Clinical trials often examine the effectiveness of a treatment outside of real world contexts. For instance, if a medicine is very effective, but has severe side effects, this likely will reduce adherence and can make the medicine less effective in the real world. The Economics 2.0 book looks at when individuals are most satisfied with a colonoscopy:
“[An] experimental was conducted with colonoscopies at a time when sedation was not yet customary. With half of the patients, the doctor left the instrument inside the colon at the end of the examination for one extra minute, without moving it. This was unpleasant, but much less painful than the colonoscopy itself. It turned out that patients who had this done to them later recalled the overall examination as less unpleasant than other test subjects who had the instrument removed earlier. Also, they were more likely to show up for follow-up examinations.“