Case Presentation: A 90-year-old-male with history of hypertension, coronary artery disease and chronic lymphocytic lymphoma presented with shortness of breath that had been getting worse over the previous few weeks, particularly at night and with lying down. He also reported increased cough and worsening leg swelling. In the ED, he was hypoxic (84% on room air), tachypneic (RR 32), afebrile, and hypotensive (BP 82/52). On exam, he had bibasilar crackles, JVP at 10 cm, and 1+ bilateral pitting edema. Initial labs were notable for: WBC 37K/uL (60% lymphocytes), creatinine 3.7 mg/dL, lactate 5.7 mmol/L, procalcitonin 0.22 ng/mL, high-sensitivity troponin T 513 and 564 ng/L, and NT-proBNP 10,417 pg/mL. Chest CT showed confluent areas of ground-glass opacity, likely pulmonary edema, consolidation in the right upper lobe, and bilateral moderate pleural effusions. He was started on broad spectrum antibiotics for suspected pneumonia, and given a small intravenous fluid bolus, with improvement in lactate levels. However, he subsequently developed escalating oxygen requirements, became more hypotensive, and was transferred to the medical ICU.In the ICU, cardiology team was consulted, who recommended pressor support and avoiding further fluid challenges given high cardiac filling pressures, and when appropriate, to begin volume reduction diuresis. Patient had another episode of hypotension, and based on his prior fluid responsiveness, ICU team favored treating for distributive shock with further volume resuscitation. Patient’s respiratory status continued to worsen. He had an echocardiogram that showed new reduced ejection fraction (30-35%), dilated IVC, estimated right ventricular filling pressures 49 mm. Despite subsequent diuresis and inotropic support, he deteriorated clinically, was transitioned to comfort measures status, and passed away.
Discussion: Management of mixed cardiogenic and septic shock is complex and requires close hemodynamic monitoring. In this case, patient was admitted with clear signs of volume overload, with possible superimposed pneumonia. He was initially treated primarily for septic shock, which in the absence of adequate assessment of volume status, led to progressive cardiopulmonary decline.As part of the UPSIDE study, adjudicators in this case thought that diagnostic error was ‘highly likely.’ The error was thought to be secondary to the following diagnostic process faults:1) Suboptimal weighing of history (orthopnea, leg swelling), physical exam (bibasilar crackles, elevated JVP) and diagnostic data (elevated NT-proBNP, pulmonary edema on chest CT) 2) Suboptimal prioritizing/delay in considering the diagnosis (mixed cardiogenic and distributive shock)3) Failure or delay in physiologic monitoring (assessment of volume status), including by consultation service (cardiology team did not re-assess the patient)These multiple diagnostic process faults lead to an error that may have contributed to patient’s death.
Conclusions: Coming up with the right diagnosis is not a single act but a ‘process’ where the primary team needs to carefully and constantly evaluate available subjective and objective data, as well as inputs from other clinical and non-clinical providers, to come up with the best possible understanding of the underlying disease process. This is essential for minimizing anchoring and other biases, and thereby reducing chances of diagnostic errors and possible harm to patients.