Diagnostic Accuracy of Point-of-Care Ultrasound Performed by Pulmonary Critical Care Physicians for the Assessment of the Right Ventricle

SESSION TITLE: Imaging SESSION TYPE: Original Investigation Poster PRESENTED ON: Wednesday, October 26, 2016 at 01:30 PM - 02:30 PM PURPOSE: Intensivists perform point of care ultrasound (POCUS) to differentiate causes of shock and respiratory failure with a high degree of accuracy. In patients with pulmonary embolism (PE), rapid and accurate assessment of the right ventricle is essential. We assessed the accuracy, quality and rapidity of pulmonary and critical care (PCCM) fellow-performed goal-directed echocardiography (GDE) by comparing their interpretation to PCCM attending interpretations and formal cardiology-reviewed trans-thoracic echocardiograms (TTE). METHODS: We performed a single center retrospective review of PCCM fellow performed GDE's on patients with acute symptomatic PE. PCCM fellows attended a 3-day introductory course in POCUS during their first year of fellowship and had everyday access to both ultrasound equipment and expert faculty mentors. GDE was performed by PCCM fellows as part of their routine evaluation of patients with acute PE and their assessment of right ventricular size (RVS) and function (RVF) was documented as either normal or abnormal. Saved videos were subsequently reviewed by two ultrasound-trained PCCM attendings both with greater than 5 years of ultrasound experience. Interpretation of the GDE by the PCCM fellow and PCCM attending was compared to a gold standard TTE performed by an expert sonographer and analyzed by a board-certified cardiologist. Sensitivity, specificity, false positive rate and area under the curve (AUC) were used to assess diagnostic accuracy. Kappa was used to evaluate inter-rater reliability. Statistical analysis was performed using R Studio. RESULTS: Of 133 patients with acute PE, sixty patients with a GDE had a cardiology-reviewed TTE available for comparison. Forty-four GDEs were available for overread by a PCCM attending. Sensitivity, specificity, false positive rate, and AUC for RVS was (fellow, n=60) 77.3, 89.5, 10.5, 0.84, (attending 1, n=44) 94.1, 85.2, 5.8, 0.88, and (attending 2, n=44) 100, 92, 0, 0.95, and for RVF, (fellow) 61.1, 97.6, 8.2, and 0.88, (attending 1) 93, 86, 13.7, and 0.87, and (attending 2) was 93.3, 89.2, 8.0, and 0.89, respectively. PCCM attendings achieved the greatest AUC and lowest false-positive rate. The inter-rater reliability of PCCM attendings on adequate GDE's (N=44) was 0.954 (p=0.001) for RVS and 0.952 (p=0.001) for RVF. Three abnormal RV's seen by PCCM fellows in patients with suspected PE resulted in the ordering of a confirmatory CTPA. The average time difference between GDE and TTE was 25 hours and 31 minutes. CONCLUSIONS: Ultrasound trained PCCM attendings and fellows accurately assess normal and abnormal RV size and function in patients with an acute PE. PCCM attendings demonstrated high inter-rater reliability and were more accurate than PCCM fellows in assessing the RV. On average, GDE was performed approximately 1 day earlier than cardiology-reviewed TTE. CLINICAL IMPLICATIONS: A rapid evaluation of the RV should be performed in all patients with PE that could lead to more informed triage and treatment decisions. This study to our knowledge is the first to assess the diagnostic accuracy of PCCM attendings and fellow's evaluation of right ventricular size and function. More studies evaluating the use of POCUS for RV assessment in other disease processes should be conducted. DISCLOSURE: The following authors have nothing to disclose: Jason Filopei, Samuel Acquah, Eric Bondarsky, Navitha Ramesh, Paru Patrawalla No Product/Research Disclosure Information