Abstract
BACKGROUND: Rapid identification of the cause of hemodynamic instability is essential in the cardiovascular intensive care unit (CV-ICU), and transesophageal echocardiography provides superior diagnostic image quality compared with transthoracic echocardiography in this setting. However, point-of-care transesophageal echocardiography (POCUS-TEE) is not routinely used, partly because of perceived training barriers. We therefore sought to summarize the etiologies of hemodynamic instability in the CV-ICU and to identify a pragmatic minimal set of TEE views required for their diagnosis. METHODS: This single-center retrospective study evaluated consecutive adult cardiovascular ICU admissions over five years. All POCUS-TEE examinations were performed as part of routine clinical care at the discretion of the treating team. Examinations in patients with hemodynamic instability were analyzed to identify the "first informative view" supporting the diagnosis. A cumulative coverage analysis was performed to derive a minimal-view set. Inter-rater reproducibility was assessed for view acquisition and management decisions. RESULTS: Among 6,898 admissions, POCUS-TEE was performed in 353 examinations (5.1%), predominantly in patients with higher severity scores compared to those without TEE. In preoperative stable patients (n = 37), POCUS-TEE altered planned surgery in 6 patients (16.2%). In postoperative stable patients (n = 101), management changes included optimization of mechanical support and anticoagulation. In postoperative-unstable examinations (n = 238), 79.4% prompted procedural interventions. A derived minimal three-view set (mid-esophageal bicaval, mid-esophageal four-chamber, and transgastric mid-short-axis) achieved 92.4% diagnostic coverage in unstable examinations. While the specific view acting as the first diagnostic window varied due to acquisition sequence (κ 0.06-0.11), the core three-view set was highly reproducible across raters (Jaccard similarity coefficient 0.60), and agreement on management decisions was moderate. No procedure-related complications were observed during bedside POCUS-TEE. CONCLUSIONS: POCUS-TEE frequently drives urgent interventions in postoperative hemodynamic instability. Although the sequence of diagnostic capture is operator-dependent, a simple three-view protocol provides > 90% diagnostic coverage. This derivation supports a physiology-based, limited-view approach for training, early competency development, and resuscitation in cardiovascular critical care.