Intra- and inter-rater reliability of the assessment of capacity for myoelectric control

Objective: To examine the reliability of the Assessment of Capacity for Myoelectric Control (ACMC) in children and adults with a myoelectric prosthetic hand. Design: Intra-rater and inter-rater reliability estimated from reported assessments by 3 different raters. Patients: A sample of convenience of 26 subjects (11 males, 15 females) with upper limb reduction deficiency or amputation and myoelectric prosthetic hands were video-taped during a regular clinical visit for ACMC. Participants' ages ranged from 2 to 40 years. Methods: After instruction, 3 occupational therapists with no, 10 weeks' and 15 years' clinical experience of myoelectric prosthesis training and follow-up independently rated the 30 ACMC items for each patient. The ratings were repeated after 2-4 weeks. Inter- and intra-rater reliability in items was examined by using weighted kappa statistics and Rasch-measurement analyses. Results: The mean intra-rater agreement in items was excellent (kappa 0.81) in the more experienced raters. Fit statistics showed too much variation in the least experienced rater, who also had only good (kappa 0.65) agreement in items. The stability of rater calibrations between first and second assessment showed that no rater varied beyond chance (> 0.50 logit) in severity. The mean inter-rater agreement in items was fair; kappa 0.60, between the experienced raters and kappa 0.47 between raters with no and 10 weeks' experience. Conclusion: Overall, the agreement was higher in the more experienced raters, indicating that reliable measures of the ACMC require clinical experience from myoelectric prosthesis training.