Brief Depression Screener Developed Using IRT for Antenatal and Postpartum Women

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Estimates of antenatal and postpartum depression vary from 7 percent to upwards of 30 percent depending on criteria for depression, the type of assessment used, and the timing of the assessment. Thus, approximately 320,000 to over 1 million women each year experience symptoms of perinatal depression. Estimates of antenatal depression are harder to obtain because of the lack of appropriate diagnostic criteria. Despite this, no screener currently exists that has been developed and tested with both antenatal and postpartum women that uses concise, simple language, and a consistent response option set. The overall aim of this project is to develop a brief screener to assess depressive symptoms among antenatal and postpartum women. The proposed screener will be easier for patients to understand, will take less time, and will be more psychometrically sound than current screeners. In Phase I we will develop approximately 30 items that are specific to depression in antenatal and postpartum women. We will test these items using cognitive interviewing among a sample of 10 antenatal and 10 postpartum women in addition to 32 general depression items we have previously developed. Experts will review the results of the cognitive interviewing to develop the item pool that will be tested in Phase II. The Phase II item pool will consist of at least 20 general depression items, 10 items specific to antenatal women, and 10 items specific to postpartum women. In Phase II, 500 antenatal and 500 postpartum women (70 percent from private sector sites and 30 percent from public sector sites) will complete the EPDS (as a screener into the study), BDI-II, the PROMIS depression items, our items, and the depression module of the SCID. IRT analyses will be conducted on these data to develop a 7-9 item screener. We anticipate that approximately 5-7 of these items will be applicable to both antenatal and postpartum women and approximately 2-3 of these items will be sample specific. The developed screener will then be given to 400 women who will take the survey twice, approximately 4-7 days apart. Conclusions regarding test-retest reliability and patient satisfaction with taking the screener using different modes of technologies will be made from this study.