Crisis resource management

Crisis resource management is the set of "non-technical skills required for effective teamwork in a crisis situation", according to the LITFL entry on this topic.   In brief, the key principles of CRM are clearly defined leadership, situational awareness, anticipation of the next step, clear closed loop communication, distribution of workload and allocation of attenion to tasks of importance. The entire (massive) psychological and theoretical framework for understanding how to improve performance in crisis situations can be summarised as "simulate the shit out of it". 

Instead of ranting aimlessly about this topic, the best approach would probably to frame the issue in terms of a critically evaluate question, which is a realistic scenario for the written exam. The theory is discussed, rationale for CRM is offered with various arguments for and against, practical approaches are discussed and evidence for the use of CRM simulation is offered.

As far as published evidence goes, there is a lot out there- enough to get lost in. The time-poor candidate should just pick a single representative paper. For example, any of the following may serve as "the one thing you read on this topic".

• Murray et al (2000)
• Gaba D, 2010 (from the Father of Modern CRM)
• Haerkens et al (2012) (does a good job of summarising the CRM concepts with special reference to the ICU environment)

Theory of crisis resource management

As there have so far been no CICM fellowship questions on this topic, I will only be brief. Major principles of CRM according to a seminal paper by Murray et al (2000) consist of the following:

• establishing leadership and support for the leader
• recognizing specific functions of a leader
• the importance of communication
• the need for continuous reassessment
• the use of all available resources
• avoidance of fixation of ideas and goals, and
• consideration of personality traits for optimal group performance.

CRM, therefore, does the following good:

• Provides a leadership structure
• Provides a shared mental model of the situation
• Improves the efficiency of coomunication

Key terms to spout during the exam seem to be "leadership structure" and "shared mental model". These terms will probably lead to ticked boxes. More tickbox statements can be found in the well-regarded chapter from Miller's Anaesthesia by Rall and Gaba (2005), quoted in LITFL:

1. Know the environment
2. Anticipate and plan
3. Call for help early
4. Exercise leadership and followership
5. Distribute the workload
6. Mobilise all available resources
7. Communicate effectively
8. Use all available information
9. Prevent and manage fixation errors
10. Cross (double) check
11. Use cognitive aids
12. Re-evaluate repeatedly
13. Use good teamwork
14. Allocate attention wisely
15. Set priorities dynamically

Elements of good CRM and roles of crisis team members

The same paper (Murray et al, 2000 ) went on to make some definitions and to assign some roles to members of a crisis management team, which seem like something the college might be interested in at some stage. They will probably ask you to describe your approach to medical leadership in a rapid response team. What are the key principles of crisis resource management?  or some wank like that. For the purpose of future cut-and-paste exercises, this wank is regurgitated below. One day I will be able to insert it into the discussion section of an SAQ with minimal modification. For those paying attention, these roles and definitions will look identical to the content of Table 1 from Murray et al (2000). 

Roles of the crisis leader

• Steps back and manages an event
• Sets clear goals
• Organizes the team
• Delegates responsibility
• Distributes work appropriately

Roles of the team member

• Assumes assigned responsibility
• Feeds back event management data
• Provides task and cognitive support
• “Owns” delegated problems
• Roles can be exchanged

Elements of communication in a crisis 

• Address people directly—introduce yourself
• Declare an emergency—urgency, not panic
• Establish your communication paths
• Use nonjudgmental comment
• Close the loop—give feedback

What is expected from the team leader

• Step back—physically and mentally
• Step back to see the whole picture
• Verbal review of patient and situation
  • Avoids fixation errors
  • Provides clarity of ideas
  • Generates new ideas

Support for the leader and team members

• Asking for help when needed is a sign of maturity, not of weakness
• Incremental help may be called
• What sources of help available?
• When and whom to call for help
• Type of help—advice, hands-on, specialized

Mobilisation of resources

• Prepare for anticipated needs—special carts, memo sheets
• Understand the infrastructure
• Know how support systems work
• Internal and external (think “outside the box”)

Rationale for the use of crisis resource management in medicine

The main reasons for the use of CRM in medical emergencies are:

• In times of stress humans perform poorly
• Analysis of poor performances reveals that lack of leadership, poor team coordination, cognitive fixation and poor communication are to blame in many cases.
• Barriers to good communication and coordination have been identified, and are largely due to structural and organisational issues than anything specifically related to medicine, individual knowledge or judgement.
• Prevention of poor performance therefore requires these barriers to be surmounted by good communication technique, clear leadership, situational awareness and a shared mental model.

