Understanding how cognitive errors and biases can result in misdiagnoses and patient harm is the "new frontier" in patient safety, says Luke A. Perkocha, MD, MBA, a pathologist at Kaiser Permanente Medical Center in San Rafael, CA.
During the last decade, the focus on reducing patient harm has been largely on correctable "systematic" causes of error such as medication errors and wrong-site surgery, says Perkocha. Solutions such as checklists, improved medication labeling, and computerized physician order entry (CPOE) have reduced the likelihood of these errors.
"Now that these 'low-hanging fruit' have been picked, focus is shifting to the cognitive processes of physicians and how predictable pitfalls or biases might lead to erroneous or missed diagnoses, thus cascading into wrong treatment and patient harm," says Perkocha. He gives two examples of recent malpractice claims involving cognitive errors:
• Two specimens were obtained from the same organ on each of two patients on the same day, but they were mixed up in the laboratory. The pathologists looking at the specimens each received one specimen on the correct patient and one on the incorrect patient.
Both pathologists, each unaware of the existence of another similar case, arrived at a diagnostic conclusion based on the first specimen they viewed. "When faced with discordant features of the second specimen, they viewed it through the prism of a diagnosis to which they were anchored," says Perkocha. Both physicians concluded that the second biopsy supported their initial conclusion.
"Both exerted great intellectual effort to explain away the discordant clues, by making allowances for artifact, different preparations, different parts of tumor sampled," says Perkocha.
Neither considered the possibility that there had been a lab error and that the diagnoses represented by the two samples they each viewed were different. They both missed the opportunity to discover the laboratory error; instead, they compounded it. "This case was a natural experiment that elegantly demonstrated the effect of anchoring bias," says Perkocha.
• When a patient presented with low back pain, the primary care physician anchored on his initial working diagnoses of musculoskeletal and disk disease and scheduled the patient for physical therapy.
"He missed the fact that the patient had a mild fever recorded by the nurse checking vital signs," says Perkocha. The possibility of a spinal epidural abscess was not considered, and the patient progressed to serious complications.
The data supporting the effectiveness of any specific approach to avoiding cognitive errors still are scarce, says Perkocha. These strategies have been suggested in the growing literature on diagnostic error to help physicians avoid cognitive errors and malpractice claims:
Be aware of overconfidence.
"Cognitive psychology studies show that we all underestimate the odds we could be wrong; physicians are particularly prone to this," says Perkocha. Some studies have even correlated a higher degree of confidence with a greater likelihood of being wrong, he adds.
"Recognizing that the odds of being wrong are not negligible, even when we feel confident, can help us act accordingly," says Perkocha.
Actively challenge your assumptions.
Perkocha suggests that physicians throw out a key early piece of evidence that led themselves or others to the diagnosis, and see if the diagnosis still fits all the other data.
Consider important potential misses.
Physicians should ask themselves, "Is there a catastrophic diagnosis I should consider?" and focus on diagnostic errors that lead to greatest harm, such as unsuspected vascular events, infections, and cancer.
Actively counter "confirmation bias" by seeking contrary evidence instead of just confirmatory evidence.
Perkocha says a good question for physicians to ask is, "What evidence, if sought, would disprove my diagnosis?"
Shift your perspective.
Physicians can ask, for example, "What if I looked at this case as if I were a cardiologist instead of a gastroenterologist? What would I think of?"
"It is often helpful to take stock of one's own state of mind, level of fatigue, and distraction at the point of making a consequential diagnosis or dismissing a symptom/compliant as trivial," says Perkocha. "This is the personal equivalent of a preoperative time-out."