A 73-year-old woman (height, 157 cm; weight, 43 kg) visited a physician 7 days before admission to our hospital for dizziness, loss of appetite, and fatigue, but no specific issues could be determined, and the patient returned home without receiving treatment. On the day of admission, she revisited the physician because she had a fever of 38°C and dyspnea. Her oxygen saturation was 86% in ambient room air, and nasopharyngeal swab polymerase chain reaction for SARS-Cov-2 revealed that she had COVID-19. The patient was transferred to our hospital under oxygen administration for respiratory management.
When the patient arrived at our hospital, her oxygen saturation was 87% (blood pressure, 152/68 mmHg; heart rate, 73 bpm) with oxygen administered at 2 L/min via a nasal cannula. We changed the oxygen administration method to OxyMaskTM (Southmedic Inc., Barrie, ON, Canada) at 7 L/min, and her oxygen saturation increased to 92%. Subsequently, she was admitted to the intensive care unit. Although she had no underlying conditions that would increase her risk for severe COVID-19, shortly after admission, her oxygen saturation dropped to approximately 90% and her respiratory rate increased to 40/min. We decided to start high-flow nasal cannula therapy at 50 L/min, FIO2 50%.
Her hemogram showed a white blood cell count of 7400/μl and hemoglobin 12.4 g/dl. Her blood biochemistry tests only showed mild elevation in transaminases (glutamic oxaloacetic transaminase, 46 U/l; glutamic pyruvic transaminase, 36 U/l), and her renal function tests and electrolytes were within the normal limits. Arterial blood gas analysis performed before high-flow oxygen therapy was suggestive of type 1 respiratory failure (pH, 7.496; PO2, 52.2 mmHg; PCO2, 28.5 mmHg). Initial chest computed tomography (CT) (Fig. 1) showed bilateral and peripheral predominant consolidation and an air bronchogram.
She was managed according to our institutional protocol (inhalational oxygen concentration, steroids, anticoagulation, tocilizumab, rehabilitation) and maintaining oxygen saturation 92%. Although she temporarily needed an FIO2 of 0.8 to maintain her peripheral oxygen saturation, we did not intubate her because she did not exert effort during ventilation, she strongly preferred not to be intubated, and her only symptom was lightheadedness when standing. On the 11th day of hospitalization, although her oxygen saturation decreased with light exertion but stabilized at rest in the supine position, we changed the oxygen administration method to OxyMaskTM 8 L/min. Although oxygen saturation was maintained when the patient was at rest and in the supine position, it dropped to lower than 80% when the patient was in the sitting position after, for instance, moving to a portable toilet. Moreover, more than 30 min was needed for oxygen saturation to increase even after starting high-flow oxygen therapy with fraction of inspiratory oxygen 0.5 or higher. However, SpO2 rapidly recovered when the patient was repositioned to the supine position. We suspected the presence of a right–left shunt, which increased with change in position, and asked the cardiologist to search for a PFO or ASD. We performed contrast-enhanced CT and echocardiography but could not find an intracardiac shunt. Afterward, although her oxygen saturation slightly decreased when she sat or stood, her oxygenation slowly improved, and she was discharged under home oxygen therapy with 0.5 L/min via a nasal cannula 28 days after admission.