| name | review-checklist |
| description | Use when verifying that a process follows all required steps before completion or handoff. |
| source | FAA Aviation Safety Checklists, WHO Surgical Safety Checklist (Haynes et al., NEJM 2009) |
| tags | ["quality-assurance","checklist","process","error-reduction"] |
Review Checklist
Systematically verify every required step before signing off on a process.
Why This Is Best Practice
Adopted by: Aviation (FAA mandatory pre-flight checklists), surgery (WHO Surgical
Safety Checklist adopted by 4,000+ hospitals), NASA Mission Control, nuclear power
(NRC operating procedures), and all industries where step omission causes irreversible
failure.
Impact: WHO Surgical Safety Checklist reduced surgical complications by 36% and deaths
by 47% across 8 hospitals in 8 countries (Haynes et al., NEJM 2009, n=3,733 patients).
Aviation accident rates dropped from 40+ fatal accidents per million flights (1950s) to
under 0.2 by 2000 — checklists identified as primary contributor (NTSB longitudinal data).
Why best: Human memory under time pressure is unreliable. External checklists eliminate
recall failures — the root cause of ~70% of aviation accidents before adoption. No
alternative (experience, expertise, vigilance) achieves the same error-reduction rate.
Sources: WHO Surgical Safety Checklist 2009, FAA Advisory Circular AC 120-21, NTSB
Steps
- List every required step in sequence before beginning the process.
- Check each step immediately after performing it — not from memory after the fact.
- If a step fails or cannot be confirmed, halt and resolve before proceeding.
- Have a second person verify critical steps independently (two-person rule).
- Record the completed checklist: who verified, when, and any deviations noted.
Rules
- Never check off a step not yet performed — even under time pressure.
- Keep checklists short enough to complete in under 5 minutes; split longer processes.
- Review the checklist itself annually to ensure steps reflect current practice.
- Do not let familiarity breed complacency — treat each checklist as a live document.
Common Mistakes
- Checking off steps from memory after completing a block — defeats the purpose.
- Using the checklist only for documentation, not as a real-time verification tool.
- Adding too many items until it becomes noise rather than a gate.
- Skipping the checklist when "everything is obviously fine" — the process assumes fallibility.
When NOT to Use
- When the process has a single step with no sequencing dependencies and no meaningful failure modes.
- When operating in a free-form creative exploration phase where rigid sequencing inhibits discovery.
- When the checklist itself has not been validated against the actual process it covers.
