Point-of-Care Ultrasound (POCUS) as an Extension of the Physical Examination in Patients with Bacteremia or Candidemia

Abstract

Background: In general, transthoracic echocardiography (TTE) is the first diagnostic test used for patients with bacteremia or candidemia and clinical signs of Infective Endocarditis (IE). Point-of-care ultrasound (POCUS) may be used in addition to physical examination for the detection of structural heart disease and valve abnormalities. Objective: To determine the diagnostic accuracy of POCUS for the detection of signs suggestive of IE, including vegetation, valvular regurgitation, structural heart disease, hepatomegaly, splenomegaly and septic embolisms, in patients with bacteremia or candidemia. Design: Observational, cross-sectional, multicenter study using convenience sampling. Setting: Six Spanish academic hospitals. Patients: Adult patients with bacteremia or candidemia between 1 February 2018 and 31 December 2020. Measurements: The reference test, to evaluate vegetation, valvular regurgitation and structural heart disease, was transesophageal echocardiography (TEE). For patients who did not undergo TEE, transthoracic echocardiography (TTE) was considered the reference test. POCUS was performed by internists, while conventional echocardiography procedures were performed by cardiologists. Results: In 258 patients, for the detection of valvular vegetation, POCUS had sensitivity, specificity, and positive and negative predictive values of 77%, 94%, 82% and 92%, respectively. For valvular regurgitation (more than mild), sensitivity was ≥76% and specificity ≥85%. Sensitivity values for the detection of hepatomegaly and splenomegaly were 92% and 92%, respectively, while those for specificity were 96% and 98%. Conclusion: POCUS could be a valuable tool, as a complement to physical examination, at the hospital bedside for patients with bacteremia or candidemia, helping to identify signs suggestive of IE.

Unexpected Causes and Complications of ST-Segment Myocardial Infarction That Highlight the Importance and Limitations of Point-of-Care Ultrasound in the Emergency Department

“…When used to assess for evidence of endocarditis in a high-risk population, POCTE has reasonable diagnostic test accuracy. In a multi-center, cross-sectional study of 258 adults with bacteremia or candidemia, POCTE performed by internists had 77% sensitivity and 94% specificity to detect valvular vegetations against the reference standard of TEE performed by cardiologists [ 16 ]. However, in this patient, the preceding symptoms of malaise and generalized illness were not fully characterized until admission to the intensive care unit, and endocarditis was not otherwise entertained upon initial presentation to the ED with STEMI and cardiogenic shock.…”

Comparative Eminence: Foundation versus Domain-Specific Model for Cardiac Ultrasound Segmentation

“…In contrast, the comparable performance of SAM on TTE and POCUS images demonstrated the advantage of leveraging the generalization capabilities of foundation models 21 in building AI solutions for the rapidly growing use of POCUS in cardiac imaging 35 . While evaluation on a larger POCUS is required to better evaluate inter-observer variations, we demonstrated a strategy to fine-tune foundation models using readily available and relatively high-quality TTE results knowing that POCUS images usually come with larger variations in operator skill levels, image quality, and scanning modalities [35][36][37][38] . We observed that fine-tuned SAM, similar to MedSAM 24 , can accurately segment cases with weak or missing boundaries, which is more common in POCUS images.…”

Multiorgan Point-of-Care Ultrasound Assessment in Critically Ill Adults

“…Here POCUS should be viewed as a rule-in rather than a rule-out test, with a specificity of 94% in one study, but notably low sensitivity. 35 The utility of lung ultrasound for the diagnosis of pneumonia has been well established, with excellent sensitivity and specificity. 11 Findings supportive of pneumonia include heavily consolidated ("hepatized") lung, subpleural consolidation, dynamic air bronchograms, and the shred sign 11 (Figures 5 and 6, Videos 5 and 6).…”

“…Here POCUS should be viewed as a rule-in rather than a rule-out test, with a specificity of 94% in one study, but notably low sensitivity. 35 …”