Making Better Decisions: Lessons from Noise by Daniel Kahneman, Oliver Sibony, and Cass R. Sunstein

The authors of Noise: A Flaw in Human Judgment* examine the random variations that occur in judgmental decisions and recommend ways to make more consistent judgments.  Variability is observed when two or more qualified decision makers review the same data or face the same situation and come to different judgments or conclusions.  (Variability can also occur when the same decision maker revisits a previous decision situation and arrives at a different judgment.)  The decision makers may be doctors making diagnoses, engineers designing structures, judges sentencing convicted criminals, or any other situation involving professional judgment.**  Judgments can vary because of two factors: bias and noise.

Bias is systematic, a consistent source of error in judgments.  It creates an observable average difference between actual judgments and theoretical judgments that would reflect a system’s actual or espoused goals and values.  Bias may be exhibited by an individual or a group, e.g., when the criminal justice system treats members of a certain race or class differently from others.

Noise is random scatter, a separate, independent cause of variability in decisions involving judgment.  It is similar to the residual error in a statistical equation, i.e., noise may have a zero average (because higher judgments are balanced by lower ones) but noise can create large variability in individual judgments.  Such inconsistency damages the credibility of the system.  Noise has three components: level, pattern, and occasion. 

Level refers to the difference in the average judgment made by different individuals, e.g., a magistrate may be tough or lenient. 

Pattern refers to the idiosyncrasies of individual judges, e.g., one magistrate may be severe with drunk drivers but easy on minor traffic offenses.  These idiosyncrasies include the internal values, principles, memories, and rules a judge brings to every case, consciously or not. 

Occasion refers to a random instability, e.g., where a fingerprint examiner looking at the same prints finds a match one day and no match on another day.  Occasion noise can be influenced by many factors including a judge’s mood, fatigue, and recent experience with other cases. 

Based on a review of the available literature and their own research, the authors suggest that noise can be a larger contributor to judgment variability than bias, with stable pattern noise larger than level noise or occasion noise.

Ways to reduce noise

Noise can be reduced through interventions at the individual or group level. 

For the individual, interventions include training to help people who make judgments realize how different psychological biases can influence decision making.  The long list of psychological biases in Noise builds on Kahneman’s work in Thinking, Fast and Slow which we reviewed on Dec. 18, 2013.  Such biases include overconfidence; denial of ignorance, which means not acknowledging that important relevant data isn’t known; base rate neglect, where outcomes in other similar cases are ignored; availability, which means the first solutions that come to mind are favored, with no further analysis; and anchoring of subsequent values to an initial offer.  Noise reduction techniques include active open-mindedness, which is the search for information that contradicts one’s initial hypothesis, or positing alternative interpretations of the available evidence; and the use of rankings and anchored scales rather than individual ratings based on vague, open-ended criteria.  Shared professional norms can also contribute to more consistent judgments.

At the group level, noise can be reduced through techniques the authors call decision hygiene.  The underlying belief is that obtaining multiple, independent judgments can increase accuracy, i.e., lead to an answer that is closer to the true or best answer.  For example, a complicated decision can be broken down into multiple dimensions, and each dimension assessed individually and independently.  Group members share their judgments for each dimension, then discus them, and only then combine their findings (and their intuition) into a final decision.  Trained decision observers can be used to watch for signs that familiar biases are affecting someone’s decisions or group dynamics involving position, power, politics, ambition and the like are contaminating the decision process and negating actual independence.

Noise can also be reduced or eliminated by the use of rules, guidelines, or standards. 

Rules are inflexible, thus noiseless.  However, rules (or algorithms) may also have biases coded into them or only apply to their original data set.  They may also drive discretion underground, e.g., where decision makers game the process to obtain the results they prefer.

Guidelines, such as sentencing guidelines for convicted criminals or templates for diagnosing common health problems, are less rigid but still reduce noise.  Guidelines decompose complex decisions into easier sub-judgments on predefined dimensions.  However, judges and doctors push back against mandatory guidelines that reduce their ability to deal with the unique factors of individual cases before them.

Standards are the least rigid noise reduction technique; they delegate power to professionals and are inherently qualitative.  Standards generally require that professionals make decisions that are “reasonable” or “prudent” or “feasible.”  They are related to the shared professional norms previously mentioned.  Judgments based on standards can invite controversy, disagreement, confrontation, and lawsuits.

The authors recognize that in some areas, it is infeasible, too costly, or even undesirable to eliminate noise.  One particular fear is a noise-free system might freeze existing values.  Rules and guidelines need to be flexible to adapt to changing social values or new data.

Our Perspective

We have long promoted the view that decision making (the process) and decisions (the artifacts) are crucial components of a socio-technical system, and have a significant two-way influence relationship with the organization’s culture.  Decision making should be guided by an organization’s policies and priorities, and the process should be robust, i.e., different decision makers should arrive at acceptably similar decisions. 

