Arthroscopic knee surgery for arthritis -- a procedure performed hundreds of thousands of times a year -- does not reduce joint pain or improve knee function, according to new research released Wednesday.
The study of 178 adults with moderate to severe arthritis found that the surgery, in which damaged bone and cartilage are removed through tiny incisions, had no benefits beyond nonsurgical treatments, such as physical therapy.
The report in the New England Journal of Medicine confirmed the results of a 2002 study that caused many orthopedic surgeons to discontinue the practice. Others continued to perform the surgeries, dismissing the earlier study because its subjects were predominantly men.
The latest report should put all doubt to rest, said Dr. Brian G. Feagan, a professor of medicine at the University of Western Ontario in Canada and a study author.
"We now have two data points saying the surgery is ineffective for arthritis," he said. "It should not be performed."
Arthroscopic knee surgery is a minimally invasive technique in which surgeons are guided by images transmitted from a tiny camera inserted into the joint. The method has a range of applications for which it is considered effective, such as repairing cartilage and ligament tears from sports injuries.
There were 985,000 arthroscopic knee surgeries in the United States in 2006, according to the most recent figures available from the national Centers for Disease Control and Prevention.
Feagan estimated that 200,000 to 300,000 of those surgeries were for arthritis. The surgery typically is performed as an outpatient procedure at a cost of about $7,000.
Arthritis, which usually begins after age 40, affects nearly 27 million Americans. In arthritis, the cartilage that protects bones at the joints breaks down, causing bones to rub together. Symptoms include pain and stiffness.
The study divided patients into two groups -- one receiving the surgery and the other serving as a control. People with large cartilage tears were excluded from the study.
All participants received one hour of physical therapy weekly for 12 weeks and were assigned to perform exercises twice daily at home. In addition, all patients used pain relievers as needed and were offered oral glucosamine and injections of hyaluronic acid, a knee lubricant. Patients were tracked for two years.
Those who received surgery showed greater improvement during the first three months, researchers said, but that advantage was gone after six months.
By the end of the study, there was no difference between the groups.
"The take-home message is that for this level of arthritis, surgery is not any better than nonsurgical treatment," said Dr. David McAllister, an associate professor of orthopedic surgery at UCLA's David Geffen School of Medicine, who was not involved in the research.
The study "is the nail in the coffin in the efficacy of the surgery," said Dr. David T. Felson, an epidemiologist at the Boston University School of Medicine.
Researchers said the results were not a complete disappointment. Both groups showed marked improvement in pain and mobility, providing evidence that physical therapy and medical treatments alone can be effective.
"I hope the remaining surgeons who are still scoping these types of knees do consider nonoperative care," said Dr. Robert B. Litchfield, a surgeon at the University of Western Ontario and a study author.
A second study in the journal, conducted by Felson and others, cautioned that the surgery may not be effective even when there is clear cartilage damage. In many cases, they found, this damage was the result of normal "wear and tear," and surgery wouldn't help.
"Identifying a tear in a person with knee pain does not mean that tear is the cause of the pain," Dr. Robert G. Marx of the Hospital for Special Surgery in New York wrote in an editorial accompanying the study.
Marx said that care should be individualized and that people with arthritis in their knees could still be candidates for surgery if it was determined that their pain was not from arthritis but from an injury.