Evaluating the precision of computer adaptive testing in longitudinal hand surgery analyses: A psychometric approach

Patient-reported outcome measures (PROMs) are essential in hand surgery for assessing patient health but it can be time-consuming. Computerized adaptive testing (CAT) offers a more efficient alternative by reducing the number of questions asked. This study sourced the data of 268 patients undergoing cubital tunnel release from the UK Hand Registry to evaluate whether CAT's inherent imprecision affects longitudinal research conclusions. Mean patient evaluation measure (PEM) scores at baseline, 2 months and 6 months from the traditional full-length assessment (10 questions) were compared with the simulated scores assuming that the CAT version (median of 2 questions) was used. Both methods showed significant improvements in PEM scores post-surgery (p < 0.01), with minimal differences between the mean scores and overlapping confidence intervals. These findings confirm that CAT replicates full-length PROM results while significantly reducing patient burden, thereby supporting its use in clinical and research settings for hand surgery.