Many organizations examine (and excoriate) bad decisions and the “bad apples” who made them.  Organizations also need to look at “good” decisions to appreciate how much their professionals disagree when making generally acceptable judgments.  Does the process for making judgments develop the answer best supported by the facts, and then adjust it for preferences (e.g., cost) and values (e.g., safety), or do the fingers of the judges go on the scale at earlier steps?

You may be surprised at the amount of noise in your organization’s professional judgments.  On the other hand, is your organization’s decision making too rigid in some areas?  Decisions made using rules can be quicker and cheaper than prolonged analysis, but may lead to costly errors. which approach has a higher cost for errors?  Operators (or nurses or whoever) may follow the rules punctiliously but sometimes the train may go off the tracks. 

Bottom line: This is an important book that provides a powerful mental model for considering the many factors that influence individual professional judgments.

*  D. Kahneman, O. Sibony, and C.R. Sunstein, Noise: A Flaw in Human Judgment (New York: Little, Brown Spark) 2021.

**  “Professional judgment” implies some uncertainty about the answer, and judges may disagree, but there is a limit on how much disagreement is tolerable.

Healthcare Safety Culture and Interventions to Reduce Preventable Medical Errors

HSS OIG report cover

We have previously written about the shocking number of preventable errors in healthcare settings that result in injury or death to patients.  We have also discussed the importance of a strong safety culture (SC) in reducing healthcare error rates.  However, after 20 years of efforts, the needle has not significantly moved on overall injuries and deaths.  This post reviews healthcare’s concept of SC and research that ties SC to patient outcomes.  We offer our view on why interventions have not been more effective.

Healthcare’s Model of Safety Culture

Healthcare has a model for SC, shown in the SC primer on the Agency for Healthcare Research and Quality’s (AHRQ) Patient Safety Network website.*  The model contains these key cultural features:

• acknowledgment of the high-risk nature of an organization's activities and the determination to achieve consistently safe operations
• a blame-free environment** where individuals are able to report errors or near misses without fear of reprimand or punishment
• encouragement of collaboration across ranks and disciplines to seek solutions to patient safety problems
• organizational commitment of resources to address safety concerns.

We will critique this model later.

Healthcare Providers Believe Safety Culture is Important

A U.S. Department of Health and Human Services (HSS) report*** affirms healthcare providers’ belief that SC is important and can contribute to fewer errors and improved patient outcomes.

AHRQ administers the Patient Safety Organization (PSO) program which gathers data on patient safety events from healthcare providers.  In 2019, the HSS Office of Inspector General surveyed hospitals and PSOs to identify the PSO program’s value and challenges.  SC was one topic covered in the survey and the results confirm SC’s importance to providers.  “Among hospitals that work with PSOs, 80 percent find that feedback and analysis on patient safety events have helped prevent future events, and 72 percent find that such feedback has helped them understand the causes of events.” (p. 10)  Furthermore, “Nearly all (95 percent) hospitals that work with a PSO found that their PSOs have helped improve the culture of safety at their facilities.  A culture of safety is one that enables individuals to report errors without fear of reprimand and to collaborate on solutions.” (p. 11) 

Healthcare Research Connects SC to Interventions to Reduced Errors

AHRQ publishes the “Making Healthcare Safer” series of reports, which represent summaries of important research on selected patient safety practices (PSPs).  The most recent (2020) edition**** recognizes SC as a cross-cutting practice, i.e., SC impacts the effectiveness of many specific PSPs. 

The section on cross-cutting practices begins by noting that healthcare is trying to learn from the experience of high reliability organizations (HROs).  HROs have many safety-enhancing attributes included committed leaders, a SC where staff identify and correct all deviations that could lead to unsafe conditions, an environment where adverse events or near misses are reported without fear of blame or recrimination, and practices to identify a problem’s scope, root causes, and appropriate solutions. (p. 17-1) 

The report identified several categories of practices that are used to improve healthcare SC: Leadership WalkRounds, Team Training, Comprehensive Unit-based Safety Programs (CUSP), and interventions that implemented multiple methods. (p. 17-13)

WalkRounds “involves leaders “walking around” to engage in face to face, candid discussions with frontline staff about patient safety incidents or near-misses.” (p. 17-16)  “Team training programs focus on enhancing teamwork skills and communication between healthcare providers . . .” (p. 17-17)  CUSP is a multi-step program to assess, intervene in, and reassess a healthcare unit’s SC. (p. 17-19)

The report also covers 17 specific areas where harm/errors can occur and highlights SC aspects associated with two such areas: developing rapid response teams and dealing with alarm fatigue in hospitals. 

