Adoptability and Accuracy of Point-of-Care Ultrasound in Screening for
Valvular Heart Disease in the Primary Care setting
Abstract
Background: Despite continued efforts, the majority of patients with
valvular heart disease (VHD) remain undiagnosed and untreated. This
study aimed to assess the adoptability and accuracy of point-of-care
handheld echocardiographic assessments (POCE) in the primary care
setting. Methods: Eleven previously untrained primary care providers
were trained to use the Vscan Extend (GE, WI) POCE to assess VHD, left
ventricular function (LVEF), and major extra-cardiac findings. Their
assessments were compared to those of 3 blinded expert readers. A total
of 180 assessments were evaluated using Kappa statistics (κ) together
with their estimated standard error, p-value, and 95% CI bounds.
Results: Identical or nearly identical agreement between previously
untrained primary providers and expert readers was evident for the
diagnosis of tricuspid regurgitation, mitral regurgitation, pericardial
effusion, volume status. These agreements were strongest in apical long
axis (κ =1, p<0.001) and parasternal long and short axis views
(k>=0.82 p<0.001), though agreement remained
robust in apical 4-chamber views (k>=0.76). The agreements
in LVEF assessment were identical in the apical long axis view (κ =1,
p<0.001) and robust in the remaining 3 views
(κ>=0.66, p<0.001). The assessments of aortic
stenosis (parasternal/long, κ =0.42, and parasternal/short, κ =0.47,
both p<0.001) were weak in their agreement. Conclusion:
Compared to expert echocardiography readers, the untrained providers’
use of POCE for VHD shows high user adoptability and diagnostic
accuracies in the primary care setting.