Analysis of patient and proxy ratings on the Dysexecutive Questionnaire: an application of Rasch analysis

Background: We provide an alternate method of analysing self-report and proxy-report data on subjective complaints of dysexecutive symptoms among a group of patients with traumatic brain injury. Objective: The purpose of this study was to examine differences in the ratings of patients and proxies on a measure of the dysexecutive syndrome and further explore the insight impairment problem in patients with traumatic brain injury. Methods: Rasch analysis was conducted on the ratings of the Dysexecutive Questionnaire (DEX) by a sample of patients and their proxies. Results: While the average scores based on patient and proxy ratings were approximately the same (mean patient raw score = 30.12 and mean proxy raw score = 31.32), differential item functioning was found in five DEX items. As a result, the relationship between measures obtained from patient and proxy ratings was only in the moderate range (intraclass correlation = 0.46). Conclusions: Identification of differential item functioning in five of the 20 DEX items reflected the different perspectives of patients and their proxies in reporting the frequency of dysexecutive behaviour and suggests that these ratings are not interchangeable.