ABSTRACT
We aimed to find the depression rating scale with the greatest accuracy when applied by psychiatrists in Iraqi Kurdistan. We recruited 200 patients with primary depression and 200 controls living in the Kurdistan region of Iraq. The Mini International Neuropsychiatry Inventory (MINI) was used as a gold standard for DSM-IV depression. We also used: the two-item and the nine-item versions of the Patient Health Questionnaire (PHQ2, PHQ9), the Hospital Anxiety and Depression Scale (HADS), the Calgary Depression Scale for Schizophrenia (CDSS) and the Centre for Epidemiological Studies Depression (CES-D) scale. Interviews were performed by psychiatrists who also rated their clinical judgement using the Clinical Global Impression (CGI) scale and other mental health practitioners. All scales and tools performed with high accuracy and reliability. The least accurate tool was the PHQ2; however, with only two items it was efficient. Sensitivity and specificity for all tools were above 90%. Clinicians using the CGI were accurate in their clinical judgement. The CDSS appeared to be the most accurate scale for DSM-IV major depression and the PHQ2 the most efficient. However, only the CDSS appeared to offer an advantage over psychiatrists' judgement.
ABSTRACT
OBJECTIVE: We aimed to redefine the optimal self-report symptoms of depression suitable for creation of an item bank that could be used in computer adaptive testing or to develop a simplified screening tool for DSM-V. METHOD: Four hundred subjects (200 patients with primary depression and 200 non-depressed subjects), living in Iraqi Kurdistan were interviewed. The Mini International Neuropsychiatric Interview (MINI) was used to define the presence of major depression (DSM-IV criteria). We examined symptoms of depression using four well-known scales delivered in Kurdish. The Partial Credit Model was applied to each instrument. Common-item equating was subsequently used to create an item bank and differential item functioning (DIF) explored for known subgroups. RESULTS: A symptom level Rasch analysis reduced the original 45 items to 24 items of the original after the exclusion of 21 misfitting items. A further six items (CESD13 and CESD17, HADS-D4, HADS-D5 and HADS-D7, and CDSS3 and CDSS4) were removed due to misfit as the items were added together to form the item bank, and two items were subsequently removed following the DIF analysis by diagnosis (CESD20 and CDSS9, both of which were harder to endorse for women). Therefore the remaining optimal item bank consisted of 17 items and produced an area under the curve (AUC) of 0.987. Using a bank restricted to the optimal nine items revealed only minor loss of accuracy (AUC = 0.989, sensitivity 96%, specificity 95%). Finally, when restricted to only four items accuracy was still high (AUC was still 0.976; sensitivity 93%, specificity 96%). CONCLUSIONS: An item bank of 17 items may be useful in computer adaptive testing and nine or even four items may be used to develop a simplified screening tool for DSM-V major depressive disorder (MDD). Further examination of this item bank should be conducted in different cultural settings.
