Agreement between ‘‘echoscopy’’ performed by non-cardiologist physicians and conventional cardiac ultrasound

Introduction: Focused cardiac ultrasound has been proposed as a useful approach for improving clinical decision making, as well as to be able to rapidly identify the ultrasound signs of a specific list of potential diagnoses. Objective: To evaluate a training program for physicians with no experience in cardiac ultrasound with the aim performing focused cardiac ultrasound using a portable device (echoscopy). Materials and methods: The results obtained from echoscopy performed by the physicians that received training were compared with those obtained with conventional cardiac ultrasound carried out by expert cardiologists. A total of 5 non-cardiologist doctors, including 1 medical student, 2 Internal Medicine residents, and 2 from Intensive Medicine, took part in a four-week training course given by a Level III Cardiology specialist. The course included: First week: Theory and basis of cardiac ultrasound (3 hours daily) Second week: Theory of acquiring images. Normal and abnormal findings (50 studies). Third week: handling of the echoscope (50 studies). Fourth week: Data collection. The study included patients scheduled for conventional cardiac ultrasound in the Non-Invasive Methods Laboratory. Two examinations were carried out on each patient. The first consisted of an echoscopy performed by a doctor that had received the training, and the second consisted of a cardiac ultrasound carried out by an expert cardiologist. The ultrasound parameters evaluated were: left ventricular ejection fraction, right ventricular dysfunction, left atrial enlargement, pulmonary hypertension, cardiac valve disease, and pericardial effusion. The results found in echoscopy versus cardiac ultrasound were compared using concordance analysis (Kappa Index). Results: The following results were obtained on the 221 studies performed: moderate agreement in left ventricular ejection fraction (κ = 0.541, p < .000), right ventricular function (κ = 0.403, p < .001), left atrial enlargement (κ = 0.413, p< .001), mitral valve and tricuspid valve disease (κ = 0.437, p< .001 and (κ = 0.466, P< .001, respectively). There was weak agreement with aortic valve disease. Pericardiac effusion and the presence of pulmonary hypertension had a poor and week agreement, respectively. Conclusions: With a limited training period, the participants with no previous experience in ultrasound techniques and using echoscopy achieved a moderate agreement in the majority of measurements when compared with conventional cardiac ultrasound performed by experts in the technique. A study with a larger number of participants is required in order to determine the ideal training period to obtain results comparable with cardiac ultrasound.