Rapid response teams (RRTs) treat deteriorating hospital patients before adverse events occur. (p. 2-1)  Weak SC and healthcare hierarchies are barriers to successful implementation of RRTs. (p. 2-10)

Alarm fatigue occurs because of high exposure to medical device alarms, many of which are loud or false alarms, that lead to desensitization, missed alarms or delayed responses. (p. 13-1)  The cultural aspects of interventions focused on all staff members (not just nurses) assuming responsibility for addressing alarms. (p. 13-6) 

Our Perspective

We have three problems with healthcare’s efforts to reduce harm to patients: (1) the quasi-official healthcare mental model of safety culture is incomplete, (2) healthcare’s assumption that it can model itself on HROs ignores a critical systemic difference, and (3) an inadequate overall system model leads to fragmented, incremental improvement projects.

An inadequate model for SC

Healthcare does not have an adequate understanding of the necessary attributes of a strong SC.  

The features listed in the introduction of this post are necessary but not sufficient for a strong SC.  SC is more than good communications; it is part of the overall cultural system.  This system has feedback loops that can reinforce or extinguish attitudes and behaviors.  The attitudes of people in the system are heavily influenced by their trust in management to do the right thing.  Management’s behavior is influenced by their goals, policy constraints, environmental pressures, and incentives, including monetary compensation.

Top-to-bottom decision making in the system needs to be consistent, which means processes, priorities, practices, and rules should be defined and followed.  Goal conflicts must be consistently handled.  Decision makers must be identified to allow accountability.   Problems must be identified (without retribution except for cause), analyzed, and permanently fixed.

Lack of attention to the missing attributes is one reason that healthcare SC has been slow to strengthen and unfavorable patient outcomes are still at unacceptable levels. 

Healthcare is not a traditional HRO

The healthcare system looks to HROs for inspiration on SC but does not recognize one significant difference between a traditional HRO and healthcare.

When we consider other HROs, e.g., nuclear power plants, off-shore drilling operations, or commercial aviation, we understand that they have significant interactions with their respective environments, e.g., regulators, politicians, inspectors, suppliers, customers, activists, etc. 

Healthcare is different because its customers are basically the feedstock for the “factory” and healthcare has to accept those inputs “as is”; in other words, unlike a nuclear power plant, healthcare cannot define and enforce a set of specifications for its inputs.  The inputs (patients) arrive in a wide range of “as is” conditions, from simple injuries to multiple, interacting ailments.  The healthcare system has to accomplish two equally important objectives: (1) correctly identify a patient’s problem(s) and (2) fix them in a robust, cost-effective manner.  SC in the first phase should focus on obtaining the correct diagnosis; SC in the second phase should focus on performing the prescribed corrective actions according to approved procedures, and ensuring that expected results occur. 

Inadequate models lead to piecemeal interventions      

Healthcare’s simplistic mental model for SC is part of an inaccurate mental model for the overall system.  The current system model is fragmented and leads researchers and practitioners to think small (on silos) when they could be thinking big (on the enterprise).  An SC intervention that focuses on tightening process controls in one small area cannot move the needle on system-wide SC or overall patient outcomes.  For more on systems models, systemic challenges, and narrow interventions, see our Oct. 9, 2019 and Nov. 9,2020 posts.  Click on the healthcare label below to see all of the related posts.

Bottom line: Healthcare SC can have a direct impact on the probabilities that specific harms will occur, and their severity if they do but accurate models of culture are essential. 

 

*  Agency for Healthcare Research and Quality, “Culture of Safety” (Sept. 2019).  Accessed May 4, 2021.  AHRQ is an organization within the U.S. Department of Health and Human Services.  Its mission includes producing evidence to make health care safer.

**  The “blame-free” environment has evolved into a “just culture” where human errors, especially those caused by the task system context, are tolerated but taking shortcuts and reckless behavior are disciplined.  Click on the just culture label for related posts.

***  U.S. Dept. of Health and Human Services Office of Inspector General, “Patient Safety Organizations: Hospital Participation, Value, and Challenges,” OEI-01-17-00420, Sept. 2019.

****  K.K. Hall et al, “Making Healthcare Safer III: A Critical Analysis of Existing and Emerging Patient Safety Practices,” AHRQ Pub. No. 20-0029-EF.  (Rockville, MD: AHRQ) March 2020.  This is a 1400 page report so we are only reporting relevant highlights.

Organizational Change and System Dynamics Insights from The Tipping Point by Malcolm Gladwell

The Tipping Point* is a 2002 book by Malcolm Gladwell (who also wrote Blink) that uses the metaphor of a viral epidemic to explain how some phenomenon, e.g., a product**, an idea, or a social norm, can suddenly reach a critical mass and propagate rapidly through society.  Following is a summary of his key concepts.  Some of his ideas can inform strategies for implementing organizational change, especially cultural change, and reflect attributes of system dynamics that we have promoted on Safetymatters.

In brief, epidemics spread when they have the right sort of people to transmit the infectious agent, the agent itself has an attribute of stickiness, and the environment supports the agent and facilitates transmission. 

