It Just Makes Sense*
The study took place at an academic medical center and enrolled 12 Pulmonary/Critical Care Medicine Fellows (PCCMF). They received training from Pulmonary/Critical Care Medicine faculty in order to perform limited transthoracic echocardiograms (LTTEs), which included assessment of right heart dysfunction via tricuspid annular plane systolic excursion, right ventricular (RV)/left ventricular ratios, and abnormal RV wall movement patterns. These approaches have been shown to reliably predict right heart dysfunction in POCUS with excellent specificity (4–6). Fellows independently performed LTTEs when pulmonary consults were requested for patients diagnosed with acute pulmonary embolism (PE) by CT-pulmonary angiogram. They noted their interpretations at the time of image acquisition, and then LTTEs were uploaded to a remote server and reviewed asynchronously by faculty experienced in POCUS. Findings indicating need for emergent intervention were reviewed with supervising faculty contemporaneously.
This retrospective study (3) included patients between the years of 2015 and 2017; 35% of 287 eligible patients (110) were included for final analysis. Patients were excluded if a PCCMF did not perform an LTTE, images obtained were of poor quality, a formal TTE (FTTE) was performed before an LTTE, a FTTE was not performed at all, or a FTTE was performed greater than 48 hours after PE diagnosis. FTTEs (interpreted by cardiologists certified in echocardiography) were considered the “gold standard” for indicating RV dysfunction. Two intensivists with greater than 5 years of POCUS experience were blinded to clinical data and outcomes; they later reviewed PCCMF images independently. PCCMFs were aware of biomarker values and CT-based evidence of RV dysfunction at the time they performed LTTEs.
Final analysis showed that blinded reviewers had a high level of agreement (κ = 0.88) and fellows reliably indicated the presence of RV dysfunction compared with FTTEs (NPV 83%). Furthermore, LTTEs appeared to be more sensitive and specific for RV dysfunction than biomarkers and were noninferior to CT-indicated RV dysfunction. In cases of clinical concern, a multidisciplinary PE response team (PERT) was activated to consider invasive interventions. During the 2-year study period, there were 18 PERT activations, 78% of which occurred as a direct result of LTTE interpretation.
The work by Filopei et al (3) demonstrates that PCCMFs can detect RV dysfunction in patients with PE, which led to earlier diagnosis than with FTTEs (median time difference 21 hr 18 min). Additionally, both PCCMFs and intensivist reviewers were able to accurately detect RV dysfunction despite lacking echocardiographic board certification. Interestingly, though this aligns with specialty-based consensus statements on POCUS for noncardiologists, RV assessment is often relegated to more advanced practitioner training (7, 8). The authors’ demonstration of PCCMF proficiency in RV analysis is unique, and it suggests that reliable RV evaluation should be part of basic POCUS curricula.
Despite the novelty of the study and its excellent design, the study does have limitations. Of their initial 287 eligible patients, only 154 patients had both types of TTE performed. A patient capture rate of 54% could suggest that PE management and PCCMF triage patterns at this hospital were relatively heterogeneous. Calculated PCCMF sensitivity was 71% (false-negative rate 11%), which some could claim is too low for a screening/triage modality.