Amaurosis Fugax Leading to Anchoring Bias

Abstract

Introduction: Anchoring bias is a cognitive bias referring to physicians practices of prioritizing information and data that support their initial impressions, even when the first impressions are wrong. These errors often occur while inheriting patients from a different geography or a long history of medical illnesses. Keeping a broad differential diagnosis, understanding changes in health and illness over time with appropriate evaluation of available laboratory and imaging data is critical to creating and appraising interventions for individuals, families and populations. Initial errors in diagnostic and therapeutic plans can be detrimental to overall outcome in critically ill patients which is even more apparent in the elderly. Case Report: A 70-year-old woman presented with fever, shortness of breath, malaise and dry cough. She was noted to be hypoxemic requiring high flow nasal cannula (HFNC). Her past history included recent diagnosis of giant cell arteritis (GCA) based on left sided acute vision loss for which she was on oral prednisone and methotrexate without pneumocystis pneumonia (PCP) prophylaxis. Her antibody panel was noted to be negative except for elevated ESR and CRP. She was found to be negative for Sars-CoV-2. Bilateral temporal artery biopsies did not reveal GCA (Figure 1). Her respiratory failure progressed into severe ARDS requiring invasive mechanical ventilation. BAL was positive for pneumocystis stain and pneumocystis culture. She was initiated on appropriate treatment for pneumocystis pneumonia and ARDS. Her ICU course was further complicated by new left sided hemiparesis and MRI showed temporal and occipital infarcts. TEE revealed immobile vegetation on all 3 aortic valve leaflets concerning for non-bacterial thrombotic endocarditis. After initiation of anticoagulation, the patient developed spontaneous retroperitoneal bleeding and progressed to multi-organ failure. Her family decided to transition her care to comfort only, and the patient succumbed to her illness. Discussion: Acute vision loss in elderly requires a thorough work up including cerebrovascular events, autoimmune and local ocular causes. A significant anchoring bias towards GCA in the absence of American College of Rheumatology s qualifying criteria for GCA led to initiation of high dose steroids and immunosuppression. Further addition of F-flurodeoxyglucose PET with temporal artery biopsy is shown to increase diagnostic fidelity in doubtful cases. Early evaluation for embolic events with an echocardiogram could have avoided development of PCP ARDS and fulminant embolic strokes in this patient. In addition, it is recommended that all patients on prednisone ≥20 mg daily for more than 30 days should receive PCP prophylaxis.