Crisis Resource Management (CRM)

Quick Recap

‣

✈️ Borrowed from aviation for a good reason

‣

📋 The highest-yield behaviors to actually practice

‣

🗣️ Don't soften your language in a crisis

‣

📊 Strong evidence for changing simulated behavior, thinner for real outcomes

‣

🔄 Debriefing is where the actual learning happens

‣

✅ Bottom line

ICU Leadership, Communication & Systems, Protocol 2. Companion to Structured ICU Handoff & Sign-out (I-PASS, Miscellaneous Topics), which addresses routine information transfer; this protocol addresses the acute crisis/resuscitation scenario specifically — the human-factors principles that determine whether a team performs well or poorly once an emergency has already started.

1. Definition

Crisis Resource Management (CRM): a structured set of cognitive and interpersonal ("non-technical") principles governing team performance during acute crises — situational awareness, communication, leadership, and resource allocation — developed specifically because technical/medical knowledge alone does not reliably translate into good real-world crisis outcomes.

Origin: adapted directly from aviation's Crew Resource Management. In the 1970s, the National Transportation Safety Board found human error contributed to over 70% of aviation accidents, and deeper investigation revealed most errors were teamwork failures, not knowledge or technical-skill deficiencies — the same core insight that motivated David Gaba and colleagues at Stanford to found Anesthesia Crisis Resource Management (ACRM) in the late 1980s, the first medical adaptation of this framework, since expanded into emergency medicine, critical care, neonatology, obstetrics, and nursing.

2. Conceptual Framework — Why Non-Technical Skills Matter as Much as Medical Knowledge

The central insight CRM is built on: a team with excellent individual medical knowledge can still perform poorly in a crisis if situational awareness, communication, and leadership break down — non-technical skills are not a "soft" addition to technical competence but an independent, measurable determinant of patient outcome. This directly parallels the aviation finding that most fatal accidents involved technically competent crews whose teamwork failed under acute stress, not their individual flying skill.

3. The 15 Key CRM Principles (Rall & Gaba)

The classic, most widely taught framework, organized as a practical checklist to apply during any crisis:

Checklist:

Know your environment — equipment location, how to operate it, who is available and when, before the crisis starts

Anticipate and plan — crisis management ideally begins before the crisis; think through likely scenarios in advance

Call for help early — do not wait until overwhelmed; calling early is a sign of good judgment, not weakness

Exercise leadership and followership — both roles are active skills; a good follower actively supports the leader's plan while still speaking up when something seems wrong

Distribute the workload — avoid any single team member (especially the leader) becoming a bottleneck performing tasks they should be delegating

Mobilize all available resources — personnel, equipment, cognitive aids

Communicate effectively — closed-loop communication (explicit acknowledgment that an instruction was heard and will be carried out) is the single most trainable, highest-yield communication behavior

Use all available information — actively seek out data rather than working from assumption

Prevent and manage fixation errors — actively guard against tunnel vision on a single diagnosis/plan despite disconfirming evidence

Cross (double) check — verify critical actions and information rather than assuming correctness

Use cognitive aids — checklists, emergency manuals, algorithms; reduces reliance on memory under stress

Re-evaluate repeatedly — crises evolve; a plan that was correct 2 minutes ago may no longer be

Use good teamwork — a general principle underlying most of the others

Allocate attention wisely — avoid over-focusing on one task/monitor to the exclusion of the broader picture

Set priorities dynamically — continuously re-rank what matters most as the situation changes, rather than following a static initial plan

4. Applying CRM in the ICU — Practical Translation

Establish a clear leader immediately — in a code/crisis, ambiguity about who is directing the response is itself a common, preventable failure mode; the leader should generally step back from hands-on tasks to maintain situational awareness of the whole picture
Closed-loop communication as the default: an instruction ("give 1mg epinephrine IV") should be met with an explicit read-back ("giving 1mg epinephrine IV now") and confirmation once completed — this single behavior is highly trainable and directly addresses the "communicate effectively" and "use all available information" principles simultaneously
Avoid "whimperatives": indirect, softened requests ("if possible, would you mind trying an IV, if that's not too much trouble?") reduce clarity and slow response in a crisis — direct, clear requests are not rudeness in this context, they are a patient-safety behavior
Cognitive aids in practice: emergency manuals/cognitive aids (e.g., the Stanford Anesthesia Cognitive Aid Group's Emergency Manual) are specifically designed to be used DURING a crisis, not just referenced afterward — assigning a team member specifically to read the cognitive aid aloud is a recognized, effective role distribution
Fixation error vigilance: explicitly and periodically ask "what else could this be" or "what am I missing" during a prolonged or non-resolving crisis — a deliberate counter to the natural tendency to commit further to an initial hypothesis

5. Evidence Base — Strong on Simulated Performance, Genuinely Thinner on Real-Patient Outcomes

Simulation-based CRM training reliably improves cognitive and interpersonal behaviors in simulated settings — this specific claim has a large, consistent evidence base across anesthesia, emergency medicine, critical care, obstetrics, and neonatology
A growing, but comparatively smaller, body of evidence extends this to real clinical team performance and adverse-event reduction — this protocol treats the distinction between "improves simulated performance" (well-established) and "reduces real-world adverse patient outcomes" (still an actively growing, less mature evidence base) as genuinely important, echoing the identical caution raised in the Breaking Bad News protocol regarding SPIKES training — both are training interventions where learner-level and simulation-level outcomes are more robustly demonstrated than downstream patient-outcome data
Survey-based follow-up evidence: in one study, 83% of CRM course graduates reported having managed a real crisis since training, and 68% reported better practice performance during emergencies as a result — self-reported, not independently verified, but a genuinely large proportion reporting perceived translation from training to real practice
Cross-specialty adoption itself is a form of evidence: CRM-based simulation training has been adopted across nearly every acute-care specialty and profession (nursing, pharmacy, EMS) over three decades — sustained, expanding real-world uptake despite the imperfect hard-outcome evidence base suggests the underlying principles are perceived as genuinely valuable by practitioners across very different clinical contexts, not just a passing trend

