Reproductive Health Surveillance in the US-Mexico Border Region, 2003-2006: The Brownsville-Matamoros Sister City Project for Women's Health ORIGINAL RESEARCH

9) in 2005. We conducted in-hospital computer-assisted personal interviews that addressed prenatal, behavioral, and lifestyle factors. We evaluated survey response rates, data quality, and other attributes of effective surveillance systems. We estimated population coverage using vital records data. Results Among the 999 women sampled, 947 (95%) completed interviews, and the item nonresponse rate was low. The study sample included 92.7% of live births in Matamoros and 98.3% in Cameron County. Differences between percentage distributions of birth certificate characteristics in the study and target populations did not exceed 2.0. Study population coverage among hospitals ranged from 92.9% to 100.0%, averaging 97.3% in Matamoros and 97.4% in Cameron County. Conclusion Results indicate that hospital-based sampling and postpartum interviewing constitute an effective approach to reproductive health surveillance. Such a system can yield valuable information for public health programs serving the growing US-Mexico border population.