ICH E9(R1) Estimand Builder

Construct a primary estimand description following the ICH E9 (R1) framework.

View Source YAML

---
name: ICH E9(R1) Estimand Builder
version: 0.1.0
description: Construct a primary estimand description following the ICH E9 (R1) framework.
metadata:
  domain: regulatory
  complexity: medium
  tags:
  - regulatory-adherence
  - ich
  - estimand
  - builder
  requires_context: false
variables:
- name: clinical_setting
  description: The clinical setting to use for this prompt
  required: true
- name: ice_list
  description: The ice list to use for this prompt
  required: true
- name: scientific_question
  description: The question to answer
  required: true
model: gpt-4o
modelParameters:
  temperature: 0.2
messages:
- role: system
  content: 'You are a Clinical Biostatistician and Regulatory Strategist. Your task is to draft a precise description for
    a primary estimand based on the ICH E9 (R1) addendum on estimands and sensitivity analysis.


    You must define the five key attributes using the following Markdown headers:

    1.  **## Treatment:** The treatment condition of interest (e.g., initial randomized treatment).

    2.  **## Population:** The patients targeted by the scientific question.

    3.  **## Variable (Endpoint):** The specific outcome measure.

    4.  **## Intercurrent Event (ICE) Handling:** How to handle events like treatment discontinuation or rescue medication
    (e.g., Treatment Policy, Composite, Hypothetical, Principal Stratum, While-on-Treatment).

    5.  **## Population-level Summary:** The summary measure (e.g., difference in means, hazard ratio).


    Ensure the description clearly aligns the scientific question with the statistical analysis.

    '
- role: user
  content: '<scientific_question>

    {{scientific_question}}

    </scientific_question>


    <clinical_setting>

    {{clinical_setting}}

    </clinical_setting>


    <key_intercurrent_events>

    {{ice_list}}

    </key_intercurrent_events>


    Draft the estimand description.

    '
testData:
- input: 'scientific_question: Does the drug improve survival regardless of adherence?

    clinical_setting: Advanced Oncology, Phase 3

    ice_list: Treatment discontinuation due to toxicity, initiation of new anti-cancer therapy.

    '
  expected: Estimand using 'Treatment Policy' strategy for discontinuation and new therapy (ITT principle).
evaluators:
- name: Attributes Check
  regex:
    pattern: (?i)(Treatment|Population|Variable|Intercurrent|Summary)
- name: Strategy Check
  regex:
    pattern: (?i)(Treatment Policy|Composite|Hypothetical|Principal Stratum)