Qualitative ultrasound training: defining the learning curve

To assess the developing competence of two trainees learning groin and shoulder ultrasound when compared to an expert practitioner.

Specific pro formas were used to record ultrasound trainee performance in each scan region and their diagnosis was compared to the opinion of the expert. The data derived were reviewed using kappa analysis and training end points were defined as a minimum 80% trainee agreement with the expert. Retrospectively, cumulative sum analysis was applied to the data to assess case-by-case performance.

For groin hernias, reporting an average of 70 examinations was required to become competent and inguinal hernias required higher numbers of examinations than femoral hernias. For shoulders, an average of 80 examinations was required and the supraspinatus and infraspinatus tendons proved the most challenging structures.

Kappa analysis demonstrated a differential in the learning curves for individual structures within each examination region. Sequential kappa scores are consistent with a sigmoid learning curve. The numbers required to achieve satisfactory agreement are suggested as required minima for ultrasound training curricula. Cumulative sum analysis provided a sensitive indicator of trainee performance, quickly highlighting individual learning difficulties when they arose. Its prospective use can ensure extra training support is instigated quickly and appropriately.