happened with no treatment at all, which means the benefit reported in decades of uncontrolled case series was likely expectation and regression to the mean, not evidence the operation works. That cuts both ways. None of these objections restores the operation for osteoarthritis. They narrow the claim to what the evidence supports: for the average patient with an arthritic knee, the surgery adds nothing to the ritual around it.
What We Didn’t Prove
This was one trial, at one center, with one surgeon and a mostly male veteran population. It does not settle the question for every patient, and it tested 2002 procedures, not today's techniques. It says nothing about arthroscopy for a traumatic meniscal tear, a knee that locks, or an injury in a young athlete, and it cannot prove that no individual patient benefits. Fifteen of the 180 patients did not complete the two-year follow-up. The trial cannot tell us how much of the placebo group's improvement was expectation and how much was the natural waxing and waning of pain.
The Bottom Line
Fake knee surgery worked as well as the real thing. In 180 patients followed for two years, neither washing out the joint nor shaving cartilage beat three small incisions and a pantomime. Two later sham-controlled trials agreed. The relief patients felt was real. The operation was not its cause. For an arthritic knee, try exercise, weight management, and physical therapy first. Treat the scalpel as a last resort, not a first step.
What You Can Do
If a surgeon recommends arthroscopy for osteoarthritis, ask what the operation adds over physical therapy: two sham-controlled trials and a meta-analysis say close to nothing. If your knee locks, catches, or gives way, that is a different problem from osteoarthritis, so get the diagnosis precise before deciding anything. Start with what the evidence supports: structured exercise, quadriceps strengthening, weight loss if you carry extra weight, and a topical anti-inflammatory. If you do have surgery, remember that much of the relief may come from the ritual itself. Plan your rehabilitation as if the exercise, not the incision, is doing the work.
Sources
- Moseley, J. B., O'Malley, K., Petersen, N. J., Menke, T. J., Brody, B. A., Kuykendall, D. H., Hollingsworth, J. C., Ashton, C. M., & Wray, N. P. (2002). A controlled trial of arthroscopic surgery for osteoarthritis of the knee. New England Journal of Medicine, 347(2), 81-88. doi:10.1056/NEJMoa013259
- Felson, D. T., & Buckwalter, J. A. (2002). Débridement and lavage for osteoarthritis of the knee. New England Journal of Medicine, 347(2), 132-133. doi:10.1056/NEJMe020055
- Kirkley, A., Birmingham, T. B., Litchfield, R. B., Giffin, J. R., Willits, K. R., Wong, C. J., Feagan, B. G., Donner, A., Griffin, S. H., D'Ascanio, L. M., Pope, J. E., & Fowler, P. J. (2008). A randomized trial of arthroscopic surgery for osteoarthritis of the knee. New England Journal of Medicine, 359(11), 1097-1107. doi:10.1056/NEJMoa0708333
- Sihvonen, R., Paavola, M., Malmivaara, A., Itälä, A., Joukainen, A., Nurmi, H., Kalske, J., Ikonen, A., Järvelä, T., Järvelin, T., Kanto, K., Karhunen, J., Knifsund, J., Kräg, A., Kinnunen, J., Laitinen, T., Lehtinen, J., Nylund, J., Reunanen, M., … Järvelä, S. (2013). Arthroscopic partial meniscectomy versus sham surgery for a degenerative meniscal tear. New England Journal of Medicine, 369(26), 2515-2524. doi:10.1056/NEJMoa1305189
- Thorlund, J. B., Juhl, C. B., Roos, E. M., & Lohmander, L. S. (2015). Arthroscopic surgery for degenerative knee: systematic review and meta-analysis of benefits and harms. BMJ, 350, h2747. doi:10.1136/bmj.h2747
- American Academy of Orthopaedic Surgeons. (2013). Treatment of osteoarthritis of the knee: Evidence-based guideline (2nd ed.). Strong recommendation against arthroscopy with lavage and/or débridement in patients with primary symptomatic osteoarthritis of the knee.
- Debate on the use of arthroscopic surgery for osteoarthritis of the knee. (2003). Proceedings (Baylor University Medical Center), 16(1), 27-29. doi:10.1080/08998280.2003.11927885