DELAYED DIAGNOSIS OF APPENDICITIS IN AN AML PATIENT

Monday, October 25, 2010
Vide Lobby (Sheraton Centre Toronto Hotel)
Robyn I. Schultz, Stephanie Grayson and Satid Thammasitboon, West Virginia University Children's Hospital, Morgantown, WV

Learning objectives: To illustrate a delayed diagnosis of appendicitis due to cognitive biases.

Case information: A 3-year-old female with acute myeloid leukemia presented with febrile neutropenia.  Broad spectrum antibiotics were initiated secondary to concern for sepsis.  The patient was receiving intensive chemotherapy.  The antibiotics were discontinued after five days of negative cultures.  The patient subsequently exhibited abdominal tenderness and nausea.  Severity of pain was difficult to interpret due to patient’s trepidation around physicians.  Over the next few days, she developed low-grade fevers.  The oncologist suspected typhlitis as the cause of abdominal pain in a febrile neutropenic patient. The pediatric surgery agreed with the working diagnosis and recommended only serial abdominal exams.  After three more days of persistent symptoms, abdominal CT scan was obtained.  Findings were consistent with appendicitis without evidence of typhlitis. The patient was taken to the OR for appendectomy.

Discussion: The diagnosis was delayed as a result of cognitive biases.  The oncologists diagnosed typhlitis based on availability heuristic.  The pattern of right lower quadrant pain in a neutropenic patient was rapidly recognized as typhlitis.  This premature closure sets in motion diagnostic momentum by which a particular diagnosis becomes established with inadequate evidence.  As the case had already been framed for typhlitis, the surgeons followed the diagnostic momentum with confidence bias acting on incomplete investigation.  The surgeons then exhibited omission bias through watchful waiting.  They were reluctant to order a potentially unnecessary abdominal film exposing the patient to radiation with an already working diagnosis in place.  A complete evaluation should have included an abdominal CT scan to rule out appendicitis, a more common diagnosis in a 3 year old with RLQ pain. The art of medicine in making the diagnosis must involve a careful analysis of the clinical scenario with cognitive pause to avoid bias when formulating a final diagnosis.