6. Structured Debriefing — Where the Learning Actually Consolidates

Debriefing after a real or simulated crisis is not optional or an afterthought — it is where CRM principles are explicitly reinforced and where the team's actual behavior (not just the medical outcome) is examined. Effective debriefing focuses on the CRM principles themselves (which of the 15 were followed well, which broke down, why) rather than solely on the medical/technical outcome, using shared vocabulary so the team can discuss non-technical performance concretely rather than vaguely ("communication broke down" vs. specifically "closed-loop communication was not used for the epinephrine order, and it was given twice").

7. Consultation Matrix

Trigger	Consult	Timing
Establishing/revising institutional CRM training program	Simulation center, nursing education, relevant department leadership	Program-level
Post-crisis debrief needed after a genuinely difficult real event	Unit leadership, cross-reference Clinician Burnout & Wellness protocol if the event was psychologically significant for staff	Immediate to near-term after the event

8. Documentation & Medicolegal Checklist

Team roles (leader, specific task assignments) documented for major resuscitation events where feasible
Debriefing conducted and key learning points documented for significant crisis events, particularly those with an adverse outcome

9. Key Guidelines

No single regulatory body mandates a specific CRM curriculum, but CRM-based simulation training is now a near-universal component of residency/fellowship training across acute-care specialties internationally, and is explicitly recommended by multiple specialty societies as part of patient safety education

10. Landmark Evidence

Source	Contribution
Gaba, Howard, Fish et al., late 1980s–present	Founded Anesthesia Crisis Resource Management (ACRM), the first medical adaptation of aviation CRM; authored the foundational 15-key-point framework, subsequently updated and widely adopted
NTSB aviation error analysis, 1970s	Found >70% of aviation accidents involved human error, predominantly teamwork failure rather than technical/knowledge deficiency — the foundational insight motivating CRM's entire premise
CRM course graduate follow-up survey	83% had managed a real crisis post-training; 68% reported better real-world practice performance
StatPearls/systematic synthesis of CRM simulation literature	Consistent evidence for improved simulated team behavior; "growing" (not yet as extensive) evidence for real-world adverse event reduction

11. Controversies

The same evidentiary gap flagged in the Breaking Bad News protocol applies here: CRM training's effect on simulated performance is well-established; its effect on real, hard patient outcomes (mortality, adverse event rates) is a genuinely less mature evidence base, even after three decades of adoption — this protocol treats broad, sustained cross-specialty adoption as suggestive but not equivalent to definitive outcome-level proof, consistent with this library's general practice of not overstating what training-intervention evidence actually shows.
Distributed/shared leadership models are an active area of continued development, particularly in contexts (e.g., obstetric or surgical emergencies) where the most senior clinician may be simultaneously required for a hands-on procedural role and the crisis-leadership role — the classic "establish a clear leader" principle does not fully resolve this specific, genuinely difficult real-world tension.
Whether CRM principles translate uniformly across genuinely different team structures and hierarchies (varying by country, specialty culture, and institutional norms around speaking up to seniority) remains a reasonable, incompletely resolved question — the framework's aviation origin assumed a particular cockpit hierarchy that does not map perfectly onto every clinical team culture.

12. References

Rall M, Gaba DM. Human performance and patient safety. In: Miller RD, ed. Miller's Anesthesia, 6th ed. Philadelphia: Elsevier Churchill Livingstone; 2005:3021-3072.
Gaba DM, Fish KJ, Howard SK, Burden A. Crisis Management in Anesthesiology. New York: Churchill Livingstone.
Crisis Resource Management Training in Medical Simulation. StatPearls. NCBI Bookshelf, 2023.
Howard S, Gaba D, Fish K, Yang G, Sarnquist F. Anesthesia crisis resource management training: teaching anesthesiologists to handle critical incidents. Aviat Space Environ Med. 1992;63(9):763-770.
Holzman RS, Cooper JB, Gaba DM, et al. Anesthesia crisis resource management: real-life simulation training in operating room crises. J Clin Anesth. 1995;7(8):675-687.
Reznek M, Smith-Coggins R, Howard S, et al. Emergency medicine crisis resource management (EMCRM): pilot study of a simulation-based crisis management course for emergency medicine. Acad Emerg Med. 2003;10(4):386-389.
Crisis Resource Management in the Delivery Room: Development of Behavioral Markers for Team Performance in Emergency Simulation. 2018.
LITFL: Crisis Resource Management (CRM). litfl.com.
Team Management Training Using Crisis Resource Management Results in Perceived Benefits by Healthcare Workers. J Contin Educ Nurs. 2007.

See also: Structured ICU Handoff & Sign-out (I-PASS, Miscellaneous Topics) for routine (non-crisis) team communication; Simulation-Based Training in Critical Care (this section) for the broader evidence base on simulation as an educational modality; Diagnostic Error & Cognitive Bias in Critical Care (this section) for the fixation-error principle explored in greater depth.