Cameron, 2013 Psychometric properties of the Quick Inventory of Depressive Symptomatology (QIDS-SR) in UK primary care.

[Altostrata]

Used in the infamous STAR*D study, the QIDS-SR test for depression has no validity, this study found. All the data and conclusions of the STAR*D study are therefore suspect.

 

J Psychiatr Res. 2013 May;47(5):592-8. doi: 10.1016/j.jpsychires.2013.01.019. Epub 2013 Feb 16.
Psychometric properties of the Quick Inventory of Depressive Symptomatology (QIDS-SR) in UK primary care.
Cameron IM, Crawford JR, Cardy AH, du Toit SW, Lawton K, Hay S, Mitchell K, Sharma S, Shivaprasad S, Winning S, Reid IC.

Source

Applied Medical Sciences (Psychiatry), University of Aberdeen, Clinical Research Centre, Royal Cornhill Hospital, Aberdeen, UK. [e-mail]

 

Abstract at http://www.ncbi.nlm.nih.gov/pubmed/23419617

It is widely believed that severity of depressive disorder should guide treatment selection and many guidelines emphasise this factor. The Quick Inventory of Depressive Symptomatology (QID-SR16) is a self-complete measure of depression severity which includes all DSM-IV criterion symptoms for major depressive disorder. The object of this study was to assess the psychometric properties of the QIDS-SR16 in a primary care sample. Adult primary care patients completed the QIDS-SR16 and were assessed by a psychiatrist (blind to QIDS-SR16) with the 17-item Hamilton Rating Scale for Depression (GRID-HAMD). Internal consistency, homogeneity and convergent and discriminant validity of the QIDS-SR16 were assessed. Severity cut-off scores for QIDS-SR16 were assessed for convergence with HRSD-17 cut-offs. Published methods for converting scores to HRSD-17 were also assessed. Two hundred and eighty-six patients participated: mean age = 49.5 (s.d. = 13.8), 68% female, mean HRSD-17 = 12.6 (s.d. = 7.6). The QIDS-SR16 exhibited acceptable internal consistency (Cronbach's alpha = 0.86), a robust factor structure indicating one underlying dimension and correlated highly with the HRSD-17 (r = 0.79) but differed significantly in how it categorised the severity of depression relative to the HRSD-17 (Wilcoxon Signed Rank Test p < 0.001). Using published methods to convert QIDS-SR16 scores to HRSD-17 scores did not result in alignment of severity categorisation. In conclusion, psychometric properties of the QIDS-SR16 were found to be strong in terms of internal consistency, factor structure and convergent and discriminant validity. Using conventional scoring and conversion methods the scale was found not to concur with the HRSD-17 in categorising the severity of depressive symptoms.

[mammaP]

I don't understand a word of this report, could someone perhaps simplify it? 

I am very interested because that my ex psychiatrist is named on it, he used to work at that hospital. 

[Altostrata]

It relates to a specific questionnaire used in the STAR*D study to diagnose depression. This study found the questionnaire was invalid.

[Altostrata]

1boringoldman.com discussion at http://1boringoldman.com/index.php/2013/05/31/a-lot-better-than-this/