In 1994, a orthopedic surgeon named J. Bruce Moseley enrolled patients with chronic knee pain into a clinical trial at a Houston VA hospital. All of them were told they would receive arthroscopic surgery — a common procedure in which a surgeon flushes debris from the knee joint and shaves damaged cartilage. Some patients received the actual operation. Others received something else entirely: Moseley made small incisions in their knees, manipulated the joint externally to simulate the sounds and sensations of surgery, irrigated the area with saline, and sewed the incisions closed. They woke up with bandaged knees, postoperative instructions, and no idea that nothing had been done. When Moseley followed up two years later, the patients who received sham surgery reported the same improvement in pain and function as those who received the real procedure. In some measures, they did slightly better.
The Moseley trial was not an isolated finding. A 2017 systematic review published in the BMJ examined 53 randomized controlled trials comparing surgical procedures against sham controls — fake incisions, simulated manipulations, anesthesia without intervention — across a wide range of conditions. In the majority of trials, the sham procedure produced outcomes statistically indistinguishable from the real one. Procedures that failed to outperform their placebos included some of the most commonly performed operations in the world: knee arthroscopy for osteoarthritis, vertebroplasty for spinal compression fractures, and shoulder surgery for certain rotator cuff conditions. These are not obscure procedures performed on rare patients. Knee arthroscopy alone is performed roughly 700,000 times annually in the United States.
The immediate temptation is to interpret these findings as proof that the procedures are useless. That conclusion is too simple. What the sham surgery trials actually demonstrate is something more complicated and, from an anthropological perspective, more interesting: the therapeutic encounter itself carries biological weight. The ritual of surgery — the hospital admission, the anesthesia, the incision, the recovery, the authority of the surgeon — produces measurable physiological effects independent of whatever mechanical correction the procedure was designed to perform. Patients are not imagining their improvement. Their pain genuinely decreases. Their function genuinely improves. The question is what, exactly, is producing that effect.
Medical anthropologists have long argued that healing cannot be reduced to its biochemical mechanisms. Every therapeutic encounter is also a social event, structured by roles, expectations, symbols, and relationships that carry meaning for the patient. Surgery is perhaps the most ritually elaborate of all medical interventions. The patient submits to unconsciousness in an environment of controlled sterility, surrounded by masked figures performing precise, coordinated actions with specialized instruments. The surgeon holds a form of authority that is almost sacerdotal in its cultural weight. When a patient emerges from that encounter having been told the problem has been fixed, the expectation of recovery is not a cognitive error — it is a reasonable response to a powerful ritual, one that the body, it turns out, takes seriously.
The ethical dimensions of this are genuinely unresolved. If a sham procedure produces outcomes equivalent to a real one, performing the real one exposes patients to unnecessary surgical risk — infection, anesthesia complications, recovery time, cost. That is a straightforward argument for abandoning procedures that fail sham-controlled trials. But the counterargument is equally uncomfortable: if the sham procedure is itself therapeutic, and if patients cannot receive the therapeutic benefit without believing they have received real surgery, then the sham procedure requires deception to function. Medicine has no framework for institutionalizing deception, even beneficial deception. Informed consent, the ethical foundation of modern clinical practice, depends on patients knowing what is being done to them.
Some researchers have proposed a partial resolution through what they call open-label placebos — patients are told explicitly that they are receiving a placebo, and improvement still occurs. The evidence for this is real but limited, and it does not map cleanly onto surgical contexts where the procedure's credibility is inseparable from its perceived mechanism. A more honest reckoning may require medicine to sit with a harder conclusion: that some of what surgery does has always been ritual, that the line between intervention and meaning is blurrier than surgical training acknowledges, and that a discipline built on biological mechanism has been, in part, practicing something closer to anthropology all along.