Intention-to-treat analysis

Intention-to-treat means a trial counts every person in the group they were first assigned to, even if they stopped the treatment or dropped out partway through.

This term is specific to research interpretation and is defined on AnxietyResearch. Shrinktionary, the network's language layer, doesn't currently carry a canonical entry for it.

It sounds odd to keep counting someone who quit. There is a good reason. People often drop out because the treatment didn't help or caused side effects. If a trial counted only the people who finished, it would quietly discard the bad results and make the treatment look better than it is. Intention-to-treat keeps the comparison honest.

Anxiety treatments, both therapy and medication, have real dropout rates. Medication can cause side effects. Therapy takes time and effort. An intention-to-treat analysis reflects what happens in ordinary conditions, not only among the people who completed everything.

A trial assigns 100 people to an SSRI. Twenty stop early because of side effects. An intention-to-treat analysis still counts all 100 in the SSRI group when measuring the result.

Its opposite is a per-protocol analysis, which counts only the people who completed the treatment as planned. Per-protocol results usually look rosier. When a trial reports both, the intention-to-treat number is the more trustworthy one for everyday decisions.