Objective: This observational study aimed to define the learning curve in goal-directed ultrasound (US) after a 2-day training course dedicated to novice emergency residents.
Materials and methods: After completion of the training program, 180 patients requiring goal-directed US examination were examined by a resident and by an experienced investigator. The main endpoints were the diagnostic agreement between the two operators for 14 clinical questions, the duration of the examinations, the number of nonaddressed questions, and the final diagnosis. All criteria were analyzed according to the experience of the resident every 10 examinations.
Results: After 30 supervised examinations, residents adequately assessed with a very good or considerable agreement global left ventricular systolic dysfunction [κ=0.92; 95% confidence interval (CI): 0.80-1], severe right ventricular dilation (κ=0.73; 95% CI: 0.37-1), inferior vena cava diameter (κ=0.88; 95% CI: 0.71-1), and pericardial effusion (κ=0.85; 95% CI: 0.55-1). In general US, 20 supervised examinations were required to diagnose intraperitoneal effusion (κ=0.81; 95% CI: 0.61-1), cholelithiasis (κ=0.73; 95% CI: 0.36-1), obstructive uropathy (κ=0.85; 95% CI: 0.56-1), bladder distention (κ=1; 95% CI: 1-1), abdominal aortic aneurism (κ=0.9; 95% CI: 0.74-1), alveolar interstitial pattern (κ=0.87; 95% CI: 0.74-0.99), consolidated lung (κ=0.83; 95% CI: 0.68-0.97), or pleural effusion (κ=0.89; 95% CI: 0.77-1). After 30 supervised examinations, the overall diagnostic accuracy was judged excellent between the two investigators, with a significant improvement during the learning curve.
Conclusion: The performance of 30 supervised and goal-oriented examinations appeared adapted to adequately answer clinical questions covered by core applications of emergency US.