ICH M15
General Principles for Model-Informed Drug Development
Standard Introduction
ICH M15 is the final harmonised guideline for planning, evaluating and documenting evidence from model-informed drug development. ICH Regulatory Members adopted the Step 4 text on 29 January 2026; FDA issued final U.S. guidance on 3 June 2026, and the EU Step 5 version became legally effective on 23 July 2026.
The guideline provides a common framework for defining the decision question and context of use, assessing model influence, the consequence of a wrong decision, model risk and model impact, and documenting model evaluation through an assessment table, Model Analysis Plan and Model Analysis Report. Other jurisdictions may use different implementation dates.
Harmonised Evidence Framework
Structures assessment around the question of interest, context of use, model influence, consequence of a wrong decision, model risk and model impact.
Risk-Proportionate Model Evaluation
Connects model risk to verification, validation, applicability assessment, robustness, uncertainty and other evidence needed to support the intended decision.
Planning and Submission Records
Defines the roles of the assessment table, Model Analysis Plan, Model Analysis Report and regulatory-interaction or submission documentation.
list_alt Core ICH M15 Evidence Elements
- Question of interest
- Context of use
- Model influence
- Consequence of wrong decision
- Model risk
- Model impact
- Predefined technical criteria
- Verification, validation and applicability assessment
- Model Analysis Plan and Model Analysis Report
- Regulatory interaction and submission documentation
Who Needs to Comply?
Drug and biologic developers, quantitative pharmacology and pharmacometrics teams, statisticians, data scientists, clinical and nonclinical experts, regulatory-affairs teams, quality functions and decision-makers who plan, evaluate, document or submit evidence produced through modelling and simulation. Applicability and legal effect depend on adoption by the regulator for the target market.
Key Requirements
Frame the Regulatory Question
State the question of interest and define a concise context of use covering the model, its role and scope, the data used to build it, and other evidence that will inform the decision.
Assess Influence, Consequence, Risk and Impact
Describe, rate and justify model influence, the consequence of a wrong decision, model risk and model impact. Use separate assessment tables when different questions of interest need distinct evaluations.
Predefine Technical Criteria
Specify question-specific criteria for evaluating the model and outcomes, explain why the proposed MIDD approach is appropriate, and make the criteria commensurate with model risk.
Evaluate the Model
Document verification, validation and applicability assessment. Address data relevance, assumptions, bias, uncertainty, robustness, performance, method-specific issues and limitations as applicable to the model and decision.
Plan and Report the Analysis
Use a Model Analysis Plan to predefine objectives, data, methods and criteria, then a Model Analysis Report to document conduct, results, evaluation, deviations, uncertainty and conclusions.
Prepare Regulatory Evidence
Provide the completed assessment table, evaluation summary, outcome of the MIDD evidence assessment and references to relevant plans and reports in interactions and submissions; seek early regulator alignment when appropriate.
Implementation Roadmap
Define the decision and team
State the development or regulatory question, decision owner, target jurisdictions and lifecycle stage. Bring together modelling, statistics, clinical or nonclinical, regulatory, data, quality and subject-matter expertise.
Complete the assessment framework
Define the context of use, rate and justify model influence and consequence of a wrong decision, derive model risk, assess model impact and record the result in the ICH assessment table.
Plan evaluation and regulator interaction
Predefine technical criteria and the model evaluation strategy in the Model Analysis Plan. Identify assumptions, data needs, uncertainty and points where early regulator alignment would reduce decision risk.
Execute and document
Develop and evaluate the model using reliable, reproducible and traceable processes. Document verification, validation, applicability, data handling, performance, robustness, uncertainty, limitations and deviations in the Model Analysis Report.
Assess evidence and submit
Update the assessment table with completed evaluation and outcomes, explain how the evidence answers the question, cross-reference plans and reports, and preserve a reviewable record for regulatory interactions and submissions.
Compliance Checklist
checklist Question and context
checklist Risk and impact
checklist Evaluation and records
checklist Regulatory communication
Which ICH M15 record answers which question?
Use the assessment table for decision context, the MAP for prospective planning and the MAR for completed analysis evidence.
| Record | Primary question | Timing |
|---|---|---|
| MIDD evidence assessment table | What decision is being supported, and what are the influence, risk, impact and assessment outcome? | Planning, interaction and submission |
| Model Analysis Plan (MAP) | What data, methods, assumptions and technical criteria will govern the analysis? | Before analysis |
| Model Analysis Report (MAR) | What was done, what did the model show, and how well did it meet the criteria? | After analysis |
| Regulatory interaction or submission document | How does the complete MIDD evidence support the regulatory question and decision? | At interaction or submission |
Common Misconceptions
Any sophisticated model automatically produces acceptable MIDD evidence.
Acceptability depends on the question, context, risk-proportionate evaluation, technical criteria, documentation and the regulator's assessment—not model complexity alone.
Model risk means the model is technically poor.
In M15, model risk describes the model outcome's contribution to a possible wrong decision and its consequences. A well-built model can still support a high-risk decision that requires stronger evaluation.
The Step 4 adoption date is one global legal deadline.
Step 4 finalises the harmonised guideline. Each ICH region implements it under its own procedures and dates.
Penalties & Enforcement
ICH M15 is a harmonised guideline, not a certification standard or stand-alone penal law. It does not create direct fines. Consequences arise through each regulator's implementation and the quality or acceptability of a development programme, regulatory interaction or submission; organizations must check the current rules and guidance in every target jurisdiction.
Frequently Asked Questions
What is ICH M15?
expand_more
ICH M15 is the harmonised guideline for assessing, evaluating and documenting evidence generated through model-informed drug development. ICH Regulatory Members adopted the final Step 4 text on 29 January 2026.
Is ICH M15 legally binding everywhere?
expand_more
No. Step 4 is the final harmonised ICH text, but each regulator implements it through its own Step 5 procedure. FDA issued final guidance on 3 June 2026, while the EU version became legally effective on 23 July 2026. Other markets must be checked separately.
Which modelling methods are in scope?
expand_more
The guideline applies to current and emerging modelling and simulation approaches. Its examples include population PK/PD, PBPK and biopharmaceutics, exposure-response, model-based meta-analysis, quantitative systems pharmacology and toxicology, agent-based and disease-progression models, and AI or machine-learning models.
What are model risk and model impact?
expand_more
Model risk combines the model's influence on the decision with the consequence of a wrong decision and helps determine evaluation needs. Model impact describes how far the proposed MIDD strategy varies from current regulatory standards or expectations. Both require a justified low, medium or high rating.
What is the difference between a MAP and a MAR?
expand_more
The Model Analysis Plan predefines the analysis objectives, data, methods, evaluation approach and technical criteria. The Model Analysis Report records what was done, results, evaluation, deviations, uncertainty, limitations and conclusions.
Does ICH M15 specify every technical method for model development?
expand_more
No. It provides general assessment, evaluation, interaction and reporting principles. It says model development should also follow accepted standards and scientific practices for the specific method, and it should be used with relevant topic-specific ICH guidelines.
Can an organization be certified to ICH M15?
expand_more
No. ICH M15 is regulatory guidance, not an organizational certification standard. Regulators evaluate the evidence and documentation in the context of development programmes, interactions and submissions.
Official Documentation
Official PDF for ICH M15
Official publication or summary for ICH M15
Official online resource
International Council for Harmonisation (ICH) guidance and reference material
Implementation toolkit
Templates, guidance, or companion resources for ICH M15