Impact of pocket ultrasound use by internal medicine housestaff in the diagnosis of dyspnea.

Abstract

Recent reports demonstrate high diagnostic accuracy of lung ultrasound for evaluation of dyspnea. We assessed the feasibility of training internal medicine residents in lung ultrasound with a pocket ultrasound device. We performed a prospective, observational trial of residents performing lung ultrasound with a pocket ultrasound. Training consisted of two 90-minute sessions of didactics and supervised bedside performance. Two residents received an additional 2 weeks of training. Residents recorded a clinical diagnosis based on admission data. Following lung ultrasound performance, an ultrasound diagnosis was recorded integrating clinical and sonographic findings. Using receiver operating curve analysis, the area under the curve was calculated for both clinical diagnosis and ultrasound diagnosis using attending physician's final discharge diagnosis as the gold standard. Five residents performed 69 exams. The AUC for ultrasound diagnosis was significantly higher than that for clinical diagnosis (0.87 vs 0.81, P < 0.01). AUCs increased using lung ultrasound for diagnoses as follows: chronic obstructive pulmonary disease (0.73-0.85, P = 0.06), acute pulmonary edema (0.85-0.89, P = 0.49), pneumonia (0.77-0.88, P = 0.01), and pleural effusions (0.76-0.96, P < 0.002). Lung ultrasound performed by residents with a pocket ultrasound improved the diagnostic accuracy of dyspnea. Two residents undergoing extended training showed a total increase in diagnostic accuracy.