Hand-held cardiac ultrasound examinations performed in primary care patients by nonexperts to identify reduced ejection fraction.

ABSTRACT

BACKGROUND: Early identification of patients with reduced left ventricular ejection fraction (LVEF) could facilitate the care of patients with suspected heart failure (HF). We examined if (1) focused cardiac ultrasound (FCU) performed with a hand-held device (Vscan 1.2) could identify patients with LVEF < 50%, and (2) the distribution of HF types among patients with suspected HF seen at primary care clinics. METHODS: FCU performed by general practitioners (GPs)/GP registrars after a training programme comprising 20 supervised FCU examinations were compared with the corresponding results from conventional cardiac ultrasound by specialists. The agreement between groups of estimated LVEF < 50%, after visual assessment of global left ventricular function, was compared. Types of HF were determined according to the outcomes from the reference examinations and serum levels of natriuretic peptides (NT-proBNP). RESULTS: One hundred patients were examined by FCU that was performed by 1-4 independent examiners as well as by the reference method, contributing to 140 examinations (false positive rate, 19.0%; false negative rate, 52.6%; sensitivity, 47.4% [95% confidence interval [CI] 27.3-68.3]; specificity, 81.0% [95% CI 73.1-87.0]; Cohen's κ measure for agreement = 0.22 [95% CI 0.03-0.40]). Among patients with false negative examinations, 1/7 had HF with LVEF < 40%, while the others had HF with LVEF 40-49% or did not meet the full criteria for HF. In patients with NT-proBNP > 125 ng/L and fulfilling the criteria for HF (68/94), HF with preserved LVEF (≥50%) predominated, followed by mid-range (40-49%) or reduced LVEF (< 40%) HF types (53.2, 11.7 and 7.4%, respectively). CONCLUSIONS: There was poor agreement between expert examiners using standard ultrasound equipment and non-experts using a handheld ultrasound device to identify patients with reduced LVEF. Asides from possible shortcomings of the training programme, the poor performance of non-experts could be explained by their limited experience in identifying left ventricular dysfunction because of the low percentage of patients with HF and reduced ejection fraction seen in the primary care setting. TRIAL REGISTRATION The study was registered at ClinicalTrials.gov (NCT02939157). Registered 19 October 2016.