As part of our first aim, we will assess the psychometric properties of the Whooley questions, by comparing the answers given to these questions against the scores obtained on the Edinburgh Postnatal Depression Scale, and we will manipulate the survey questionnaire layout of the Whooley Questions and the Edinburgh Postnatal Depression Scale in order to ensure that this survey design choice does not affect data quality. As part of our second aim, we will compare two prospective sampling protocols on patient compliance and engagement with the app.
We will use a parallel, randomised control trial study design. Participation in each part of the study (i.e., iPads in antenatal clinics, or app running on own handset) will be independent from each other. Those participants consenting to get involved in the part of the study assessing the use of iPads in antenatal clinics will be randomly assigned to complete (i) an app version of the Whooley questions and the Edinburgh Postnatal Depression Scale in which the questionnaires are presented using a scrolling layout (i.e., App screening - Scrolling), or (ii) an app version of the Whooley questions and the Edinburgh Postnatal Depression Scale in which the questionnaires are presented using a paging layout (i.e., App screening - Paging).
Participants consenting to get involved in the part of the study assessing the use of an app running on participants' own devices will be randomly allocated to one of two sampling protocols: (i) a prospective sampling protocol of 6 months consisting of the Edinburgh Postnatal Depression Scale and 5 momentary questions related to mood; or (ii) a prospective sampling protocol of 6 months consisting of the Edinburgh Postnatal Depression Scale.
We will utilise a block randomisation procedure (with blocks of 4) to generate our allocation sequence. Random numbers will be generated using Stata 13.0. Researchers conducting participant recruitment will not be involved in this randomisation procedure in order to avoid recruitment bias.