Reliability and validity of the Dutch version of the Glasgow Anxiety Scale for people with an Intellectual Disability (GAS-ID)

Background In the Netherlands, no self-report screening questionnaire for anxiety in people with intellectual disabilities (ID) was available yet. Therefore, we have translated the Glasgow Anxiety Scale for people with an Intellectual Disability (GAS-ID) into Dutch and studied its reliability and validity in adults with borderline, mild or moderate ID. Method Test-retest reliability was studied in 66 participants, convergent validity against the Anxiety sub-scale of the Hospital, Anxiety and Depression Scale (HADS-A) in 96, and criterion validity against psychiatric diagnosis in 195 participants. Results Internal consistency was =0.86 and test-retest reliability ICC=0.89 (95% CI: 0.82-0.93). Correlation with the HADS-A was r=0.61 (95% CI: 0.47-0.72); sensitivity was 83.9% (95% CI: 72.2-91.2) and specificity was 51.8% (95% CI: 43.6-59.9) using a cut-off score of 17. Missed diagnoses (false-negatives) were mostly specific phobias. Of the false-positives, 38 of 66 participants (58%) had another psychiatric diagnosis. Conclusions The Dutch version of the GAS-ID is a reliable screening instrument with satisfactory sensitivity, but moderate specificity for anxiety disorders. Although specificity for anxiety disorders is only moderate, high scores appear to be indicative of other psychiatric problems too, justifying referral for psychiatric assessment of false-positives.