Firefighters, tacit knowledge and rapid decisions

For the past week, Cape Town firefighters have relentlessly been battling fires across the city and it appears that every hour on the news there is another fire to manage.  It’s been said that every fire is different, so have you ever wondered how firefighters decide on the best course of action with so many variables and very little time to exhaust all the possible options and outcomes?

In the late 70’s Gary Klein explored the above question. One of his first discoveries was that the laboratory models of decision-making were not sufficient for the conditions that firefighters work in.  Furthermore during interviews many firefighters were unable to explain how they reached their conclusions and some experienced firefighters stated that they’ve never made any decisions. This led to some fieldwork observing the firefighters and conducting post-incident interviews to explore whether and how firefighters make decisions.  The findings and further work on decision-making resulted in the Recognition Primed Decision-making model (RPD), naturalistic decision-making and recently the shadowbox concept.

Initially Klein hypothesized that firefighters narrow their options to two and then choose the best option of these.  Findings demonstrated that 80-90% of the firefighters used the same RPD strategy to decide:  They would only generate one option based on cues in the environment. Once they’ve selected their option, they would mentally imagine the next step e.g. how the fire are likely to spread, which houses are likely to collapse and then act on that.

Klein discovered that in high-stress, uncertain situations the level of expertise of the decision-maker plays a major role. Decision-making are based on tacit knowledge where the ‘new’ situation contains something familiar or a cue that allows the decision-maker to act on a ‘prototype’ of what worked before.  Experienced firefighters can thus rapidly match current situation to past pattern and literally instantaneously and seemingly intuitively know what to do.

As an example, Klein refers to an interview where firefighters had to enter a house that was on fire.  Within seconds the commander ordered everyone to leave the house, literally as the last firefighter escaped, the floor collapsed.  In retrospect, the commander explained that the fire was too quiet and that was unusual, he also felt that his ears were too hot.  The combination of these impressions prompted the reaction.  At the time he didn’t know what was wrong, but he knew something was amiss.  It turned out that the heart of the fire was not on the first floor where they entered, it was in the basement that they didn’t know about.  If the commander didn’t pick up on the subtle cues, the entire team could’ve been killed.

Because fires grow exponentially, the faster the firefighter can react, the more likely it is that the fire can be contained. It was found that in these situations, the decision-maker opts for the first workable idea and not necessarily the best option (satisficing).  It should be mentioned that experienced decision-makers are very likely to generate a plausible first option. Laboratory decision models that require comprehensive evaluation and multiple options are simply not feasible as the fire would be out of control by the time a decision has been taken.  It’s thus unlikely that reflective deliberation is the key to successful decisions whilst firefighting.

All of the above was described in the Recognition Primed Decision model (RPD).  In the RPD model, a suitable reaction is immediately considered and recognized.  RDP relies on the intuitive pattern finding, where people use their experience to match the situation to patterns that they have learned.  This is followed by conscious analysis in the form of visualizing or mental simulation to ‘test’ the option. Both intuitive pattern finding and conscious analysis are required. According to the RPD model it becomes a cycle where the decision-maker creates a mental picture of what a reasonable solution looks like, the actions required to reach the ideal outcome and then mentally evaluates the effects of the action.

 

figure-65-recognition-primed-decision-making-model-from-a-recognition-primed-decision

RPD model

 

These findings have been replicated in other environments including military command and control, management of offshore oil installations and neonatal critical care nurses.  These environments are all dynamic, constantly shifting, outcomes are uncertain and decisions are time-critical.

I think that the RDP model enforces the importance of tacit knowledge in high stress environments.  This is knowledge that is deeply rooted, often people are not aware of the extent of their knowledge e.g. firefighters saying that they don’t make decisions. Tacit knowledge is the know-how that is not contained in the procedure manuals or policies.  It is what allows people to perform tasks and provide local knowledge without focused attention.  There is only one way to obtain tacit knowledge and that is through experience, and preferably this experience should be guided by an experienced mentor.  Sadly tacit knowledge is lost when experienced people leave an organization or industry.

Emergency care is an industry in which tacit knowledge is really important. Even just driving an ambulance into some neighborhoods requires tacit and local knowledge.  Rapid and effective decisions are required prehospital, during major incidents and in overcrowded emergency centers. In these situations the decision-making density is high; decisions are made amidst constant, dynamic change.  The information available to the decision-maker is incomplete, ambiguous or not available. Decisions are often required within seconds rather than minutes.

The RPD has two preconditions namely

1)      There needs to be adequate cues in the environment to assist the decision maker

2)       The opportunity in an environment to learn from the cues as it takes time to reach the point of expertise to make these decisions

These preconditions can be used as pointers on how to improve decision-making in emergency care situations.  The first is building a repertoire of patterns that will assist decision-making.  A great method that is used clinically is simulation training; this could and should be extended to operational decisions during crises situations. In the prehospital environment this is already well-established during training.

