Validity and reliability of the Turkish version of the adult ADHD self-report screening scale for DSM-5

Abstract

Background: Approximately half of the children with Attention-deficit/hyperactivity disorder (ADHD) continue to meet diagnostic criteria in adulthood. The prevalence of adult ADHD is reported between 2.5% and 4.4% and is associated with significant impairment in quality of life and increased psychiatric comorbidity. ADHD in adults remains mostly undiagnosed and/or untreated despite the availability of effective treatments. Majority of people who do not receive necessary treatment are in the non-clinical, or non-psychiatric clinical population. Screening is an important step for diagnosing adults with ADHD. Yet there are no valid and reliable screening questionnaires calibrated for DSM-5 in Turkish. Aims: we aimed to test the reliability and the validity of the ASRS-5 screening questionnaire designed according to DSM-5 in the Turkish population. Study design: methodological and cross-sectional study Methods: The translation was carried on according to the WHO-CIDI translation guide using a linguistic adaptation approach. We used a convenience sampling method to recruit an adult ADHD group (n=68) and a control group (n=68). The participants completed a sociodemographic form, six items ASRS-5 and a previous version 18 items ASRS v1.1 for the concurrent validity analysis. For the diagnostic validity clinical diagnosis made by psychiatrists according to DSM-5 criteria was used. Internal consistency and item-total correlation coefficients, exploratory factor analyses, correlation with ASRS v1.1 and ROC curve analysis were conducted. Results: the internal consistency measured by Cronbach alpha was 0.869. Item-total correlation coefficients were calculated to be between 0.602 and 0.717 and the correlations were statistically significant (p<0.0001). The ASRS-5 showed to have a unidimensional factor structure explaining the 60.54% of the variance. The correlation between ASRS-5 and ASRS v1.1 total score was calculated as 0.992 (p<0.0001), between ASRS-5 and ASRS v1.1 attention deficit sub-dimension as 0.868 (p<0.0001). In the ROC analysis of ASRS-5, the area under the curve was found to be 0.916. The cut-off score was calculated as 9/10 with a sensitivity of 85.2% and specificity as 89.7%. Conclusion: our data suggests that ASRS-5 is a valid and reliable self-report measure to assess and screen ADHD in Turkish population. It may be useful for both clinical and population studies.