Gipson and Hindin Respond

We appreciate the opportunity to respond to the issues raised by Burnie et al. regarding our mixed-method analysis of pregnancy termination in Bangladesh. With respect to ascertainment of pregnancy status, all women within the Sample Registration System (SRS) households are asked to report on their menstrual status during the quarterly visits. If amenorrheic, women are queried as to the possible cause (e.g., use of injectable contraception, suspected pregnancy). If a woman is unsure of her pregnancy status, her record is flagged in the SRS database for follow-up. If her pregnancy is confirmed, she is asked for the date of her last menstrual period. Albeit imperfect, this method is less subject to recall bias as compared with retrospective reporting of pregnancies and pregnancy loss.1,2 A more detailed explanation of the SRS and pregnancy ascertainment are described elsewhere.3,4 Attrition over the study period was relatively low as compared with other longitudinal data sets. The sample was comprised of couples in which both spouses answered the 1998 fertility questions. Of these couples, over 82% remained active through the 5-year study period. Even when comparing the sample women to all women who participated in the 1998 survey (n = 5 273), there were no significant differences (P < .05) apart from a slightly higher mean number of children among our sample (2.45 vs 2.37 children; P value = 0.035). We agree with Burnie et al. that walking distance does not wholly define access. However, given the existing secondary data available for this analysis and literature linking higher mobility with lower fertility among Bangladeshi women,5 we felt walking distance to be the most appropriate available proxy measure to assess the potential physical and psychological barriers for women in accessing health services in this context. During qualitative data collection, there were nightly staff meetings in the field, in addition to further crystallization of study findings by the entire study team through peer review of transcripts and memo-writing on emerging and dominant themes. The themes identified during the fieldwork served as a basis for the results presented. Lastly, we would like to concur with Burnie et al. that this study is not meant to be representative of all of Bangladesh, or even all of rural Bangladesh. Despite limitations, these data shed light on the contextual circumstances of pregnancy termination. We hope that our study prompts further investigation and a more holistic understanding of pregnancy termination within and outside of Bangladesh

Environmental health · Family medicine · Global Maternal and Child Health · Medicine

• Collecting Data on Pregnancy Loss

  John B Casterline•Studies in Family Planning•1989

  Add to list

• Individual and Community Aspects of Women's Status and Fertility in Rural Bangladesh

  Deborah Balk•Population Studies•1994

  Add to list