Binary gate checklist. All items must pass before proceeding to next phase.

Population, Intervention, Comparator, Outcomes, Study design - framework for research questions.

The audit trail showing exactly where a number came from (e.g., "Table 2, p.1234").

Daily adversarial checking by 2 rotating members to catch errors in extraction.

Emergency protocol: 72 hours to fix a failed DoD gate or terminate with documented reasons.

A measurable marker (e.g., cholesterol) used as proxy for patient-important outcomes (e.g., death).

Scope lock - no new studies can be added after this point.

Heterogeneity statistic. Low <30%, Moderate 30-60%, High >60%.

Complete reproducible package: code, data, instructions. Anyone can run and get identical results.

Structured document with options, evidence, and trade-offs for team decision-making.

International registry for systematic review protocols. Timestamps your pre-specified methods.

Reporting guideline for systematic reviews. Includes the flow diagram showing study selection.