Development and initial validation of a computerized adaptive test prototype for ECG interpretation using item response theory

Aims Accurate electrocardiogram (ECG) interpretation is essential for appropriate clinical decision-making, yet competency remains suboptimal among many healthcare professionals. Although computerized adaptive testing (CAT) has the potential to provide rapid, individualized, and standardized assessment of ECG interpretation skills, its implementation requires a well-calibrated item bank based on item response theory. This study aimed to establish item parameters, evaluate the latent structure underlying ECG interpretation competency using multidimensional item response theory (MIRT), and develop an initial CAT prototype for ECG competency assessment. Methods and results We analyzed responses from 535 healthcare professionals who completed a 50-item online ECG test previously developed and administered via an online platform. Participants represented diverse clinical roles and experience levels. Dimensionality was examined using exploratory and confirmatory MIRT analyses. A unidimensional model and a two-factor model both emerged as plausible latent structures. Item parameters, including discrimination and difficulty, were estimated from item characteristic curves. Of 50 items, 47 demonstrated strong discrimination, and difficulty parameters were appropriately distributed. Model fit analyses demonstrated that the selected model adequately represented the response data and provided a suitable basis for CAT development. Conclusions Although both unidimensional and two-factor models demonstrated acceptable statistical performance, the unidimensional model was selected as the initial operational framework because of its simplicity and suitability for computerized adaptive testing. The calibrated ECG item bank enabled development of a preliminary CAT prototype. These findings support the feasibility of CAT as an efficient and standardized approach for assessing ECG interpretation competency across healthcare professionals. Further studies are warranted to validate its performance and educational utility in broader clinical settings.