ABSTRACT
BACKGROUND: The positive predictive rate of the Japanese version of the Edinburgh Postnatal Depression Scale (EPDS) is lower than those of other versions. This study aimed to confirm whether the EPDS Japanese version reflects actual symptom frequency and to examine the possibility of improving the positive predictive rate. METHODS: This is a methodological study aimed at improving the positive predictive value of EPDS. The participants were 63 non-pregnant and 382 pregnant women. They answered the 10 questions of the Japanese version of the EPDS and reported the specific number of days as the frequency. The EPDS score (EPDS-S) and the frequency score (FREQ-S) were calculated for three factors of emotion: anhedonia, anxiety, and depression. RESULTS: The positive rates of the EPDS-S and FREQ-S in pregnant women were 6% and 8%, respectively, which were lower than those in non-pregnant women (17%). On comparing the EPDS-S with the FREQ-S, a significant underestimation of frequency was observed in approximately 3% of pregnant women. The FREQ-S showed better internal consistency than the EPDS-S. Among the factors of emotion, women tended to rate anhedonia lower in the EPDS-S than in the frequency scale. CONCLUSION: Pregnant women tended to report a lower frequency on the Japanese version of the EPDS than their actual symptom frequency, which was especially true for those with a desire to self-harm. The combined use of the FREQ-S and EPDS-S can prevent underestimation and help improve the detection rate of depression.