Argument from analogy

The use of CRM in medicine originated with David Gaba, who in the 1980s adapted some practices from the aviation industry to use in the setting of anaesthetic crises. ​His 2010 editorial explores this history.Originally, the "C' in CRM stood for "Crew" or "Cockpit". The original article by Lauber et al (187) is not available, but - in brief- it was an exploration of the various ways in which air crews had failed each other in a series of aircraft disasters from the 1970s. The analogies between pilots and anaesthetists are extensive:

• Both have an emphasis on safety
• Both have hours of boredom punctuated by moments of terror
• Both need to work in a complex team environment

Thus, the same approach can be extrapolated to the medical environment, with some important modifications. For instance, the piloting team of an aircraft will be usually very familiar with the equipment and flight path, allowing a lot more "automated" behaviour. 

Objectives of crisis resource management

The whole point of this mode of thinking is to ensure better patient outcomes, i.e. the prevention of situations in which a patient is killed by the actions of an incompetent but well-meaning committee. Such a group is typically composed of individials who independently work better than they do as a team. In other words, a team should be at least equal  to the sum of its parts. 

Specific errors which are addressed by good CRM include:

• Ineffective communication
• Duplication of effort
• Fixation errors (i.e. cognitive bias)
• Inefficient use of resources
• Poor access to vital information

Objections to the application of CRM in medicine

Though the high pressure environment of the aircrafty cockpit is sexy and hot with excitingly militaristic dialogue, the mindless application of aviation safety principles into medicine can be criticised as inappropriate. In his blog entry from 2012, Andy Buck makes the argument that a medical crisis is not not the same environment as an the aircraft cockpit. A cockpit is a structured environment where everybody has discrete and unalterable roles, whereas a resuscitation is a dynamic environment with multiskilled operators who can adapt to different roles. Further, Buck objects to the negative connotation of the term "crisis". 

Evidence for the utility of CRM and CRM training

Does this concept actually help anybody? That is difficult to say. It is not exactly susceptible to investigation by randomised controlled trial. Evidence does exist, of a sort - but it is all indirect. For instance, Clark et al (2014) had used a validated CRM assessment tool (the Ottawa GRS) to assess emergency medicine residents. CRM training was associated with a gradual increase in the Ottawa GRS scores, particularly the domains of leadership, problem solving, and resource utilization. How does this translate in patient survival? We may never know this. It would probably be unethical to randomise a group of patients to receive critical care from a group of practitioners who have never had CRM training when a CRM-trained group is also available. 

Pragmatic rules for good crisis resource management

Avoidance of mitigated speech

Wishy-washy gibberish like "can somebody please get some adrenaline" and "can we have a think about the futility of resuscitation here" are examples of mitigated speech, a behaviour of people who perceive themselves to be in a position of diminished power or authority (or perhaps in possession of undeserved authority).  It is an attempt to downplay or sugarcoat the meaning of a statement. Examples of such speech in resuscitation scenarios can include the following:

• Command: "Fluids will be adminstered"
• Obligation: "We need to administer some fluids"
• Suggestion: "Why don't we give some fluid?"
• Question: "Does anybody agree that fluids would be appropriate?"
• Preference: "Perhaps we could consider some fluids as one of the options"
• Hint: "I wonder if this patient might be somewhat volume-depleted"

As opposed to:

• Direct order: "Give one litre of Plasmalyte as a stat bolus, please."