People

An epidemic thrives on three different types of people: people who connect with lots of other people, people who learn about a new product or idea and are driven to tell others, and persuasive people who sell the idea to others.  All these messengers drive contagiousness although all three types are not required for every kind of epidemic.

Stickiness

A virus needs to attach itself to a host; a new product promotion needs to be memorable, i.e., stick in people’s minds and spur them to action, for example Wendy’s “Where’s the beef?” campaign or the old “Winston tastes good . . .” jingle.  Information about the new product or idea needs to be packaged in a way that makes it attractive and difficult to resist.

Context

General and specific environmental characteristics can encourage or discourage the spread of a phenomenon.  For a general example in the social environment consider the Broken Windows theory which holds that intolerance of the smallest infractions can lead to overall reductions in crime rates.

At the more specific level, humans possess a set of tendencies that can be affected by the particular circumstances of their immediate environment.  For example, we are more likely to comply with someone in a uniform (a doctor, say, or a police officer) than a scruffy person in jeans.  If people believe there are many witnesses to a crime, it’s less likely that anyone will try to stop or report the criminal activity; individual responsibility is diffused to the point of inaction.      

Our Perspective

We will expand some of Gladwell’s notions to emphasize how they can be applied to create organizational changes, including cultural change.  In addition, we’ll discuss how the dynamics he describes square with some aspects of system dynamics we have promoted on Safetymatters.

Organizational change

Small close-knit groups have the potential to magnify the epidemic potential of a message or idea.  “Close-knit” means people know each other well and even store information with specific individuals (the subject matter experts) to create a kind of overall group memory.  These bonds of memory and peer pressure can facilitate the movement of new ideas into and around the group, affecting the group’s shared mental models of everything from the immediate task environment to the larger outside world.  Many small movements can create larger movements that manifest as new or modified group norms.

In a product market, diffusion moves from innovators to early adopters to the majority and finally the laggards.  A similar model of diffusion can be applied in a formal organization.  Organizational managers trying to implement cultural changes should consider this diffusion model when they are deciding who to appoint to initiate, promote, and promulgate new or different cultural values or behaviors.  Ideally, they should start with well-connected, respected people who buy into the new attributes, can explain them to others, and influence others to try the new behaviors.

System dynamics

This whole book is about how intrusions can disrupt an existing social system, for good or bad, and result in epidemic, i.e., nonlinear effects.  This nonlinearity helps explain why systems can be operating more or less normally then suddenly veer into failure.  Active management deliberately tries to create such changes to veer into success.  Just think about how social media has upset the loci of power in our society: elected leaders and experts now have larger megaphones but so does the mob. 

That said, Gladwell presents a linear, cause-and-effect model for change.  He does not consider more complex system features such as feedback loops or deliberate attempts to modify, deflect, co-opt or counteract the novel input.  For example, a manager can try to establish new behaviors by creating a reinforcing loop of rewards and recognition in a small group, and then recreating it on an ever-larger scale.

Bottom line: This is easy reading with lots of interesting case studies and quotes from talking head PhDs.  The book comes across as a long magazine article. 

 

*  M Gladwell, The Tipping Point (New York: Back Bay Books/Little, Brown and Co.) 2000 and 2002.

**  “Product” is used in its broadest sense; it can mean something physical like a washing machine, a political campaign, a celebrity wannabe, etc.

Implications of Randomness: Lessons from Nassim Taleb

Most of us know Nassim Nicholas Taleb from his bestseller The Black Swan. However, he wrote an earlier book, Fooled by Randomness*, in which he laid out one of his seminal propositions: a lot of things in life that we believe have identifiable, deterministic causes such as prescient decision making or exceptional skills, are actually the result of more random processes. Taleb focuses on financial markets but we believe his observations can refine our thinking about organizational decision making, mental models, and culture.

We'll begin with an example of how Taleb believes we misperceive reality. Consider a group of stockbrokers with successful 5-year track records. Most of us will assume they must be unusually skilled. However, we fail to consider how many other people started out as stockbrokers 5 years ago and fell by the wayside because of poor performance. Even if all the stockbrokers were less skilled than a simple coin flipper, some would still be successful over a 5 year period. The survivors are the result of an essentially random process and their track records mean very little going forward.

Taleb ascribes our failure to correctly see things (our inadequate mental models) to several biases. First is the hindsight bias where the past is always seen as deterministic and feeds our willingness to backfit theories or models to experience after it occurs. Causality can be very complex but we prefer to simplify it. Second, because of survivorship bias, we see and consider only the current survivors from an initial cohort; the losers do not show up in our assessment of the probability of success going forward. Our attribution bias tells us that successes are due to skills, and failures to randomness.

Taleb describes other factors that prevent us from being the rational thinkers postulated by classical economics or Cartesian philosophy. One set of factors arises from how are brains are hardwired and another set from the way we incorrectly process data presented to us.

