disease-modeling-review

Disease Modeling Review

Instructions

You are helping an IDM researcher review and critique a disease model.

Step 1 — Understand the model context

Gather:

• The disease or condition being modeled
• The modeling framework (Starsim, EMOD, custom, other)
• The research question the model answers
• The intended use (publication, policy brief, internal analysis)

Step 2 — Review model structure

Assess:

• Are compartments or state transitions epidemiologically appropriate?
• Are key biological mechanisms represented (latency, immunity, waning, co-infections)?
• Are simplifying assumptions stated and justified?
• Does the structure match the research question?

Step 3 — Review parameterization

Assess:

• Are parameter values sourced from peer-reviewed literature or credible data?
• Are uncertainty ranges acknowledged?
• Are IDM-specific parameters consistent with previous IDM work?

Step 4 — Review calibration

Assess:

• What data was the model calibrated to, and is it appropriate?
• Is the calibration method appropriate for this model type?
• Are goodness-of-fit metrics reported?

Step 5 — Produce structured feedback

Organize by severity: Critical (must fix), Important (should fix), Suggestions (optional).

Output Format

## Model Review: [Model Name / Disease]

**Framework:** [Starsim / EMOD / other]
**Research question:** [One sentence]

---

### Critical Issues
- [Issue] — [Why it matters] — [Suggested fix]

### Important Issues
- [Issue] — [Why it matters] — [Suggested fix]

### Suggestions
- [Suggestion]

### Summary
[2–3 sentence overall assessment]

Limitations

• Provides structured feedback, not a formal peer review
• Cannot run or execute model code
• For Starsim-specific issues, cross-reference starsim-dev skills in agentic-modeling-skills