Comparison of a self-rated and a clinician-rated measure for identifying depression among adolescents in a primary-care setting.

OBJECTIVE

To compare the diagnostic accuracy of a self-rated and a clinician rated measure of depression for primary care use in school setting by pediatricians.

METHODS

Two tools for screening depression were administered to early adolescents in three schools. These included the self-rated Beck Depression Inventory (BDI), pediatrician rated Children's Depression Rating Scale-Revised (CDRS-R), and ICD-10 clinical interview by a psychiatrist as reference standard. These tools were compared for their overall performance using Areas Under the Curve (AUC) of Receiver Operating Characteristic (ROC) curves. The optimal screening threshold score for both tools were identified from their sensitivity and specificity plotted for all threshold scores. For the optimal cut-off scores, the diagnostic accuracy parameters like sensitivity, specificity, predictive values, likelihood ratio and diagnostic odds ratio were calculated using contingency table.

RESULTS

The area under the curve for BDI was 0.67 and CDRS was 0.50 suggesting that BDI as a screening tool has better diagnostic accuracy. The optimal screening threshold score for BDI was 18 with a sensitivity of 63 and specificity of 70. For the CDRS-R cut-off score of 59, the sensitivity was 36 and specificity was 82 respectively. Using both tools concurrently improved the diagnostic accuracy.

CONCLUSIONS

Using the ROC characteristics and various validity indices, the authors showed that BDI has better sensitivity and CDRS-R a better specificity. It might be prudent to use both these instrument simultaneously to improve the identification of depression in primary care settings like school health clinic.