The brain wiring issues include the work of Daniel Kahneman who describes how we use and rely on heuristics (mental shortcuts that we invoke automatically) to make day-to-day decisions. Thus, we make many decisions without really thinking or applying reason, and we are subject to other built-in biases, including our overconfidence in small samples and the role of emotions in driving our decisions. We reviewed Kahneman's work at length in our Dec. 18, 2013 post. Taleb notes that we also have a hard time recognizing and dealing with risk. Risk detection and risk avoidance are mediated in the emotional part of the brain, not the thinking part, so rational thinking has little to do with risk avoidance.

We also make errors when handling data in a more formal setting. For example, we ignore the mathematical truth that initial sample sizes matter greatly, much more than the sample size as a percentage of the overall population. We also ignore regression to the mean, which says that absent systemic changes, performance will eventually return to its average value. More perniciously, ignorant or unethical researchers will direct their computers to look for any significant relationship in a data set, a practice that can often produce a spurious relationship because all the individual tests have their own error rates. “Data snoops” will define some rule, then go looking for data that supports it. Why are researchers inclined to fudge their analyses? Because research with no significant result does not get published.

Our Perspective

We'll start with the obvious: Taleb has a large ego and is not shy about calling out people with whom he disagrees or does not respect. That said, his observations have useful implications for how we conceptualize the socio-technical systems in which we operate, i.e., our mental models, and present specific challenges for the culture of our organizations.

In our view, the three driving functions for any system's performance over time are determinism (cause and effect), choice (decision making), and probability. At heart, Taleb's world view is that the world functions more probabilistically than most people realize. A method he employs to illustrate alternative futures is Monte Carlo simulation, which we used to forecast nuclear power plant performance back in the 1990s. We wanted plant operators to see that certain low-probability events, i.e., Black Swans**, could occur in spite of the best efforts to eliminate them via plant design, improved equipment and procedures, and other means. Some unfortunate outcomes could occur because they were baked into the system from the get-go and eventually manifested. This is what Charles Perrow meant by “normal accidents” where normal system performance excursions go beyond system boundaries. For more on Perrow, see our Aug. 29,2013 post.

Of course, the probability distribution of system performance may not be stationary over time. In the most extreme case, when all system attributes change, it's called regime change. In addition, system performance may be nonlinear, where small inputs may lead to a disproportionate response, or poor performance can build slowly and suddenly cascade into failure. For some systems, no matter how specifically they are described, there will inherently be some possibility of errors, e.g., consider healthcare tests and diagnoses where both false positives and false negatives can be non-trivial occurrences.

What does this mean for organizational culture? For starters, the organization must acknowledge that many of its members are inherently somewhat irrational. It can try to force greater rationality on its members through policies, procedures, and practices, instilled by training and enforced by supervision, but there will always be leaks. A better approach would be to develop defense in depth designs, error-tolerant sub-systems with error correction capabilities, and a “just culture” that recognizes that honest mistakes will occur.

Bottom line: You should think awhile about how many aspects of your work environment have probabilistic attributes.

 

* N.N. Taleb, Fooled by Randomness, 2^nd ed. (New York: Random House) 2004.

** Black swans are not always bad. For example, an actor can have one breakthrough role that leads to fame and fortune; far more actors will always be waiting tables and parking cars.

Setting the Bar for Healthcare: Patient Care Goals from the Joint Commission

Joint Commission HQ

The need for a more effective safety culture (SC) in the field of healthcare is acute: every year tens of thousands of patients are injured or unnecessarily die while in U.S. hospitals. The scope of the problem became widely known known with the publication of “To Err is Human: Building a Safer Health System”* in 2000. This report included two key observations: (1) the cause of the injuries and deaths is not bad people in health care, rather the people are working in bad systems that need to be made safer and (2) legitimate liability concerns discourage the reporting of errors, which means less feedback to the system and less learning from mistakes.

It's 20 years later. Is the healthcare system safer than it was in 2000? Yes. Is safety performance at a satisfactory level? No.

For evidence, we need look no further than a Nov. 18, 2019 blog post** byDr. Mark Chassin, president and CEO of the Joint Commission (JC), the entity responsible for establishing standards for healthcare functions and patient care, and evaluating, accrediting, and certifying healthcare organizations based on their compliance with the standards.

Dr. Chassin summarized the current situation as follows: “The health care industry has directed a substantial amount of time, effort, and resources at solving the problems, and we have seen some progress. That progress has typically occurred one project at a time, with hard-working quality professionals applying a “one-size-fits-all” best practice to address each problem. The resulting improvements have been pretty modest, difficult to sustain, and even more difficult to spread.”

Going forward, he says the industry can make substantial progress by committing to zero harm, overhauling the organizational culture, and utilizing proven process improvement techniques. He singles out the aviation and nuclear power industries for having similar commitments.

