INTRODUCTION: Dynamic changes of arterial waveform-derived variables (stroke volume variation (SVV), during positive pressure ventilation (PPV) have been shown to predict volume responsiveness in critically ill patients ; however, their use in diagnosing right ventricular (RV) failure is underappreciated.
CASE PRESENTATION: A 58 year old male was admitted with fever, chills, tachycardia and leukocytosis with left upper quadrant pain and chest wall pain after a fall. Chest pain was sharp and stabbing over the left chest and better when lying on the left side. Past medical history was significant for alcohol and heroin addiction. He takes no medications and has no allergies. At presentation temperature was 103.6F, heart rate 123, blood pressure 150/97. He had dyspnea with respiratory rate of 20 and oxygen saturation 96% on room air. Chest exam was normal, except tenderness to palpation on the left side. He had right upper quadrant tenderness as well as old and recent intravenous track marks. Laboratory data revealed a creatinine of 1.7, White blood cells were 14,500 with neutrophils of 12,000, platelets of 105,000 and total bilirubin of 5.1 with indirect 4.6. Admission ECG and chest radiograph were normal. Patient was admitted with a diagnosis of pneumonia, receiving antibiotics, but failed to improve over the next 3 days. He further deteriorated with worsening shortness of breath, tachycardia, diaphoresis, and increasing oxygen requirements. On day 3 he was cold, clammy, diaphoretic, tachycardic, tachypneic, hypoxemic, and hypotensive. He was intubated and transferred to medical ICU. Chest radiograph revealed large left lower lobe opacity with both effusion and consolidation. White blood cell count was 23,000 with 17,000 neutrophils, total bilirubin had remained unchanged and troponin 0.28. Two sets of blood cultures from admission had no growth. Impression was sepsis with septic shock secondary to pneumonia and parapneumonic effusion/empyema. Chest tube was placed evacuating approximately 1.3 liters of clear yellow fluid. A central venous catheter and radial arterial line were placed and despite large volume (6 liters crystalloid) resuscitation, and vasopressor support, patient remained hypotensive. Arterial pressure cardiac output monitoring revealed cardiac index (CI) of 1.2 and stroke volume index (SVI) of 10 (normal 35-50), SVV was elevated at 49% suggesting he remained on the preload dependent portion of the Frank- Starling curve. Central venous pressure (CVP) was elevated at 17. Because of elevated CVP along with high SVV and persistent low cardiac output that did not respond to volume, RV failure was suspected and a bedside echocardiographic evaluation revealed RV failure with high pulmonary pressures suggestive of pulmonary embolism. Pulmonary angiogram confirmed the diagnosis and after thrombectomy, hemodynamic parameters normalized (CI 2.8, SVI 35, SVV 12, and CVP 5) and vasopressors were discontinued.
DISCUSSION: The utility of dynamic variables in the intensive care unit has been well documented during PPV. Positive pressure ventilation increases intrathoracic pressures, decreases RV preload, and increases RV afterload, resulting in diminished RV and subsequently left ventricular (LV) stroke volumes. When the LV operates on the steep portion of the Starling curve, SVV is a sensitive indicator of preload dependence. One of the demonstrated limitations of SVV in predicting fluid responsiveness is pulmonary hypertension with severe RV dysfunction. Our case emphasizes the physiology behind this limitation, aiding the diagnosis of acute right ventricular failure from massive pulmonary embolism.
CONCLUSIONS: This case illustrates not only an atypical presentation of pulmonary embolism, but also the utility of variables obtained from arterial waveform analysis to diagnose right ventricular failure, which ultimately led to the correct diagnosis of pulmonary embolism. Specifically illustrated here is the differential diagnosis of elevated SVV that does not respond to volume resuscitation.
Reference #1 Marik PE, et al. Crit Care Med 2009;37:2642-7.
DISCLOSURE: The following authors have nothing to disclose: Bogdan Tiru, William McGee
No Product/Research Disclosure Information