A study has found that surgery is no better than more conservative treatment to relieve knee pain caused by arthritis.

In the study, being published Thursday in The New England Journal of Medicine, 86 patients who had the operation fared no better over two years than 86 who had physical therapy and took medications to dampen inflammation.

The results of the study are in line with those from a study published in 2002. But experts are divided about what effects the two studies will have.

Some say the new study just confirms what they already knew. Others say they hope that doctors who did not believe the 2002 study will be persuaded by this one to stop doing the operations.

The 2002 study, by the Department of Veterans Affairs, had a different design: instead of assigning patients to surgery or medical treatment, it assigned them to real surgery or a sham operation. The real surgery was found to be no better than the sham one.

That study was denounced by many orthopedic surgeons, but Medicare decided in 2003 to stop paying for the operation. Still, because doctors can be reimbursed for the procedure by modifying what they say is the patient's problem, it is not clear whether most doctors stopped doing the operation, or how many such operations are being done. There is no national system for keeping track.

The surgery involves making small incisions in the knee, inserting an arthroscope to see the joint, and then flushing debris from the knee or shaving rough areas of cartilage and cleansing the joint.

It seemed to make sense that the debris and rough areas were contributing to knee pain, and when the department's study said the operation was useless, many simply did not believe it.

"What happened after our study was that organized orthopedics rallied the troops to try and discredit our study as much as possible," said Dr. Bruce Moseley, the 2002 study's principal investigator, who is now at the Richmond Bone and Joint Clinic in Texas. "People continued to practice the way they practiced."

But the federal Centers for Medicare and Medicaid Services were convinced.

"It was one of the very rare occasions that CMS actually narrowed coverage from its existing policy," said Dr. Sean Tunis, the centers' chief medical officer at the time. "The VA trial showing no benefit was very influential."

Since then, said Dr. Barry Straube, the current chief medical officer, the number of operations may have declined.

Medicare pays for the surgery under several cost codes. For one, arthroscopies that involved shaving knee cartilage, Medicare paid for 27,697 arthroscopies in 2002. In 2006, the number was 6,466, Straube said.

Dr. E. Anthony Rankin, president of the American Academy of Orthopaedic Surgeons, said most orthopedists appreciated the surgery's limitations. "As a tool for treating arthritis alone, it probably isn't a good tool," he said.

But others say the operation remains popular.

In fact, said Dr. David T. Felson, a professor of medicine and epidemiology at Boston University School of Medicine, the operation seems to have "become even more popular."

Dr. Brian G. Feagan, head of the clinical trials unit at the Fowler Kennedy Sport Medicine Clinic in London, Ontario, and an author of the new study, said Canadian doctors, too, were not much affected by the department's study.

"It really didn't change practice," Feagan said. Surgeons continued to believe in arthroscopy for arthritis pain, he added, and "they are doing a lot of it."

Now, with the new study, "I think practice will change," he said, adding, "It's pretty hard to ignore two studies that say the same thing."

Another study, also published Thursday in The New England Journal of Medicine, found that even when an MRI scan of an arthritis patient's knee showed damaged cartilage, that injury might have nothing to do with knee pain.

The study, led by Felson, involved 991 middle-age and elderly people in Framingham, Mass. Sixty-three percent of participants with knee pain from arthritis had a torn or destroyed meniscus, the wedge-shaped piece of cartilage that helps stabilize the knee. But 60 percent of those with arthritis but without knee pain also had a damaged meniscus.

"In patients with arthritis, almost everybody has meniscal tears," said Dr. Martin Englund of Boston University, the study's lead author. "We are so drilled to think, 'Oh, a meniscal tear — that must be painful,' or 'That's the cause of the pain.' But it may be involved in the disease process itself. There are many, many other reasons for pain in knee, but the meniscus is the structure we focus on and see."

Dr. Robert G. Marx, an orthopedist at Hospital for Special Surgery in Manhattan, who wrote an editorial accompanying the papers, cautions that there are different sorts of meniscal tears and that the decision on surgery can require clinical judgment.

"It can be very effective for patients who have osteoarthritis but are complaining from other problems in the knee, most commonly a large meniscal tear or a loose flap of cartilage," Marx said. "The challenge for the surgeon is to pick the patients appropriately." But others said they hoped the studies would persuade many orthopedists to be more judicious in their use of the surgery for arthritis.

"If it doesn't change care, it speaks poorly for the medical community's willingness to take evidence into account," Felson said.