But achieving substantial, sustained improvement is a big lift. To get a feel for how big, let's look at the 2020 goals and strategies the JC has established for patient care in hospitals, in other words, where the performance bar is set today.*** We will try to inform your own judgment about their scope and sufficiency by comparing them with corresponding activities in the nuclear power industry.

1. Identify patients correctly by using at least two ways to identify them.

This is a major challenge in a hospital where many patients are entering and leaving the system every day, being transferred to and from different departments, and being treated by multiple individuals who have different roles and ranks, and are treating patients at different levels of intensity for different periods of time. There is really no analogue in the closed, controlled personnel environment of a power plant.

2. Improve staff communication by getting important test results to the right staff person on time.

This should be a familiar challenge to people in any organization, including a power plant, where functions may exist in different organizational silos with their own procedures, vocabulary, and priorities.

3. Use medicines safely by labeling medicines that are not labeled, taking extra care with patients on blood thinners, and managing patients' medicine records for accuracy, completeness, and possible interactions.

This is similar to requirements to accurately label, control, and manage the use of all chemicals used in an industrial facility.

4. Use alarms safely by ensuring that alarms on medical equipment are heard and responded to on time.

In a hospital, it is a problem when multiple alarms are going off at the same time, with differing degrees of urgency for personnel attention and response. In power plants, operators have been known to turn off alarms that are reporting too many false positives. These situations call out for operating and maintenance standards and practices that ensure all activated alarms are valid and deserving of a response.

5. Prevent infection by adhering to Centers for Disease Control or World Health Organization hand cleaning guidelines.

The aim is to keep bad bugs from circulating. Compare this prctice to the myriad procedures, personnel, and equipment dedicated to ensuring nuclear power plant radioactivity is kept in an identified, controlled, and secure environment.

6. Identify patient safety risks by reducing the risk for suicide.

Compare this with the wellness, fitness for duty, and behavioral observation programs at every nuclear power plant.

7. Prevent mistakes in surgery by making sure that the correct surgery is done on the correct patient and at the correct place on the patient’s body, and pausing before the surgery to make sure that a mistake is not being made.

This is similar to tailgate meetings before maintenance activities and using the STAR (Stop-Think-Act-Review) approach before and during work. Think of the potential for error in mirror-image plants; people are bi-lateral but subject to the similar risks.

Our Perspective

The JC's set of goals is thin gruel to show after 20 years. In our view, efforts to date reflect two major shortcomings: a lack of progress in defining and strengthening SC, and a lack of any shared understanding of what the relevant system consists of, how it functions, and how to improve it.

Safety Culture

Our July 31, 2020 post on When We Do Harm by Dr. Danielle Ofri discussed the key attributes for a strong healthcare SC, i.e., one where the probability of errors is much lower than it is today. In Ofri's view, the primary cultural attribute for reducing errors is a willingness of individuals to assume ownership and get the necessary things done, even if it's not in their specific job description, amid a diffusion of responsibility in their task environment. Secondly, all members of the organization, regardless of status, should have the ability (or duty even) to point out problems and errors without fear of retribution. The culture should regard reporting an adverse event as a routine and ordinary task. Third, organizational leaders, including but not limited to senior managers, must encourage criticism, forbid scapegoating, and not allow hierarchy and egos to overrule what is right and true. There should be deference to proven expertise and widely held authority to say “stop” when problems become apparent.

The Healthcare System

The healthcare system includes the providers, the supporting infrastructure, external environmental factors, e.g., regulators and insurance companies, the patients and their families, and all the interrelationships and dynamics between these components. An important dynamic is feedback, where the quality and quantity of output from one component influences performance in other system components. System dynamics create homeostasis, fluctuations, and all levels of performance from superior to failure. Other organizational variables, e.g., management decision-making practices and priorities, and the compensation scheme, provide context for system functioning. For more on system attributes, please see our Oct.9, 2019 post or click the healthcare label.

Bottom line: Compare the JC's efforts with the vast array of safety and SC-related policies, procedures, practices, activities, and dedicated personnel in your workplace. Healthcare has a long way to go.

* Institute of Medicine (L.T. Kohn et al), “To Err Is Human: Building a Safer Health System” (Washington, D.C.: The National Academies Press) 2000. Retrieved Nov. 5, 2020.

** M. Chassin, “To Err is Human: The Next 20 Years,” blog post (Nov. 18, 2019).  Retrieved Nov. 1, 2020.

*** The Joint Commission, “2020Hospital National Patient Safety Goals,” simplified version (July, 2020). Retrieved Nov. 1, 2020.

How to Consider Unknown Unknowns: Hints from McKinsey

Our July 31, 2020 post on medical errors discussed the importance of the “differential diagnosis” where a doctor thinks “I believe this patient has X but what else could it be?” We can usually consider that as a decision situation with known unknowns, i.e., looking for another needle in a haystack based on the available evidence. But what if you don’t know what you don’t know? How do you create other possibilities, threats or opportunities, or different futures out of thin air? A 2015 McKinsey article* provides some suggestions for getting started. There is nothing really new but it reiterates some important points we have been making here on Safetymatters.

