Successful five-item triage for the broad spectrum of mental disorders in pregnancy - a validation study.

BACKGROUND

Mental disorders are prevalent during pregnancy, affecting 10% of women worldwide. To improve triage of a broad spectrum of mental disorders, we investigated the decision impact validity of: 1) a short set of currently used psychiatric triage items, 2) this set with the inclusion of some more specific psychiatric items (intermediate set), 3) this new set with the addition of the 10-item Edinburgh Depression Scale (extended set), and 4) the final set with the addition of common psychosocial co-predictors (comprehensive set).

METHODS

This was a validation study including 330 urban pregnant women. Women completed a questionnaire including 20 psychiatric and 10 psychosocial items. Psychiatric diagnosis (gold standard) was obtained through Structured Clinical Interviews of DSM-IV axis I and II disorders (SCID-I and II). The outcome measure of our analysis was presence (yes/no) of any current mental disorder. The performance of the short, intermediate, extended, and comprehensive triage models was evaluated by multiple logistic regression analysis, by analysis of the area under the ROC curve (AUC) and through associated performance measures, including, for example, sensitivity, specificity and the number of missed cases.

RESULTS

Diagnostic performance of the short triage model (1) was acceptable (Nagelkerke's R(2)=0.276, AUC=0.740, 48 out of 131 cases were missed). The intermediate model (2) performed better (R(2)=0.547, AUC=0.883, 22 cases were missed) including the five items: ever experienced a traumatic event, ever had feelings of a depressed mood, ever had a panic attack, current psychiatric symptoms and current severe depressive or anxious symptoms. Addition of the 10-item Edinburgh Depression Scale or the three psychosocial items unplanned pregnancy, alcohol consumption and sexual/physical abuse (models 3 and 4) further increased R(2) and AUC (>0.900), with 23 cases missed. Missed cases included pregnant women with a current eating disorder, psychotic disorder and the first onset of anxiety disorders.

CONCLUSIONS

For a valid detection of the full spectrum of common mental disorders during pregnancy, at least the intermediate set of five psychiatric items should be implemented in routine obstetric care. For a brief yet comprehensive triage, three high impact psychosocial items should be added as independent contributors.