Back to Glossary
Research Definition

Intention-to-Treat

Also known as: ITT, ITT analysis, Intent-to-treat

Intention-to-Treat is a clinical trial analysis method where all participants are analyzed in the groups to which they were originally randomized, regardless of whether they completed treatment, switched groups, or adhered to the protocol. ITT analysis preserves randomization benefits and reflects real-world effectiveness.

Last updated: January 28, 2026

How Intention-to-Treat Works

The ITT Principle

“Once randomized, always analyzed”

SituationITT Approach
Patient stops taking drugStill counted in drug group
Patient switches to other armCounted in original group
Patient lost to follow-upIncluded with imputation
Patient violates protocolRemains in assigned group

Why ITT Matters

Preserves Randomization Randomization creates comparable groups. Excluding patients based on post-randomization events can reintroduce bias and make groups incomparable.

Reflects Reality In real practice, patients:

  • Skip doses
  • Stop medications
  • Don’t follow instructions perfectly

ITT results better predict real-world effectiveness.

Relevance to Peptides

ITT in GLP-1 Agonist Trials

STEP 1 Analysis Populations

PopulationDefinitionWeight Loss
ITTAll randomized-14.9%
Per-protocolCompleted as designedHigher
CompletersFinished studyHigher still

Why Different Populations Show Different Results

Patients who:

  • Stop due to side effects: Often had less weight loss
  • Drop out early: Weight regain not captured
  • Complete study: Generally better adherence, better results

ITT provides the most conservative (and realistic) estimate.

Handling Missing Data

MethodApproach
Last observation carried forwardUse last measured value
Multiple imputationStatistically estimate missing values
Mixed modelsAccount for all available data
Worst-case scenarioAssume no benefit for dropouts

ITT vs Per-Protocol Analysis

Comparison

AspectITTPer-Protocol
IncludesAll randomizedProtocol completers only
PreservesRandomizationTheoretical drug effect
EstimatesEffectivenessEfficacy
Bias riskLowerHigher
Real-world relevanceHigherLower

When Each is Used

  • ITT: Primary analysis for regulatory decisions
  • Per-protocol: Supportive analysis showing maximal effect
  • Both reported: Provides complete picture

Understanding Trial Reports

What to Look For

“In the intention-to-treat population (n=1,961), mean weight change was -14.9% for semaglutide vs -2.4% for placebo”

This tells you:

  • All randomized patients included
  • Results represent real-world expectation
  • Drug effect is 12.5 percentage points

Comparing ITT to Completers

If ITT shows 15% weight loss but completers show 18%:

  • 15% is more realistic expectation
  • 18% represents those who tolerate and adhere
  • The 3% gap reflects dropout impact

Frequently Asked Questions

Isn’t it unfair to count patients who stopped taking the drug?

ITT actually provides a fairer test. If a drug has intolerable side effects causing many patients to stop, that’s important information. Excluding those patients would make the drug look better than it performs in reality. ITT captures both efficacy and tolerability.

Which results should I believe - ITT or per-protocol?

ITT is generally more reliable for predicting real-world outcomes. Per-protocol shows what happens under ideal conditions. For making treatment decisions, ITT is more relevant because it accounts for the reality that not everyone tolerates or adheres to treatment perfectly.

Related Peptides

Related Terms

Disclaimer: This glossary entry is for educational purposes only and does not constitute medical advice. Always consult a qualified healthcare provider for medical questions.