The authors begin by noting executives’ decision making processes often coalesce around “managing the probable,” i.e., attempting to fit a current decision into a model that has worked before. The questions they ask and the data they seek tend to narrow, not expand, the decision and its context. This is an efficient way to approach many everyday decisions but excessively simple models are not appropriate for complicated decisions like how to approach a changing market or define a market that does not yet exist. All models constrain the eventual solution and simple models constrain it the most, perhaps leading to a totally wrong answer.

Decision situations that are dramatically different, complex, and uncertain require a more open-ended approach, the authors call it “leading the possible.” In such situations, decision makers should acknowledge they don’t know how uncertain environmental conditions will unfold or complex systems will evolve. The authors propose three non-traditional mental habits to identify and explore the possibilities.

Ask different questions

Ask questions that open up possibilities rather than narrowing the discussion and constraining the solution. Sample questions include: What do I expect not to find? How could I adjust to the unexpected? What might I be discounting or explaining away too quickly? What would happen if I changed one or more of my core assumptions? We would add: Is fear of missing out prodding me to move too rashly or complacency allowing me to not move at all?

As Hans Rosling said: “Beware of simple ideas and simple solutions. . . . Welcome complexity.” (see our Dec. 3, 2018 post)

Take multiple perspectives

Decision makers, especially senior managers, need to escape the echo chamber of the sycophants who surround them. They should consider how people who are very different from themselves might view the same decision situation. They can consult people who are knowledgeable but frustrating or irritating, or outside their usual internal circle such as junior staff, or even dissatisfied customers. Such perspectives can be insightful and surprising.

Other thought leaders have suggested similar approaches. For example, Ray Dalio proposes thoughtful disagreement where decision makers seek out brilliant people who disagree with them to gain a deeper understanding of decision situations (see our April 17, 2018 post) or Charlan Nemeth on the usefulness of authentic dissent in decision situations (see our June 29, 2020 post).

Recognize systems

The authors’ appreciation for systems thinking mirrors what we’ve been saying for years. (For example, see our Jan. 6, 2017 post.) Decision makers should be looking at the evolution of the forest, not examining individual trees. We need to acknowledge and accept that “Elements in a system can be connected in ways that are not immediately apparent.” The widest view is the most powerful but people have “been trained to follow our natural inclination to examine the component parts. We assume a straightforward and linear connection between cause and effect. Finally, we look for root causes at the center of problems. In doing these things, we often fail to perceive the broader forces at work.”


The authors realize that leaders who can apply the new habits may have different attributes than earlier senior managers. Traditional leaders are clear, confident, and decisive. However, their preference for managing the probable leaves them more open to being blindsided. In contrast, new leaders need to exhibit “humility, a keen sense of their own limitations, an insatiable curiosity, and an orientation to learning and development.”

Our Perspective

This article promotes more expansive mental models for decision making in formal organizations, models that deemphasize reliance on reductionism and linear, cause-effect thinking. We applaud the authors’ intent.

McKinsey is pretty good at publishing small bite “news you can use” articles. However, they do not contain any of the secret sauce for which McKinsey charges its clients top dollar.

Bottom line: Some of you don’t want to read 300 page books on management so here’s an 8 page article with a few good points.


* Z. Achi and J.G. Berger, “Delighting in the Possible,” McKinsey Quarterly (March 2015).

Culture in Healthcare: Lessons from When We Do Harm by Danielle Ofri, MD

In her book*, Dr. Ofri takes a hard look at the prevalence of medical errors in the healthcare system.  She reports some familiar statistics** and fixes, but also includes highly detailed case studies where errors large and small cascaded over time and the patients died.  This post summarizes her main observations.  She does not provide a tight summary of a less error-prone healthcare culture but she drops enough crumbs that we can infer its desirable attributes.

Healthcare is provided by a system

The system includes the providers, the supporting infrastructure, and factors in the external environment.  Ofri observes that medical care is exceedingly complicated and some errors are inevitable.  Because errors are inevitable, the system should emphasize error recognition and faster recovery with a goal of harm reduction.

She shares our view that the system permits errors to occur so fixes should focus on the system and not on the individual who made an error.***  System failures will eventually trap the most conscientious provider.  She opines that most medical errors are the result of a cascade of actions that compound one another; we would say the system is tightly coupled.

System “improvements” intended to increase efficiency can actually reduce effectiveness.  For example, electronic medical records can end up dictating providers’ practices, fragmenting thoughts and interfering with the flow of information between doctor and patient.****  Data field defaults and copy and paste shortcuts can create new kinds of errors.  Diagnosis codes driven by insurance company billing requirements can distort the diagnostic process.  In short, patient care becomes subservient to documentation.