An excellent article by Brindley et al (2011) has a table (Table 1) of practical communication strategies, among which one can find the following pearls:

• Announcements (of actions, to ensure that a double check of the action occurs, eg. "I am given adrenaline as a bolus". This is a extrapolation of pilot training, who are taught to "fly by voice", i.e. to routinely announce their every action)
• Graded assertiveness (commanding language invites for people to either obey or refuse, rather than stand around doing nothing in response to a mitigating statement)
• Confirmation (repeating the order back to the leader, to ensure it has been understood correctly)
• Closed loop communication (which helps ensure that ordered tasks are actually carried out)

Standardised communication technique

"Avoid the fluff" is the best way to summarise this piece of advice. The best resource for this seems to be a blog entry from resus.me, "Learning to Speak Resuscitese". The author (Cliff Reid) points to the standardised set of phrases used by pilots and air traffic controllers which is designed to minimise ambiguity. "Abort", "acknowledge",  "confirm" et cetera - these words make the team members sound inhuman and robotic, which is good because inhuman robots are fast, efficient and error-resistant. This is what you want in a resuscitation team.

It all sounds good, but thus far there is no strong evidence that it works. Yamada et al (2014) could not demonstrate any positive influence on error rate from standardised communication, at least in the setting of simulated neonatal resuscitation. The only benefit seemed to be the earlier initiation of chest compressions.

Debriefing

Debriefing can be defined as a "post-experience analytic process, ...evaluating and integrating lessons learned into one’s cognition and consciousness". This is the definition offered by Roxane Gardner (2013), who otherwise promotes debriefing as "a lynchpin in the process of learning". The concept is historically founded in WWI, when Samuel Lynn Atwood Marshall became frustrated with the inaccurate reconstruction of wartime events and began to hold post-combat interviews immediately after missions. 

According to Rudolph et al (2008), the debriefer serves as a ‘‘cognitive detective’’ who "uses observations of a participant’s or team’s performance and outcomes, and works backwards to identify what frames drove their action". This technique is known as "advocacy-inquiry". In case this ever comes up in the exam, the three stages of debriefing described by this author are as follows:

1. Reactions
   • Clear the air
   • Review the facts
   • Set the stage for addressing learning objectives
2. Understanding
   • Explore what happened
   • Unpack frames through advocacy–inquiry
   • Apply good judgment and teach, moving participants to new understanding or skills
   • Generalize lessons learned to other situations
3. Summarize
   • Review lessons learned
   • Discuss take-aways, lessons learned that will be applied in future events

This should take place in an open safe environment, where the members can feel comfortable discussing matters which might be personally embarrassing or ego-damaging. Debriefing must be confidential and all input needs to be acknowledged as valuable. Thank of it as a big hippy group-hug session. 

Simulation and teaching of CRM principles

Østergaard et al (2011) presents an excellent overview of simulation training for crisis resource management. Key features include:

• A scenario with well-defined learning objectives
• A "patient", usually a mannequin
• An instructor, ideally with experience of the "real" scenario- to act as the facilitator
• A setting, designed to simulate a relevant environment 
• A suspension of disbelief, depending on the quality of the setting
• Audiovisual recording of the scenario for the purposes of debriefing

"Suspension of disbelief" and realism in medical simulation

The simulation mannequin may look humanoid, but it has numerous unrealistic elements, no matter how sophisticated it may be. The skin is rubbery, the breath sounds are weird, the pulse is mechanical, the chest clicks when compressed and the airway has sticky residue from previous "intubations" using that horrible silicone spray.  And on top of that you are working with water-filled syringes labelled as "adrenaline", a Nerf defibrillator which delivers no shocks, and monitoring equipment which does not need to be attached to the patient. How are you supposed to get into that?  

Well. Simulation does not have to be physically realistic to be effective. It merely needs to have semantic realism, i.e. conveyed by realistic-sounding concepts and reinforced by feedback. For instance, the mannequin does not need to spurt blood to simulate haemorrhage: it would be enough for the heart rate oin the monitor to increase, and for the instructor to say that there is a pool of melaena under the sheets. In fact Dieckmann et al (2007) recommend that the pursuit of physical realism should be far down the list of priorities for the medical simulation operator.