J. Bruce Moseley, the Baylor College of Medicine orthopedic surgeon who performed the operations for the study, said he believed doctors should stop doing the procedure, called arthroscopic debridement, for people with arthritis. Orthopedic surgeons have been debating whether it works for about a decade, he said.
He added, however, that other types of arthroscopic surgery for the knee - those that repair certain cartilage or ligament tears or remove loose bits of bone or cartilage - were effective.
"This study is not an indictment of arthroscopic surgery as a whole," said Moseley, who is also team doctor for the Houston Rockets and Houston Comets basketball teams. "There are definitely some conditions that clearly get benefit from arthroscopic surgery."
Peter Sharkey, an orthopedic surgeon at Thomas Jefferson University Hospital, said he had questioned the value of arthroscopic surgery for people with arthritis for years. "This research pretty much puts the nail in the coffin," he said.
However, he and John D. Kelly IV, a Temple University orthopedic surgeon, said there may be times when people who have arthritis, but also have another mechanical problem, may benefit.
Douglas Jackson, a Long Beach, Calif., knee specialist who is past president of the American Academy of Orthopedic Surgeons, said he wondered if the study's participants - all men treated at a VA hospital - were representative of most people who get the operation.
Moseley said they were.
The only patients with arthritis who he thinks benefit from arthroscopic surgery are those with a specific type of tear called a bucket-handle tear and those with torn anterior cruciate ligaments.
Many arthritic patients want to try everything possible before resorting to a knee replacement. But Moseley said he believed the desire to avoid knee replacement was no reason to do a surgical procedure that he said had no benefit. In his experience, he said, most people end up getting a knee replacement, too.
Robert W. Jackson, chief of orthopedics at Baylor University Medical Center in Dallas, has sparred with Moseley in the past and remains a fan of the arthroscopic procedure. "I've done thousands of these in people with osteoarthritic knees, and they really are better," he said. The key is in choosing people in the relatively early stages of the disease, he said.
Moseley has been skeptical of arthroscopic debridement and lavage since he was a resident in the late '80s. "I just didn't quite understand why people were reporting so much benefit when seemingly there wasn't very much done," he said.
In the procedure, the surgeon debrides - scrapes or sands the surface of the joint - then lavages, washing away loose debris with saline solution. It is done through three tiny incisions with the aid of an arthroscope, a viewing instrument.
Over the years, Moseley wondered whether it might be lavage, not debridement, that was making patients feel better. But Nelda Wray, a pulmonary doctor who leads Baylor's Section of Health Services Research, suggested that the placebo effect might be at work. Moseley thought she might be right.
They conceived an unusual experiment to test their theory: 180 patients were divided into three groups. One received lavage only and one received debridement and lavage. The third group got placebo surgery that mimicked the real thing, including the three 1-centimeter incisions. All the patients stayed in the hospital overnight and were cared for by nurses who did not know which group they were in.
"We designed a rigorous trial that's going to literally end up taking on an industry that's worth more than a billion dollars a year," said Wray, lead researcher for the team from Baylor and Houston Veterans Affairs Medical Center.
She said the evidence was "simply overwhelming" that the arthroscopic procedure for people with arthritis was no better than fake surgery. She concluded that the money spent on such surgery "might be put to better use." Insurers should begin evaluating whether they want to continue covering the procedure, she said.
In addition, she said, the work shows there is a significant placebo effect in surgery.
Placebo groups are routinely used in trials of new drugs but rarely to evaluate operations.
Moseley said procedures that would be good candidates for placebo comparisons would be those done to relieve low back, neck, spine or neurological pain, tendinitis and impingement in the shoulder.
While it is intriguing that patients felt better after a fake operation, Wray said there was no way doctors could justify giving patients placebo procedures. Plus, even the placebo operation has some risk of complications, such as infection.
She believes the placebo effect comes largely from patients believing they will get well. If so, the main thing more research into the phenomenon may accomplish is to "improve the way doctors communicate with patients."
Patients in this study's placebo group were not angry when doctors told them the truth, she said. "Overwhelmingly, their attitude was: 'I don't care what you did to me. I am so glad I'm better.' "
Contact Stacey Burling at 215-854-4944 or firstname.lastname@example.org.