Dr. Saam Morshed, a professor of orthopedic surgery at the University of California, San Francisco, who was not involved in the study, said, “I think it’s fair that we hold the mirror up to ourselves and scrutinize effectiveness for some of these operations. It’s important to understand where we have gaps in knowledge of the efficacy or nonefficacy of common surgical treatments.”

At the same time, he said, “It’s also important to understand that just because there isn’t a randomized trial supporting a given treatment, that doesn’t mean that the treatment is not effective.” Hip surgery, he said, is a good example. There may be no randomized trials of hip surgery, but there is overwhelming observational evidence for its effectiveness compared with nonsurgical treatment.

In other common procedures, the picture may be different. An arthroscopic operation to repair the anterior cruciate ligament, or ACL, in the knee, among the most common sports-injury surgeries in the United States, has a rate of success as high as 97 percent in some studies. But when the operation was compared with nonsurgical treatments, the review found, there was little difference in pain scores or the need for further surgical or nonsurgical treatment.

The researchers describe a large review of studies of the operation to repair the rotator cuff, the group of tendons and muscles that keeps the upper arm bone in the shoulder socket. Compared with exercise and steroid injections, the review found, there was little or no clinically significant difference in pain, function, quality of life or patient satisfaction with the results.

Some studies were randomized controlled trials, giving one group of patients real surgery and a matched group a placebo operation. In two such studies of surgery for shoulder impingement, a condition that causes pain on raising the arm, there was no difference between surgery and placebo surgery in patient-reported outcomes or adverse events.