Abstract

Background. The present study provides additional data on the psychometric properties of the 30-item Inventory of Depressive Symptomatology (IDS) and of the recently developed Quick Inventory of Depressive Symptomatology (QIDS), a brief 16-item symptom severity rating scale that was derived from the longer form. Both the IDS and QIDS are available in matched clinician-rated (IDS-C 30 ; QIDS-C 16 ) and self-report (IDS-SR 30 ; QIDS-SR 16 ) formats. Method. The patient samples included 544 out-patients with major depressive disorder (MDD) and 402 out-patients with bipolar disorder (BD) drawn from 19 regionally and ethnicically diverse clinics as part of the Texas Medication Algorithm Project (TMAP). Psychometric analyses including sensitivity to change with treatment were conducted. Results. Internal consistencies (Cronbach's alpha) ranged from 0·81 to 0·94 for all four scales (QIDS-C 16 , QIDS-SR 16 , IDS-C 30 and IDS-SR 30 ) in both MDD and BD patients. Sad mood, involvement, energy, concentration and self-outlook had the highest item-total correlations among patients with MDD and BD across all four scales. QIDS-SR 16 and IDS-SR 30 total scores were highly correlated among patients with MDD at exit ( c =0·83). QIDS-C 16 and IDS-C 30 total scores were also highly correlated among patients with MDD ( c =0·82) and patients with BD ( c =0·81). The IDS-SR 30 , IDS-C 30 , QIDS-SR 16 , and QIDS-C 16 were equivalently sensitive to symptom change, indicating high concurrent validity for all four scales. High concurrent validity was also documented based on the SF-12 Mental Health Summary score for the population divided in quintiles based on their IDS or QIDS score. Conclusion. The QIDS-SR 16 and QIDS-C 16 , as well as the longer 30-item versions, have highly acceptable psychometric properties and are treatment sensitive measures of symptom severity in depression.