Intention-to-Treat Analysis

Biostatistician analyzing randomized trial groups at a secure workstation

Intention-to-Treat Analysis Explainer

Intention-to-treat analysis compares randomized trial participants in the groups to which they were originally assigned, regardless of whether they completed, switched, or fully followed the assigned treatment. The principle preserves the starting comparison created by randomization. If people are moved between analysis groups based on what happened after assignment, the groups can become different in ways related to prognosis and the estimated treatment effect can be biased.

The approach often answers a treatment-policy question: what is the effect of assigning this strategy under the conditions of the trial, including real problems such as nonadherence or crossover? It does not necessarily estimate the biological effect among people who take every dose exactly as planned. Per-protocol or as-treated analyses ask different questions and can be informative, but simple exclusions can break randomization and need careful methods and interpretation.

Intention-to-treat is not a magic cure for missing data. If outcomes are unavailable, including a participant in the assigned group does not create the missing measurement. Researchers try to maintain follow-up, describe missingness, use prespecified methods, and perform sensitivity analyses. Terms such as modified intention-to-treat can hide different exclusions, so reports should state exactly who was analyzed and how post-randomization events were handled.

Readers should connect the analysis to the question. An intention-to-treat estimate can reflect the practical effectiveness of offering a treatment, while another estimand may address effects if treatment were taken as intended. Both require transparent assumptions. The enduring value of the principle is protection of the randomized comparison: assignment happens before outcomes and should not be rewritten in response to them.

Participants remain in their original randomized groups even if their treatment use later changes.

This helps preserve the balance that randomization was designed to create.

The estimate often describes the effect of assigning a treatment strategy under trial conditions.

It may differ from the effect among participants who adhered perfectly.

Assignment can be retained, but an unmeasured outcome remains missing.

Follow-up, prespecified methods, and sensitivity analyses are needed to assess the uncertainty.

Explore more "Explainers"

Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.