Posted on: Thursday, July 11, 2002
Popular surgery on knees disputed
By Susan Okie
The Washington Post
An operation performed about 300,000 times annually on U.S. patients with arthritis is completely ineffective, according to an unusual new study that compared the procedure with phony surgery.
The study by researchers at the Houston Veterans Affairs Medical Center compared two versions of arthroscopic knee surgery for osteoarthritis with a placebo operation in which patients were given a sedative and only received small skin incisions. All patients reported reduced pain and better knee function, but there was no difference in outcome between those who had real surgery and those who got the placebo procedure.
Comparing arthroscopic knee surgery with a sham operation was the only way to find out whether the procedure, which costs about $5,000, was helping people with osteoarthritis, said surgeon J. Bruce Moseley of Baylor College of Medicine, who operated on all of the patients.
"There are a lot of people who do this surgery" for osteoarthritis, said Moseley, who reported the findings in the New England Journal of Medicine. "I would like to think ... they would (now) change the way they practice. Now that I know that really nobody is a good candidate, I won't do it."
Surgeons began doing the procedure to try to relieve pain in arthritis and it became increasingly popular after doctors published reports of hundreds of cases where patients reported improvement.
But, like most surgical procedures, the operation was never before compared to anything that could rule out a placebo effect, the improvement that people get from a treatment because they expect it to work. Placebo controlled studies of surgical procedures are rare because it is difficult to devise sham operations that fool participants, and because researchers cannot ethically subject people to phony operations that carry significant risks.
In this case, researchers convinced patients they had had surgery without subjecting them to more than minimal risks. The 180 participants did not know until the end of the two-year study which treatment they received, but they were told in advance that they might not get real knee surgery and had to write out a statement showing that they understood. Forty-four percent of patients who were invited to participate refused, said Baylor ethicist Baruch Brody.
"We were happy about that," he said. "It was real evidence that people were understanding. ... If the refusal rate had gotten too low, we were going to stop the trial."
The study's results are "compelling" and should encourage similar tests of other popular operations for subjective symptoms such as pain, said Carolyn Clancy, acting director of the federal Agency for Health Care Research and Quality. Such research can save lives and help other patients "avoid risks of surgery that's not actually doing a lot," she said.
The study tested two kinds of arthroscopic knee surgery for osteoarthritis. Both are done through small incisions by a surgeon who views the inside of the joint through an instrument called an arthroscope. In one type, known as lavage, the surgeon uses large quantities of saline to wash particles and inflammatory enzymes out of the joint. In the other, called debridement, the surgeon also uses instruments to smooth out rough joint cartilage and remove loose debris.
Study participants were randomly assigned to the three treatments. Those receiving real surgery had general anesthesia. Those in the placebo group received an intravenous drug to make them sleep and a local anesthetic for the skin incisions. Moseley kept each patient in the operating room for the normal length of an operation, called for various instruments and moved the leg around as he would during the real procedure.
When questioned about knee pain, arthritis pain and total body pain during the two-year follow-up period, all patients reported the same degree of improvement. On objective tests measuring their ability to walk and climb stairs, there was no significant improvement in any group.
