Background & Aim   A basic concern regarding the use of portable point-of-care echocardiography by general practitioners (GPs) has been the lack of data demonstrating that it would yield clinically relevant information on patients’ outcomes. In the frame of the cluster-randomized 4-arm Handheld-BNP study 44 GPs underwent standardized, quality-controlled training enabling them to perform and interpret simplified two-dimensional and Color-Doppler echocardiography (ECHO). This secondary analysis aimed to assess GPs’ ECHO findings in diagnostically naïve patients presenting in primary care with signs and symptoms suggestive of heart failure (HF), and investigate their prognostic significance during long-term follow-up.

Methods In two of four Handheld-BNP study arms GPs performed ECHO with a portable device (OptiGo^®, Philips Medizinsysteme Ultraschall, 2.5 MHz phased-array transducer, B- and Color-Doppler-mode, no spectral Doppler). From imaging the left ventricular (LV) parasternal long and short axis views and the apical four chamber view they reported in a semi-quantitative fashion LV dimensions (normal, dilated), LV wall thickness (LVWT, normal, increased), LV ejection fraction (LVEF, normal, reduced), left atrial dimension (LA, normal, dilated) and mitral regurgitation (MR, none or small, severe). In one of the study arms, GPs also measured B-type natriuretic peptide (BNP). Based on patients’ clinical evaluation and information derived from ECHO (and BNP in 50% of the patients), GPs decided about the presence of HF (yes or no). Survival status was assessed for the following 72.2 (68.5 – 72.2) months. Events were censored at 72 months.

Results: A total of 425 patients were included (68±11years, 61% female). Using ECHO, GPs diagnosed no abnormality in 194 (46%) patients (group 1), 1 abnormality in 92 (22%, group 2), and >1 abnormality in 134 (32%) (group 3).  LVEF was described as reduced in 87 (21%) patients, LV dimension as dilated in 145 (35%), LVWT as increased in 143 (34%), and LA as dilated in 61 (15%). Severe MR was found in 3 patients (1%). GPs diagnosed HF in 177 (42%) patients. Over 72 months, a total of 64 patients (15%) died, 14 (7.5%) in group 1, 15 (16.3%) in group 2, and 35 (26.1%) in group 3, respectively (Logrank p<0.001)). Compared to group 1 (reference) HR in group 2 was 2.36 [1.14-4.89] (p=0.021), and in group 3 HR was 4.04 [2.17-7.51] (p<0.001), respectively (Figure A). The HR for GP diagnosis of HF was 3.67 [2.15-6.27] (<0.001). Age-adjusted HRs regarding all-cause mortality of groups 2 and 3 were 1.84 [0.88-3.82] (p=0.10) and 2.58 [1.36-4.86] (p=0.004), respectively. Age-adjusted HR for the GPs’ diagnosis of HF was 2.49 [1.44 -4.30] (p=0.001).

Conclusion In diagnostically naïve patients presenting in primary care with signs and symptoms suggestive of HF, overall all-cause mortality during 72 months was high, but particularly increased if GPs detected ECHO abnormalities at baseline. Patients with a GP diagnosis of HF had also a significantly higher all-cause mortality risk. Simplified semi-quantitative ECHO may support GPs to identify patients at increased mortality risk and help to triage patients’ for further diagnosis and intensified therapy. 

Figure: Survival according to A) number of abnormal findings in hand-carried echocardiography and B) heart failure diagnosis by General Practitioner. Time = months