Prospective Testing of the Brisbane Postnatal Depression Index

Abstract

Over 50 percent of women have one or more risk factors for postnatal depression during pregnancy or in the perinatal period, but only 10 to 15 percent become clinically depressed. The objective of this study was to prospectively test the Brisbane Postnatal Depression Index (referred to here as Index), to validate a theoretical index that was developed earlier, and to establish whether the index could be introduced as a clinically useful method to detect women who may be at risk for developing postnatal depression. Antenatally, women were asked about social support and about personal and family history of mental illness, including postnatal depression. Responses were scored according to predefined ratings on the Index. In the postnatal wards, 353 women were recruited and their scores for "blues," social support, feelings about the baby, and satisfaction with the birth process were added. Sixteen weeks after hospital discharge, women were asked to complete the Edinburgh Postnatal Depression Scale. The Brisbane Index was validated by the number of women scoring more than 12 on the Edinburgh Postnatal Depression Scale at 16 weeks postpartum who were correctly predicted by a score of more than 6 on the Index. Sensitivity, specificity, positive predictive value, and negative predictive value for the Index, using >6 as a cutoff point, were calculated. "Ease of use" was assessed informally with participants and staff responsible for administration and scoring the instrument. Compared with results from the derivation study, prospective testing of the index showed an improvement in sensitivity from 36.3 to 47.5 percent and a small decrease in specificity, but no improvement on the positive predictive value from 39.8 to 39.6 percent. The Brisbane Postnatal Depression Index was validated in a prospective sample, but its sensitivity and specificity require improvement before introduction as a measure of prediction.