Detection of Pulmonary Embolism During Cardiac Arrest-Ultrasonographic Findings Should Be Interpreted With Caution.

OBJECTIVES: The aim of this study was to test the hypothesis that the right ventricle is more dilated during resuscitation from cardiac arrest caused by pulmonary embolism, compared with hypoxia and primary arrhythmia.

DESIGN: Twenty-four pigs were anesthetized and cardiac arrest was induced using three different methods. Pigs were resuscitated after 7 minutes of untreated cardiac arrest. Ultrasonographic images were obtained and the right ventricular diameter was measured.

SETTING: University hospital animal laboratory.

SUBJECTS: Female crossbred Landrace/Yorkshire/Duroc pigs (27-32 kg).

INTERVENTIONS: Pigs were randomly assigned to cardiac arrest induced by pulmonary embolism, hypoxia, or primary arrhythmia.

MEASUREMENTS AND MAIN RESULTS: There was no difference at baseline. During induction of cardiac arrest, the right ventricle dilated in all groups (p < 0.01 for all). The primary endpoint was right ventricle diameter at the third rhythm analysis: 32 mm (95% CI, 29-36) for pulmonary embolism which was significantly larger than both hypoxia: 23 mm (95% CI, 20-27) and primary arrhythmia: 25 mm (95% CI, 22-28)-the absolute difference was 7-9 mm. Physicians with basic training in focused cardiac ultrasonography were able to detect a difference in right ventricle diameter of approximately 10 mm with a sensitivity of 79% (95% CI, 64-94) and a specificity of 68% (95% CI, 56-80).

CONCLUSIONS: The right ventricle was more dilated during resuscitation when cardiac arrest was caused by pulmonary embolism compared with hypoxia and primary arrhythmia. However, the right ventricle was dilated, irrespective of the cause of arrest, and diagnostic accuracy by physicians with basic training in focused cardiac ultrasonography was modest. These findings challenge the paradigm that right ventricular dilatation on ultrasound during cardiopulmonary resuscitation is particularly associated with pulmonary embolism.