Test–retest reliability of the Danish Adult Reading Test in patients with comorbid psychosis and cannabis-use disorder

Abstract

BACKGROUND: The New Adult Reading Test is a common instrument for assessing pre-morbid IQ for patients with, for instance, schizophrenia. However, test-retest reliability has not been established for patients dually diagnosed with psychosis and substance use disorder. Furthermore, test-retest reliability of the Danish adaptation has never been established in any population. AIMS: To determine the test-retest reliability of the Danish Adult Reading Test (DART) (adapted from the National Adult Reading Test, NART) for patients dually diagnosed with psychosis and cannabis-use disorder. METHODS: This was a secondary analysis of the CapOpus randomized trial. As part of the trial, 103 patients were randomized, and completed the DART up to three times. Pearson's r and pairwise t-tests were calculated. RESULTS: DART score was independent of randomization, cannabis-use frequency and psychopathology. Scores at the last interview were slightly higher than at the first two. Correlation over time was very strong (0.8 < r <0.9) for all pairwise comparisons of interviews. Variations in DART scores and estimated pre-morbid IQ over time were sometimes of borderline statistical significance but not of clinical relevance. CONCLUSIONS: DART and NART have high test-retest reliability, but apparently non-systematic, clinically irrelevant variation over time does occur. CLINICAL IMPLICATION: The Danish adaptation of the New Adult Reading Test possesses good test-retest reliability, making it an appropriate choice for assessment of pre-morbid IQ, and in patients with dually diagnosed psychosis and cannabis-use disorder.