The other precondition is the one that I think emergency care and greater healthcare industry often disregards: tacit knowledge.  It takes time to build adequate experience which allows good decisions.  Consider the vast amount of nurses, paramedics and firefighters that has left the industry in the past ten years, I daresay that the industry disregards the expertise that comes with 20 – 30 years of hands-on experience in the field.  These experts are not only leaving because of better salaries, but it’s for better work conditions, opportunities, career progression and to feel acknowledged. Tacit knowledge should be aggressively guarded by the industry!

Novices see only what is there; experts can see what is not there – Klein

Read more by searching for any work by Gary Klein including his work on macro cognition; sources of power, gaining insights, naturalistic decision-making and sense-making

https://www.researchgate.net/profile/Gary_Klein6/publication/235418838_A_Recognition_Primed_Decision_RPD_Model_of_Rapid_Decision_Making/links/565512fd08aefe619b1a44f0.pdf

 

Advertisements

Making sense…learning about sense-making

Remember the movie ‘Back to the future’ where Marty (Michael J. Fox) travels through time, whilst figuring out what is happening and how to influence the future whilst simultaneously dealing with the current situation?

It reminds me of sense-making. Sense-making is about how we give meaning to a situation. Sense-making is based in our decisions and beliefs and it is influenced by aspects such as power relations, hierarchy, autonomy, linguistics, temporalization (focus on now, the past or the future), justification, culture and traditional society, equality, bargaining power, locus of control, openness to change. Sense-making is done individually or collectively in organizations and communities.  Capturing these underlying nuances in our perceptions allows deeper insights into why an organization or community are angry, failing, succeeding, innovating, etc.

Brenda Dervin, Gary Klein, Karl Weick and David Snowden are a few of the contributors to sense-making. They all describe their own theory about sense-making either as individual or collective activity.  Snowden combined several sense-making theories into one tool.  For my studies I am using his tool to capture community perspectives regarding cost as a barrier to access emergency care. 

Today I’m sharing my understanding of Brenda Dervin’s metaphor for sense-making.  Dervin sees sense-making as the individual activity of information seeking, processing, recreating and application. Information is described as a tool designed by humans whilst making sense of reality.  Methods of communication and the application of information (knowledge) are key aspects of sense-making.

The metaphor 

timespace

It starts in a time-space milieu, implying constant energy and movement. This environment is never static and time-space energy propels us forward or pulls us back, always fluctuating and never stable.

Moving through time-space we have Mr. Squiggly.  Humans are depicted as squiggly because we are caught between certainty and uncertainty.  Thus there is a constant flux within the person (internal) and between the person and the environment (external).  Mr. Squiggly carries an umbrella, symbolizing mind-set, perceived constraints and enablers. (I feel that it should be a backpack – demonstrating our baggage).

Mr Squiggly.png

Whilst moving through time-space, a gap or barrier is encountered. For the purposes of my work, the gap is a life-threatening injury or illness requiring urgent care. The gap can however be any barrier or difficulty in daily life (in an organization it could be new policies, a new boss etc.). The gap forces Mr. Squiggly to stop until a way has been found to ‘bridge’ the gap and reach an outcome.

The outcome depends on how the gap is bridged.  Some potential outcomes are not obvious in the beginning and it may only become apparent retrospectively, influencing future decisions and beliefs.

Moving from situation to outcome requires a bridge. The building blocks used to build the bridge consists of different types of blocks.  One being the individual mind-set, during a life-threatening emergency it would include individual beliefs about health, healthcare, medication, culture. Building blocks also include inputs from others, the stories within the community about a time that something similar happened to someone else, financial hardships suffered, patient outcome etc.  Mr. Squiggly consciously and unconsciously use all the information in the form of building blocks to create the bridge.

Once the bridge is built, Mr. Squiggly can leap across the bridge to the outcome; this is aptly called gap-bridging.  The building blocks of information is now applied, thus knowledge are created. Dervin use the term ‘verbing’ as an important gap-bridging and sense-making tool.  Everyday examples of verbing include the use of words for example emailing, googling or ubering. Verbing thus occurs when a noun is turned into a verb, creating action or experience.  It certainly fits a methodology where knowledge is seen as a context-specific sense-making activity in a specific point in time and space.

 teh gap.png

I’m using sense-making to capture the various roles within a community perceive gaps, bridge gaps and view the outcomes during a life-threatening injury or illness. Perceptions will be captured using a type of narrative enquiry, where the participant is asked to tell a descriptive story.  After telling the story, the storyteller explains the meaning of their own story by indexing it onto a predesigned framework.  This is very different from ‘traditional’ research where the investigator assigns meaning to the stories. Neutral questioning is used to guide the indexing.  Questions are framed in such a way to capture the nuanced aspects of hierarchy, autonomy, equality etc.  After capturing, the data are combined and the software allows for it to be visually displayed, enabling easy identification of patterns and trends.  More details on the software are a story for another day.

This technique has been applied to monitor, evaluate, communicate, and create feedback loops in projects, organizations, communities and even broader society.

