Sugar Pills That Worked With the Word Placebo on the Bottle

In late 2009, Ted Kaptchuk's team at Harvard handed 37 patients with irritable bowel syndrome a brown bottle. The label said "placebo pills" and instructed them to take two, twice daily. The clinicians explained, on the record, that the pills contained no active ingredient. The 43 patients in the control arm got nothing.

Three weeks later, the placebo group reported a 92-point improvement on the IBS Symptom Severity Scale. The no-treatment group reported 46 points. The effect size at endpoint was 0.79, large enough that Kaptchuk wrote it was comparable to results from trials of the best available IBS medications.

The finding broke a piece of textbook orthodoxy. Placebos were supposed to need deception to work -- the patient had to believe they were getting something real. Kaptchuk's group has since replicated the open-label result in chronic low back pain, episodic migraine, allergic rhinitis, cancer-related fatigue, and menopausal hot flashes. A 2021 follow-up in IBS found open-label placebo and double-blind placebo produced statistically indistinguishable improvements.

The usual caveat: this is mostly subjective, fluctuating symptoms -- the kind of conditions where reporting bias and regression to the mean live. Open-label placebos do not shrink tumors or replace insulin. What they keep doing is matching active drugs on patient-reported measures for the messy chronic conditions that fill primary-care waiting rooms. The pills don't seem to need the lie. They might just need the ritual.