Aim: The aim of the present study was to examine the possibility of screening apprehensive pregnant women and mothers at risk for post-partum depression from an analysis of the textual data in the Mother and Child Handbook by using the text-mining method.
Methods: Uncomplicated pregnant women (n = 58) were divided into two groups according to State-Trait Anxiety Inventory grade (high trait [group I, n = 21] and low trait [group II, n = 37]) or Edinburgh Postnatal Depression Scale score (high score [group III, n = 15] and low score [group IV, n = 43]). An exploratory analysis of the textual data from the Maternal and Child Handbook was conducted using the text-mining method with the Word Miner software program. A comparison of the 'structure elements' was made between the two groups.
Results: The number of structure elements extracted by separated words from text data was 20 004 and the number of structure elements with a threshold of 2 or more as an initial value was 1168. Fifteen key words related to maternal anxiety, and six key words related to post-partum depression were extracted.
Conclusion: The text-mining method is useful for the exploratory analysis of textual data obtained from pregnant woman, and this screening method has been suggested to be useful for apprehensive pregnant women and mothers at risk for post-partum depression.
Keywords: Edinburgh Postnatal Depression Scale score; Mother and Child Handbook; State-Trait Anxiety Inventory; post-partum depression; text-mining method.
© 2016 Japan Society of Obstetrics and Gynecology.