A Surgery Everyone Knows Works
If you are over 40 and your knee hurts, there is a decent chance someone will recommend arthroscopic surgery. A surgeon threads a camera and small instruments through two or three incisions, trims away torn meniscal tissue, flushes the joint with saline, and sends you home the same afternoon. About 700,000 of these procedures are performed annually in the United States alone, generating an estimated $4 billion in direct medical costs each year. Patients routinely report feeling better, orthopedic societies endorse the procedure, and insurance covers it without much argument.
One problem. The surgery is a placebo. When researchers in Finland tested the procedure the way pharmaceutical companies test drugs, by randomizing patients to either the real surgery or a convincing fake, the surgery did not outperform the placebo at any time point over a decade of follow-up.
What the FIDELITY Trial Found
Raine Sihvonen and Teppo Järvinen at the University of Helsinki enrolled 146 patients between 2007 and 2013 across five orthopedic clinics in Finland. All had persistent knee pain lasting at least three months that had not responded to conservative treatment, and all had MRI-confirmed degenerative medial meniscus tears without established osteoarthritis. If meniscectomy works for anyone, it should work for this population: torn meniscus, persistent symptoms, no arthritis complicating the picture.
Randomization happened in the operating room after every patient received a diagnostic arthroscopy confirming eligibility, and real surgery patients got a standard partial meniscectomy with a mechanized shaver and meniscal punches while sham patients got an elaborate performance in which the surgeon asked for all instruments, pushed a bladeless shaver against the patella outside the knee, manipulated the joint, ran the suction, and kept the patient in the operating room for the same duration as a real meniscectomy. The blinding held. Only 13% of sham patients guessed they had received sham surgery, identical to the guessing rate in the real surgery group.
At 12 months, every primary endpoint came back negative. The numbers tell the story. Lysholm knee scores improved by 21.7 points in the meniscectomy group versus 23.3 in the sham group. WOMET scores improved by 24.6 versus 27.1 points. Knee pain after exercise dropped by 3.1 versus 3.3 points. None of these differences reached statistical significance, and the confidence intervals excluded clinically meaningful effects. Both groups improved substantially, and both reported high satisfaction. Skin incisions, the sounds of instruments, and a surgeon manipulating the knee produced the same recovery as removing the torn tissue.
A Decade Later, Surgery Looked Worse
Skeptics argued benefits might emerge with longer follow-up, predicting that the meniscectomy group would stay stable while the sham group deteriorated as untreated tears worsened, and Sihvonen's team tracked the same patients for a full decade to find out, with 91% completing 10-year assessments.
Published in the New England Journal of Medicine in April 2026, the results showed the opposite. Surgery fared worse. Eighty-one percent of patients in the meniscectomy group showed radiographic progression of osteoarthritis versus 70% in the sham group, twelve percent of surgery patients had undergone knee replacement or osteotomy compared to 4% in the sham arm, and WOMET scores favored the sham group by 9.4 points. Removing meniscal tissue reduces the knee's shock-absorbing capacity, concentrating mechanical loads on a smaller contact area, and over a decade that biomechanical insult compounds.
Not an Isolated Finding
In 2002, J. Bruce Moseley and colleagues at the Houston Veterans Affairs Medical Center randomized 180 patients with knee osteoarthritis to arthroscopic debridement, arthroscopic lavage, or placebo surgery where surgeons made skin incisions and splashed saline without inserting instruments. Over 24 months, at no single time point did either surgical group report less pain or better function than placebo. The evidence was unambiguous.
Kirkley and colleagues at Western Ontario ran a similar trial in 2008 with 178 patients. Two-year WOMAC scores were nearly identical between surgery-plus-therapy and therapy alone. Then came the definitive synthesis: a 2022 Cochrane systematic review by O'Connor and colleagues pooled the randomized evidence and concluded that arthroscopic surgery for degenerative knee disease provides a clinically irrelevant improvement in pain of 2.4 mm on a 100-mm scale, while carrying real risks of deep venous thrombosis, infection, and pulmonary embolism.
Where Surgeons Push Back
Orthopedic surgeons have not accepted these findings quietly, and their strongest objection concerns patient selection. FIDELITY enrolled patients with degenerative tears and no established osteoarthritis, and surgeons argue the procedure remains appropriate for traumatic tears in younger patients or for specific mechanical symptoms like a displaced "bucket-handle" fragment physically blocking knee extension. A 2024 Norwegian trial found that meniscectomy outperformed sham in patients aged 35 to 55, though the sample was small and awaits replication.
That distinction matters, but it does not rescue the procedure for most recipients, because the vast majority of meniscal surgeries are not performed for locked knees but for the far more common scenario of a middle-aged person with chronic pain and an MRI showing a degenerative tear, a population for which four randomized trials, one Cochrane review, and a BMJ clinical practice guideline uniformly say the surgery does not help. Practice lags the evidence. Several non-orthopedic guideline bodies have recommended discontinuing it, while the AAOS and BASK continue to endorse it.
What We Didn't Prove
FIDELITY studied degenerative medial meniscus tears in adults 35 to 65 without established osteoarthritis, and the results cannot be extended to traumatic tears, lateral meniscus tears, or younger populations where tissue biology differs. All five sites were Finnish, and cultural factors around pain reporting and rehabilitation compliance may limit generalizability. Radiographic changes reported at 10 years do not always correlate with symptoms: some patients with severe X-rays experience minimal pain, and vice versa. Nine percent of each group was lost to follow-up, and although attrition was balanced, selective dropout could theoretically bias results.
The Bottom Line
Arthroscopic partial meniscectomy for degenerative meniscal tears is one of the most commonly performed surgeries on Earth, with an estimated $4 billion spent annually in the United States alone. It has now been tested against sham surgery in a double-blind trial with 10 years of follow-up, and it did not reduce pain, did not improve function, did not prevent osteoarthritis, and was associated with higher rates of joint replacement. For the majority of patients receiving this surgery, skin incisions and the sounds of instruments produce the same recovery as removing the torn meniscal tissue.
What You Can Do
If you have been told you need arthroscopic knee surgery for a degenerative meniscal tear, ask your surgeon three questions. First: is this a traumatic tear with mechanical locking, or a degenerative tear found on MRI? For degenerative tears, randomized evidence consistently shows no benefit. Second: have you tried structured physical therapy for at least 12 weeks? Kirkley's 2008 trial and the Cochrane review found therapy outcomes equivalent to surgery. Third: would the recommendation change without the MRI? Degenerative meniscal tears appear on imaging in roughly 35% of asymptomatic adults over 50, meaning the visible tear may be incidental. For clinicians, the 2017 BMJ guideline (doi:10.1136/bmj.j1982) provides a strong recommendation against arthroscopy for degenerative knee disease in nearly all patients.