Friday, May 21, 2021

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.

Tuesday, February 2, 2021

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. 


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.


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.


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.