I am using sense-making because I’m passionate about the voice of the ‘voiceless’ in organizations and communities.  I feel that tacit knowledge is often overlooked.  Sense-making provides a tool to capture many voices, combine the different perspectives and seek common ground, emerging trends or underlying moods. This is powerful, whether in an organization, community or development project. It prevents the implementation of one perspective, ‘outside’ views or only top-down approaches. Inclusive sustainable implementation requires more than one story, one perspective and more than one type of knowledge.

 As Chimamanda Ngozi Adichie says in her powerful TED talk about the danger of a single story:

The single story creates stereotypes, and the problem with stereotypes is not that they are untrue, but that they are incomplete. They make one story become the only story.”

References

I adapted the pictures from: Facing a gappy situation. Sense-making methodology: communicating communicatively with campaign audience.  Dervin, B. 2003.  In Dervin B, Foreman-Wernet L (Eds). Sense-making methodology reader: selected writings of Brenda Dervin, Cresskill, NJ: Hampton Press 2003.

http://www.ted.com/talks/chimamanda_adichie_the_danger_of_a_single_story?language=en

It’s more complex than it seems….19 babies received the wrong vaccinations

A week ago there was an incident at one of the private hospitals.  Nineteen new-born babies received the incorrect BCG vaccination and are now at risk for contracting Tuberculosis and suffering other side effects.   According to the media reports, the vaccine was confused with another drug used for a different patient population (adult and new-born) with different indications. But vials are similar in appearance.

No doubt the parents are furious and the media has a duty to report and allow for public comment.  I keep telling myself not to read the comments alas I still read it.  The readers has crucified the nursing staff involved and are blaming the lazy, stupid, can’t read, how-hard-can-it-be-to-do-your-job nurses.   Stupid nurse, fire the nurse.  End of story.  Problem solved.

Errors and failures in complex systems such as hospitals are not that simple.  It is not linear with a direct and clear relationship between cause and effect.  And neither are the interactions between the various functions within the system linear either.   In complex systems there is no single most “fundamental” reason for an error.  There is an interaction of several causal factors originating from various places within the system.  The inquisitive investigator will find patterns between apparently unrelated incidents prior to this incident.

The impact of a blame culture or over simplified cause and effect investigation results in system failures being “blamed” on the last “link” prior to the incident.  This last link is typically the doctor or nurse involved in patient care.

Possible contributing factors to the vaccination error:

  • The drug manufacturers.  There has been an increasing push on manufacturers to distinguish medication and provide visual cues on ampoules as warning prompts.  For instance the difference between medication manufactured for adult, paediatric and neonatal should be visible by either colour or tag.
  • The pharmacist that dispensed the drug had a duty to check and label it.  The question should be asked that if an adult drug ended up in a new-born unit, was there a dispensing error?
  • Responsible person receiving and unpacking the medication in the unit.  This person has to check medication received against requested medication.  The role is performed by an administrator with no nursing background.  If the packaging was familiar and the name similar, they would miss the error or assumed it is a generic.
  • The first nurse to administer vaccination using the vial.  There is policies dictating that nurses perform the 5R’of medication administration to ensure that it’s the right drug, right dosage, right route for the right patient, at the right time.  Was this done, and could the nurse complete the check without interruption?
  • Normalizing deviance.  If the same constraints are experienced on a regular basis it is accepted as normal.  When it’s accepted as a “new” norm it creates an environment where failure or error is inevitable.  Nursing shortages has become normal; this has an impact on safe medication administration practises as well as the frequency of interruptions during this risky time.  This normalizing deviance results in medication error becoming inevitable; the question is simply when it is going to occur, not if.
  • Latent factors includes management decisions, staffing levels, general policy and procedure, design of the unit, the pharmacy and the hospital, staff related factors such as perceived time pressure to perform tasks, fatigue, amount of recent shifts worked, was this an overtime shift, how many hours since the last uninterrupted break.  Level of experience of the nurse, pharmacist, ward administrator.  These latent factors can be dormant in a complex system prior to it causing failure.

Errors involving drugs must be the most common hospital safety event and it has been researched widely.  The high error rate may be because there is an over reliance on human factors with a disregard of heuristics during medication preparation and administration.  We need to avoid relying on the flawed human aspects of decision making.  When making rapid decisions we apply cognitive shortcuts. An example would be reading the first few letters on a familiar looking vial, kept in the normal storage space in a container marked with the vaccine name.  So we make an assumption that this must be the right drug and we continue to the next decision. The innovative safety-vests stating do-not-disturb can reduce the pressure to make rapid decisions before the next interruption.

I certainly felt that I could relate to this incident.  It could’ve been me making the error.  I wish that I could say that I always do rigorous medication checks; or that I trust my own decision-making towards the end of a twelve hour shift with limited breaks.  I wish that I could say that at the end of back to back shifts I’m still as conscientious as during the first shift.  I wish that I could say that frequent interruptions do not affect the task that I’m busy with.

 In summary

According to statistics by the Institute for Healthcare Improvement approximately 5% of medical error and harm is caused by incompetence or negligence.  The other 95% involves conscientious, competent and possibly caring individuals.  Mistakes such as this error were probably caused by a bad system and not a bad healthcare professional.

There is a saying that: “a bad system will beat a good person every time.”