“Humans err,” said Dr. Saul N. Weingart, director of the Center for Patient Safety at Dana-Farber Cancer Institute, in a sweeping analysis of medical error, in a Grand Rounds lecture for NIH fellows in the Clinical Center in 2006. Cognitive psychologists recognize three models of error in human performance; skill-based, rule-based, and knowledge-based. Human factor scientists call skill-based errors slips and rule-based and knowledge-based errors mistakes.
Recognize missteps for what they are, said Weingart. Skill-based behavior is rapid, effortless, unconscious—things you do automatically, such as brushing your teeth, buttoning your shirt, driving to work. Characteristic slips involve
· Capture (you’re supposed to do one sequence of steps, A,B,C,F, and instead you do ABCD, because ABCD is more common—as when you load up the family car for vacation and absentmindedly drive toward the office)
· Description error (correct action, wrong object—pouring syrup in your coffee or milk on your pancakes)
· Associative activation (having the wrong cue trigger a particular behavior—the doorbell rings and you answer the phone)
· Loss of activation (striding purposefully into a room and not remembering why you are there).
“These slips are common across all cultures,” says Weingart. “They are part of the human phenotype.” We’re built to perform many functions on an unconscious level so that we can think about other things that are more interesting and important. “But those automatisms allow us to get into trouble. And they are degraded by the things you’d expect, such as fatigue, illness, alcohol, sleep deprivation, and boredom.”
Rule-based behavior (when it’s raining, take an umbrella) is important in medical training. Using mnenomics, we train physicians in rule-based behaviors with code algorithms, said Weingart. “The trick is to apply it in the correct setting.” Rule-based mistakes occur when we apply the wrong rule — for example, going down a code algorithm but choosing the wrong arm.
Knowledge-based behavior involves solving novel problems, coming at them afresh. Knowledge-based mistakes, of which there are many, have to do with biased memory and intellectual convenience. We like to use the solution we used last time, or we don’t like to change our mind once we’ve come up with an initial hypothesis, or we’re overconfident about what we decide.
“Training, communicating, and reminding are weak interventions. Only by probing deeply can we understand the conditions that lined up to allow something unfortunate to happen. You’ve got to change the system.”
“Don’t rely on memory or vigilance,” said Weingart, “or expect people to perform well when fatigued.” Simplify routines, reduce handoffs, build in redundancies and provide better information. Create forcing functions (making it impossible to do action B without precondition A). Make use of best practices—bar coding, for example, and repeating verbal orders (“the Starbucks rule: I’d like a low-fat no-foam latte”).
Knowing we are faulty machines and that errors are inevitable, said Weingart, we must design medical care systems that make those errors transparent and reduce the number of preventable adverse events and close calls—paying special attention to high-risk patients, “those with the least physiological reserve.”
It’s important, says Weingart, to take care of yourself, take care of the system, and reduce blame and shame. “If we’re not free to talk about errors,” he said, “we’ll never learn from them.”
Pat McNees wrote this article for NIH Clinical Center News, September 2006. It appeared as "Expert explores the anatomy of medical error."
For a fuller discussion of the three models, see Lucius Leape, “Error in Medicine” JAMA, Dec. 21, 1994, 1851-57.
Further reading:
• Reducing Medical Errors (summary of ways to prevent medical errors, from a 1999 Institute of Medicine report, To Err Is Human:Building a Safer Health System.
• Institute of Medicine reports
• More stories about medical errors and quality control
• Controversies in Laboratory Medicine: A Series From the Institute for Quality in Laboratory Medicine (John R. Butterly, MD, Richard E. Horowitz, MD, Medscape, 2006)
• After the Error: Disclosure Responsibilities and Controversies (Sue Evans, MD, MPH, Yale School of Medicine slides, 6-15-14)
• Medical errors and the Institute of Medicine (IOM) (Premier). An excerpt:
Medication errors can occur at any stage of medication administration. These include:
--Ordering: wrong dose, wrong choice of drug,
--Transcribing: wrong frequency of drug administration, missed dose because medication is not transcribed,
--Dispensing: drug not sent in time to be administered at the time ordered, wrong drug, wrong dose,
--Administering: wrong dose of drug administered, wrong technique used to administer the drug, and
--Monitoring: not noting the effects of the given medication.
• Medical errors snare more than one victim (Jamesetta Newland, The Nurse Practitioner, Sept. 2011)
Saul N. Weingart, MD, PhD, a national expert on reducing the prevalence and burden of medical error, was vice president for patient safety and director of the Center for Patient Safety at the Dana-Farber Cancer Institute and an associate professor at Harvard Medical School when he lectured in Lipsett Auditorium, in the NIH Clinical Center, on August 2, 2006.