Other changes can have unforeseen consequences.  For example, scheduling fewer working hours for interns leads to fewer diagnostic and medication errors but also results in more patient handoffs (where half of adverse medical events are rooted.)    

Aviation-inspired checklists have limited applicability

Checklists have reduced error rates for certain procedures but can lead to unintended consequences, e.g., mindless check-off of the items (to achieve 100% completion in the limited time available) and provider focus on the checklist while ignoring other things that are going on, including emergent issues.

Ofri thinks the parallels between healthcare and aviation are limited because of the complexity of human physiology.  While checklists may be helpful for procedures, doctors ascribe limited value to process checklists that guide their thinking.

Malpractice suits do not meaningfully reduce the medical error rate

Doctors fear malpractice suits so they practice defensive medicine, prescribing extra tests and treatments which have their own risks of injury and false positives, and lead to extra cost.  Medical equipment manufacturers also fear lawsuits so they design machines that sound alarms for all matters great and small; alarms are so numerous they are often simply ignored by the staff.

Hospital management culture is concerned about protecting the hospital’s financial interests against threats, including lawsuits.  A Cone of Silence is dropped over anything that could be considered an error and no information is released to the public, including family members of the injured or dead patient.  As a consequence, it is estimated that fewer than 10% of medical errors ever come to light.  There is no national incident reporting system because of the resistance of providers, hospitals, and trial lawyers.

The reality is a malpractice suit is not practical in the vast majority of cases of possible medical error.  The bar is very high: your doctor must have provided sub-standard care that caused your injury/death and resulted in quantifiable damages.  Cases are very expensive and time-consuming to prepare and the legal system, like the medical system, is guided by money so an acceptable risk-reward ratio has to be there for the lawyers.***** 

Desirable cultural attributes for reducing medical errors

In Ofri’s view, culture includes hierarchy, communications skill, training traditions, work ethic, egos, socialization, and professional ideals.  The primary cultural attribute for reducing errors is a willingness of individuals to assume ownership and get the necessary things done amid a diffusion of responsibility.  This must be taught by example and individuals must demand comparable behavior from their colleagues.

Providing medical care is a team business

Effective collaboration among team members is key, as is the ability (or duty even) of lower-status members to point out problems and errors without fear of retribution.  Leaders must encourage criticism, forbid scapegoating, and not allow hierarchy and egos to overrule what is right and true.  Where practical, training should be performed in groups who actually work together to build communication skills.

Doctors and nurses need time and space to think

Doctors need the time to develop differential diagnosis, to ask and answer “What else could it be?”  The provider’s thought process is the source of most diagnostic error, and subject to explicit and implicit biases, emotions, and distraction.  However, stopping to think can cause delays which can be reported as shortcomings by the tracking system.  The culture must acknowledge uncertainty (fueled by false positives and negatives), address overconfidence, and promote feedback, especially from patients.

Errors and near misses need to be reported without liability or shame.

The culture should regard reporting an adverse event as a routine and ordinary task.  This is a big lift for people steeped in the hierarchy of healthcare and the impunity of its highest ranked members.  Another factor to be overcome is the reluctance of doctors to report errors because of their feelings of personal and professional shame.

Ofri speaks favorably of a “just culture” that recognizes that unintentional error is possible, but risky behavior like taking shortcuts requires (system) intervention, and negligence should be disciplined.  In addition, there should not be any bias in how penalties are handed out, e.g., based on status.

In sum, Ofri says healthcare will always be an imperfect system.  Ultimately, what patients want is acknowledgement of errors and apology for them from doctors.

Our Perspective

Ofri’s major contribution is her review of the evidence showing how pervasive medical errors are and how the healthcare industry works overtime to deny and avoid responsibility for them.

Her suggestions for a safer healthcare culture echo what we’ve been saying for years about the attributes of a strong safety culture.  Reducing the error rates will be hard for many reasons.  For example, Ofri observes medical training forges a lifelong personal identity and reverence for tradition; in our view, it also builds in resistance to change.  The biases in decision making that she mentions are not trivial.  For one discussion of such biases, see our Dec. 18, 2013 review of Daniel Kahneman’swork.

Bottom line: After you read this, you will be clutching your rosary a little tighter if you have to go to a hospital for a major injury or illness.  You are more responsible for your own care than you think.

*  D. Ofri, When We Do Harm (Boston: Beacon Press, 2020).

**  For example, a study reporting that almost 4% of hospitalizations resulted in medical injury, of which 14% were fatal, and doctors’ diagnostic accuracy is estimated to be in the range of 90%.

***  It has been suggested that the term “error” be replaced with “adverse medical event” to reduce the implicit focus on individuals.

****  Ofri believes genuine conversation with a patient is the doctor’s single most important diagnostic tool.

***** As an example of the power of money, when Medicare started fining hospitals for shortcomings, the hospitals started cleaning up their problems.