• Journal Article

Measuring antenatal depression and anxiety: Findings from a community-based study of women in Hyderabad, Pakistan

Citation

Karmaliani, R., Bann, C., Mahmood, M. A., Harris, H., Akhtar, S., Goldenberg, R. L., & Moss, N. (2007). Measuring antenatal depression and anxiety: Findings from a community-based study of women in Hyderabad, Pakistan. Women & Health, 44(3), 79-103. DOI: 10.1300/J013v44n03_05

Abstract

This study evaluated the psychometric properties of two possible measures of depression and anxiety among pregnant women in Pakistan for use in the Global Network for Women's and Children's Health Research project, a collaborative, international multi-site research network investigating methods for improving pregnancy and birth outcomes in developing countries. The first measure, the Aga Khan University Anxiety and Depression Scale (AKUADS), is an Urdu language scale originally developed for the general Pakistani population, whereas the second measure, the How I Feel scale, was designed for pregnant women in the United States. In an earlier pilot study, we found that the two scales demonstrated similar levels of diagnostic validity. Because neither scale was designed for the specific population of interest, item response theory analyses were conducted to evaluate the psychometric properties of the scales at three levels of measurement: scale, item, and response option. The study results provide insights that may be useful to researchers or clinicians developing or using scales in this population. In particular, our findings suggest that scales designed for populations with lower literacy, such as our target population, may improve data quality by including no more than three response options (e.g., almost always, sometimes, and never) and keeping the direction of item wording consistent throughout the scale. Based on the results from the current study, we recommend a short form of the AKUADS which removes poorly functioning items and reduces respondent burden while retaining the reliability and validity of the longer form.