LAB SAFETY CONCEPTS
The Paradox of Learning from Errors
Why Laboratories Should Embrace Errors as Learning Opportunities
James S. Hernandez, MD, MS
One thing that I have come to recognize is that I can always find lessons on how to manage my laboratory better from other professions. An example of this came to me from Donald M. Berwick, MD, who now heads the Centers for Medicare and Medicaid Services, in his plenary talk at the 11th Annual National Forum on Quality Improvement in Health Care (1).
In 1949, 13 smokejumpers, ages 17–28, lost their lives fighting a forest fire in Mann Gulch, just north of Helena, Montana. The men were fighting in unfamiliar terrain, and the fire was spreading quicker than they could outrun the flames. Their front man, Wag Dodge, invented a solution on the spot, which he later referred to as an escape fire. He purposely lit the bush in front of him and urged the other firefighters to join him in the burnt out safe area. Because their tools were so heavy, he also told the men to drop their Pulaskis, a combination of an axe and pick that is used to fight forest fires. None of the smokejumpers followed Dodge’s radical directions. Most perished, but Dodge survived.
Berwick used this tragedy as an analogy for a similar situation in healthcare. Just like men trapped in a forest fire, medical errors can cause healthcare workers to panic. Berwick’s message to conference participants was to do what is instinctively counterintuitive: meet our errors head on and learn from them. He went on to observe that even though the best medical systems make errors, experience delays, and rely on misinformation, creativity is one of the features that make these organizations resilient and adaptable. By innovating and learning from medical errors, Berwick said, healthcare professionals can create a more ideal system. In the analogy of the Mann Gulch fire, we must drop our Pulaskis, the weighty assumptions that are holding us back.
Like smokejumpers who were asked to drop
their Pulaskis when trapped in a fire,
laboratory professionals need to drop
their weighty assumptions about error reporting,
in order to improve patient safety.
Creating Good from Bad
The point of Berwick’s analogy was that healthcare professionals need to be encouraged to develop innovative ideas that will create a better healthcare system. “Our challenge is to have the courage to name clearly and boldly the problems we have,” Berwick said. “We must find ways to do that without either marginalizing the truth-teller or demoralizing the good people working in these bad systems.” This lesson is especially applicable to errors in the laboratory.
As laboratory professionals, we need to embrace a culture that encourages staff to report and correct lab errors without fear of punishment. Unfortunately, this is not the situation in many laboratories today. Supervisors instinctively desire to find and chastise the “guilty” party, creating an environment that hinders progress toward eliminating errors.
To me, the Just Culture described by David Marx (2) is one of the most important and commonly overlooked methods of error reporting (3). In this model, most errors are traced back to flaws in systems, not humans. For example, let’s look at how the substitution test from the Just Culture could be applied to an error discovered in the laboratory. The lab supervisor would ask, “Is it likely that a similarly trained and experienced bench technologist could make the same mistake, given the same scenario?” If the answer is yes, then the supervisor must look at the bench technologist’s history of violating procedures. If the bench technologist has no history of violating procedures, he or she should not be blamed for the error. Instead, this test would indicate a flaw in the system that the bench technologist would not be responsible for, and viewed as an issue for the laboratory director and management to solve.
Openly Discussing Errors
Another way to confront our errors in the laboratory is to resurrect morbidity and mortality conferences. Such conferences are a close approximation of a systems post-mortem exam, and although common prior to the 1990s, today most laboratories no longer hold this type of meeting.
However, in my institution, we have found that conducting regular morbidity and mortality conferences is very useful for finding and correcting errors in both anatomic pathology and laboratory medicine. For example, we recently discussed a case that led to lab guidelines on assessing specimens for suspected contamination. Our discussion of the case revealed that the patient complained of nausea and vomiting, but did not appear to be critically ill. He had a history of diabetes mellitus and used an insulin pump. The central lab reported the patient’s glucose as 1000 mg/dL along with a note saying “possible contamination; recollection requested.” Our investigation also revealed that a point-of-care glucose test had been performed on the patient at the same time with similar results, and that he did not have an intravenous line in place at the time of the tests.
In evaluating this case, we came to the surprising conclusion that the specimen was in fact not contaminated and that the high glucose level was a true positive. As a result of this thorough investigation, which included applying the substitution test, we determined that any medical technologist could have mistakenly thought this very high glucose level was due to contamination. This analysis caused us to develop guidelines to aid in evaluating specimens with suspected contamination (See Table).
Guidelines for Assessing Suspected Contamination
Use the following guidelines to help evaluate for suspected contamination of samples:
- Na <130 mmol/L and Cl <100 mmol/L and K >5.5 mmol/L
- Glucose >800 mg/dL and creatinine <0.6 mg/dL
- Na >180 mmol/L and K <2.5 mmol/L
The technologist contacts the nurse/physician in charge of the patient and asks the following questions:
- Does the patient have an IV?
- If a high glucose value is in question: Was a bedside point-of-care glucose performed and if so, what is the value?
Based on these two questions, the nurse/physician and the technologist together decide on the need to reject and recollect the specimen.
- If a critical value is in question, the clinical pathologist is notified immediately.
- If recollection is required, the phlebotomist for lab collects, or nurse for nurse collects, must be informed that “this is a critical value redraw.”
- If recollection is ordered, all tubes from the original draw will be recollected and results checked.
Laboratory supervisors and directors can also learn other valuable lessons from our experience. First, do not assume that medical technologists understand either what a contaminated specimen is or the criteria used to evaluate specimens for contamination. We should encourage medical technologists to contact a pathologist or their lab’s medical director if they have questions about whether or not any specimen is contaminated.
Second, open communication between the lab and nursing staff is critical in these situations. Laboratorians should feel comfortable asking nurses if the patient has an intravenous line and if a point-of-care glucose has been performed, and nurses should respond accordingly.
Success in the Future
Flaws and errors lace all medical systems, but we, as healthcare professionals, must strive to overcome them by being innovative and adaptable. As laboratory information is key to patient care, our mission should be to learn from our mistakes. By keeping patient safety an absolute priority, we can not only boost laboratory productivity, but also encourage excellence from our staff.
Dr. Hernandez is a member of the Patient Safety Focus Editorial Board.
- Berwick, Donald M. Escape Fire, Plenary Address, Eleventh Annual National Forum on Quality Improvement in Health Care, New Orleans, MA: December 9, 1999.
- Marx, David. “Patient Safety and the Just Culture: A Primer for Health Care Executives.” April 17, 2001. Trustees of Columbia University in the City of New York.
- Hernandez, J. “How to Modify Staff Behavior That Puts Patients At Risk: The Just Culture Model” Clinical Laboratory News, 2009, Vol 35, 17.