Objective: To assess the operating characteristics of the Beck-Depression Inventory-II (BDI-II) and the BDI-II cognitive subscale (BDI-II-cog) in screening for major depression (MDD) in post-myocardial infarction (MI) patients.
Methods: Between October 2003 and July 2005, 131 post-MI patients admitted to an urban academic medical center completed the BDI-II and a semi-structured interview for depression within 72 hours of symptom onset. Sensitivity, specificity, positive and negative predictive values, overall correct classification, and likelihood ratios for various cutoff values on both scales were evaluated by comparing scores to interview diagnosis of MDD. Receiver-operator curves (ROC) were also calculated and area under the curve (AUC) measured.
Results: The optimal cutoff value for the BDI-II was > or = 16, with a sensitivity of 88.2% and a specificity of 92.1%. Cutoff values of > or = 3 or > or = 4 were both acceptable for the BDI-II-cog (sensitivity = 88.2% and 82.4%, respectively; specificity = 81.6% and 88.6 %, respectively). AUC was 0.96 for the BDI-II and 0.89 for the cognitive subscale.
Conclusions: Effective depression screening is important in post-MI patients because of depression's independent association with morbidity and mortality following MI. Our results suggest that the BDI-II and its cognitive subscale are effective tools for screening for MDD in post-MI patients.