Recognize missteps for what they are, said Weingart. Skill-based behavior is rapid, effortless, unconscious—things you do automatically, such as brushing your teeth, buttoning your shirt, driving to work. Characteristic slips involve
· Capture (you’re supposed to do one sequence of steps, A,B,C,F, and instead you do ABCD, because ABCD is more common—as when you load up the family car for vacation and absentmindedly drive toward the office)
· Description error (correct action, wrong object—pouring syrup in your coffee or milk on your pancakes)
· Associative activation (having the wrong cue trigger a particular behavior—the doorbell rings and you answer the phone)
· Loss of activation (striding purposefully into a room and not remembering why you are there).
“These slips are common across all cultures,” says Weingart. “They are part of the human phenotype.” We’re built to perform many functions on an unconscious level so that we can think about other things that are more interesting and important. “But those automatisms allow us to get into trouble. And they are degraded by the things you’d expect, such as fatigue, illness, alcohol, sleep deprivation, and boredom.”
Rule-based behavior (when it’s raining, take an umbrella) is important in medical training. Using mnenomics, we train physicians in rule-based behaviors with code algorithms, said Weingart. “The trick is to apply it in the correct setting.” Rule-based mistakes occur when we apply the wrong rule — for example, going down a code algorithm but choosing the wrong arm.
Knowledge-based behavior involves solving novel problems, coming at them afresh. Knowledge-based mistakes, of which there are many, have to do with biased memory and intellectual convenience. We like to use the solution we used last time, or we don’t like to change our mind once we’ve come up with an initial hypothesis, or we’re overconfident about what we decide.
“Training, communicating, and reminding are weak interventions. Only by probing deeply can we understand the conditions that lined up to allow something unfortunate to happen. You’ve got to change the system.”
“Don’t rely on memory or vigilance,” said Weingart, “or expect people to perform well when fatigued.” Simplify routines, reduce handoffs, build in redundancies and provide better information. Create forcing functions (making it impossible to do action B without precondition A). Make use of best practices—bar coding, for example, and repeating verbal orders (“the Starbucks rule: I’d like a low-fat no-foam latte”).
Knowing we are faulty machines and that errors are inevitable, said Weingart, we must design medical care systems that make those errors transparent and reduce the number of preventable adverse events and close calls—paying special attention to high-risk patients, “those with the least physiological reserve.”
It’s important, says Weingart, to take care of yourself, take care of the system, and reduce blame and shame. “If we’re not free to talk about errors,” he said, “we’ll never learn from them.”
Pat McNees wrote this article for NIH Clinical Center News, September 2006. It appeared as "Expert explores the anatomy of medical error."
For a fuller discussion of the three models, see Lucius Leape, “Error in Medicine” JAMA, Dec. 21, 1994, 1851-57.
Further reading:
• Reducing Medical Errors (summary of ways to prevent medical errors, from a 1999 Institute of Medicine report, To Err Is Human:Building a Safer Health System.
• Institute of Medicine reports
• More stories about medical errors and quality control
• Controversies in Laboratory Medicine: A Series From the Institute for Quality in Laboratory Medicine (John R. Butterly, MD, Richard E. Horowitz, MD, Medscape, 2006)
• After the Error: Disclosure Responsibilities and Controversies (Sue Evans, MD, MPH, Yale School of Medicine slides, 6-15-14)
• Medical errors and the Institute of Medicine (IOM) (Premier). An excerpt:
Medication errors can occur at any stage of medication administration. These include:
--Ordering: wrong dose, wrong choice of drug,
--Transcribing: wrong frequency of drug administration, missed dose because medication is not transcribed,
--Dispensing: drug not sent in time to be administered at the time ordered, wrong drug, wrong dose,
--Administering: wrong dose of drug administered, wrong technique used to administer the drug, and
--Monitoring: not noting the effects of the given medication.
• Medical errors snare more than one victim (Jamesetta Newland, The Nurse Practitioner, Sept. 2011)
Saul N. Weingart, MD, PhD, a national expert on reducing the prevalence and burden of medical error, was vice president for patient safety and director of the Center for Patient Safety at the Dana-Farber Cancer Institute and an associate professor at Harvard Medical School when he lectured in Lipsett Auditorium, in the NIH Clinical Center, on August 2, 2006.