Back to all papers

Comparative Diagnostic Accuracy of Point-of-Care Ultrasound for Detecting Reduced Left Ventricular Systolic Function: A Systematic Review and Meta-Analysis.

November 17, 2026pubmed logopapers

Authors

Khan MA,Shellenberger RA,Ross IR,Patel YB,Bellino AC,El Nayir M,Nwaezeapu KI,Daniel E,Zamarripa P,Wang S,Moon WJ

Affiliations (5)

  • Trinity Health Ann Arbor Hospital, Department of Internal Medicine, Ann Arbor, MI, USA.
  • Washington University School of Medicine in St. Louis, Department of Medicine, St. Louis, MO, USA.
  • University of Illinois College of Medicine at Peoria, OSF Saint Francis Medical Center, Department of Cardiology, Peoria, IL, USA.
  • Michigan State University, East Lansing, MI, USA.
  • Saint Louis University School of Medicine, Department of Internal Medicine, St. Louis, MO, USA. [email protected].

Abstract

To assess the diagnostic accuracy of point-of-care ultrasound (POCUS) for detecting reduced left ventricular systolic function (LVSF) using visual qualitative assessment (VQA), E-point septal separation (EPSS), and artificial intelligence (AI)-assisted assessment, compared with transthoracic echocardiography (TTE). We searched PubMed, Ovid MEDLINE, Embase, and the Cochrane Library from January 1, 1985, through July 8, 2025. Eligible studies compared POCUS-assessed LVSF with left ventricular ejection fraction measured by TTE (or cardiac MRI in one study). Data were pooled using random-effects meta-analysis to determine the diagnostic odds ratio (DOR) and likelihood ratios (LR) for each of the three methods. Twenty-eight studies, including 4,647 participants, met inclusion criteria. Nine studies evaluating VQA yielded a DOR of 40.53 (95% CI, 15.07 to 109.00; P < 0.001) and a positive LR (LR + ) of 6.90 (95% CI, 3.75 to 12.69). Seven studies evaluating EPSS resulted in a DOR of 21.77 (95% CI, 10.25 to 46.28; P < 0.001) and a LR + of 5.70 (95% CI, 2.89 to 11.22). Five studies evaluating AI-assisted POCUS resulted in a DOR of 47.35 (95% CI, 16.87 to 132.93; P < 0.001) and a LR + of 8.37 (95% CI, 4.40 to 15.93). Meta-regression analyses revealed no statistically significant differences in diagnostic performance among the three methods. Subgroup analysis by operator experience level also showed no statistically significant difference between novice and experienced operators for each method. VQA, EPSS, and AI-assisted POCUS all demonstrate moderate-to-high diagnostic accuracy for identifying reduced LVSF, with no significant differences among methods on meta-regression or by operator experience level. Moderately strong LR + across all three methods supports the use of POCUS for ruling in reduced LVSF. EPSS cutoffs at the lower and upper ends of the reported range showed higher sensitivity and specificity, respectively, suggesting utility as a rule-in or rule-out tool where formal echocardiography is not readily available. PROSPERO CRD42024626542.

Topics

Journal ArticleSystematic Review

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAISlice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.