Unraveling Diagnostic Error: Delving Deeply to Identify Hidden Human Factors

About 90 minutes after the clinician’s first evaluation, a nurse, while rechecking the child’s vital signs, became concerned: The child was mildly hypotensive, somnolent, and displayed increased lethargy. The physician immediately reevaluated the child and noted a marked deterioration in clinical status, most ominously the presence of higher fever (39oC), hyperventilation, and a sweet odor to her breath. The lab work returned simultaneously, revealing severe anemia with sickled cells, marked leukocytosis, hyperglycemia, and ketoacidosis. The child was urgently admitted to the ICU after a total ED duration of nearly five hours, with a diagnosis of diabetic ketoacidosis, impending coma, sickle cell anemia, and possible septic shock. She was subsequently found to have pneumococcal sepsis and nearly died. Fortunately, she survived, though her hospitalization lasted two weeks.

An investigation into this incident concluded that though there were problems with the triage process and laboratory, the physician was primarily in error and thus responsible for this child’s deterioration and near demise. The investigation concluded that the physician “should have called the lab to find out why the lab studies were not available at the time she first examined this child, and a more thorough examination would have detected the sweet breath so characteristic of diabetic ketoacidosis.” Further, the investigation charged, the physician had “lost track of her obligations.” She was chastised by the hospital’s chief medical officer for “sloppy management” and “marginal competence.” The physician became despondent, anxious, and depressed—a second victim (Scott et al., 2009).

The ED staff was terribly upset with this conclusion, and its physicians instituted a secondary investigation. In addition to the findings regarding the flawed triage process—in which a young inexperienced nurse had been put in charge of triage for the first time without designated backup—and the laboratory delays, this investigation identified a broader texture of the ED environment that affected the physician’s performance, including highly significant contributing factors that collectively resulted in causality at the bedside.

At the time she conducted her examination, the physician was in the 10th hour of a 12-hour shift and had not had a meal or a break in eight hours. She was handling seven patients, two of whom were awaiting admission. She had been asked by the respiratory ICU to insert an arterial line for a patient pending admission, as the ICU staff was backed up, and she had just had a telephone confrontation with a gynecology consultant who was refusing to see a teenage patient with fever and left adnexal rebound tenderness until a pelvic ultrasound was performed, even though radiology was backed up by 60–90 minutes.

The secondary investigation concluded that hunger, fatigue, task saturation, inordinate workload pressure, insufficient system and structure processes impeding the admission of patients, and an interprofessional conflict due to the gynecologist’s belligerence all combined to create a perfect storm—one where other physicians of similar training and experience might well have made the same “errors” in diagnosis. The ED physician felt vindicated by her colleagues.

Yes, the physician was responsible for the care of this child and thus can be held accountable, but only by understanding the human factors degrading her performance can this accountability be put into appropriate context. The acknowledgment of these hidden human factors identified specific opportunities for improvements that could enhance the ED work environment and processes, the well-being of its staff, and (most importantly) the quality and safety of patient care.