A 20 year-old male presented to a community hospital with multiple days severe post-prandial nausea, vomiting, and abdominal pain. Symptoms were not relieved with pain medications or anti-emetics. The patient underwent esophagogastroduodenoscopy (EGD) showing distal duodenal ulceration suspicious for Crohn’s disease (CD) and biopsies were taken which were reported to be negative. The patient then developed hematemesis, hematochezia, and a localized petechial rash and was transferred to our hospital given the severity of gastrointestinal bleeding. At this time the patient underwent another EGD with colonoscopy showing diffuse ulceration in his stomach, duodenum, terminal ileum, and colon. The presence of edematous terminal ileum with cobblestoning and skip lesions was most concerning for CD. However, biopsies taken showed ischemic changes without evidence of CD. New petechiae developed with subsequent evolution into confluent purpura over the abdomen, legs and ankles bilaterally. Patient history, laboratory, and urine studies were negative for arthralgias, infection and renal disease. Given ischemic changes noted on bowel biopsy a vasculitis work-up was initiated including skin biopsy that revealed a leukocytoclastic vasculitis with direct immunofluorescence positive for IgA deposition in vessels. A diagnosis of HSP was made and the patient was started on high dose corticosteroids with resolution of gastrointestinal issues.
Discussion:
Henoch-Schonlein Purpura (HSP) is a systemic vasculitis related to IgA deposition in small vessels characterized by palpable purpura, polyarthralgias, renal disease, and abdominal pain. In adults, HSP is rare making early diagnosis challenging, especially when gastrointestinal issues precedes the rash. Diagnostic challenge can be complicated by cognitive bias such as the anchoring effect when the clinical presentation so closely mimics another entity. Cognitive bias has been associated with diagnostic inaccuracies leading to errors in disease management and therapeutic choice. EGD with colonoscopy appeared grossly consistent with CD and CD has a peak in incidence at approximately 20 years of age. The initial presentation appeared consistent with CD causing anchoring, which ultimately delayed diagnostic work-up when an atypical petechial rash presented. With clinical suspicion for HSP and an initial skin biopsy that indicates leukocytoclastic vasculitis, follow-up with direct immunofluorescence has a sensitivity and specificity for HSP of 0.81 and 0.83 respectively and a positive predictive value as high as .84 when timed early after the appearance of lesions.
Conclusions:
This case presentation highlights the importance of recognizing cognitive bias such as the anchoring effect particularly when atypical findings are present. Recognition of anchoring in this case could have lead to earlier diagnostic evaluation of this patient’s rash while decreasing cost of care and time to correct diagnosis.