Birth certificate data quality: A study of maternal medical risk




Restrepo, Elizabeth

Journal Title

Journal ISSN

Volume Title



Live births in the United States have been registered since 1900, using the Standard Certificate of Live Birth. This form is the means by which uniformity of content in documents used to obtain birth data is accomplished. Birth certificate data are used for many purposes including maternal and perinatal health surveillance, policy research, and program development; resource allocation; and innumerable demographic tabulations. The accuracy of birth certificate data is of critical importance to the use of the birth certificate as a primary source of population-based maternal and perinatal data, but the quality of data recorded on the birth certificate has been questioned by data users for over a half century. An upcoming revision of the birth certificate, for release in 2003, and a recent change in birth certificate data collectors in one large hospital provided the motivation and opportunity to examine data recorded on the birth certificate by both medical records clerks and registered nurse data collectors.

The purpose of this retrospective descriptive study was to examine and describe the data recorded in the “Medical Risk Factors for This Pregnancy” section on the Texas birth certificate, for one institution, in 1999, when clerks collected the birth certificate data, and in 2000, when registered nurses collected the data. Utilization of the check-box and text-field formats by clerks and by nurses was the measure of data quality in this study. The sample population consisted of births recorded in the months of May through December, inclusive, for the years 1999 and 2000, in one North Texas institution. Secondary analysis of existing birth data was conducted using exploratory, descriptive, and non-parametric statistical methods.

Chi-square analysis of recording of medical risk factors for each data collection year demonstrated that the reporting of maternal medical risk was significantly greater in 2000 when registered nurses collected the birth certificate data (p < .05). The 2000 maternal population was, therefore, characterized as higher risk than the 1999 maternal population.



Health and environmental sciences, Communication and the arts, Birth certificate, Data quality, Maternal, Medical risk