Performance Improvement 1999. Health Statistics

02/01/1999

TITLE: Race Reporting on Birth Records by Multiracial Women: An Evaluation of Question Format Alternatives

ABSTRACT: This report examines the implications of changing the format of the race item on the birth record filled out for every birth in the United States at the location of birth. Procedures for filling out birth certificates vary from State to State and, in some States, from hospital to hospital. To provide uniform information to the vital registration system, the Federal government, in cooperation with the States, developed the U.S. Standard Certificate of Live Birth, which records relevant information about the race and origin of the parents, but not the baby. This study: (1) evaluates how women of multiracial and Hispanic backgrounds interpret the race question on the birth certificate; (2) evaluates experimental race questions such as the use of a multiracial cue and the use of a "mark all that apply" format; and (3) provides information about the response processes that women use when answering questions about race. Nine States and the District of Columbia recruited 763 women to participate in this study. The study tests the effects of different questions on race both in a mail and a follow-up telephone survey. The mother's self-reported race is compared with what she reported on the actual birth record. Further, for those residents with two children, the study compared consistency of race reporting between the last two birth records. The report concludes that: (1) the percentage of women reporting more than one race increases when the race questions include a multiracial cue or "mark all that apply" format; (2) women who report more than one race are more likely to enter two specific races rather than a term such as "multiracial"; (3) among the mothers whose parents belonged to two or more racial groups, self-reports of race on the last two birth records were inconsistent 25 percent of the time and; (4) self-reports of race were different 40 percent of the time between the birth record and the standard version of the mail survey.

AGENCY SPONSOR: National Center for Health Statistics

FEDERAL CONTACT: Susan Schechter-Ryan

PHONE NUMBER: 301-436-7111

PIC ID: 5918

PERFORMER ORGANIZATION: Westat, Inc. Rockville, MD
 

TITLE: The Effect of Condition Sequencing Order on Cause-of-Death Statistics

ABSTRACT: This study evaluated whether the direction of the order in which causes of death are sequenced on the death certificate influences either selected underlying cause of death or the number of causes recorded. Physicians read one of four case vignettes and completed either the standard or an experimental medical certification section. On the standard format, certifiers were to enter temporally-backward sequences of conditions. On the experimental format, certifiers were to enter temporally-forward sequences. For each response, underlying cause was selected according to the rules of the International Classification of Diseases and the number of entered causes was counted. The results were that the certificate format did not influence significantly the distribution of selected underlying causes of death. Significantly more causes were recorded, on average, on the experimental certificates than on the standard certificates, but the difference in number was small. The data do not indicate that reversing the sequencing order would elicit better cause of death information for purposes of tabulating underlying cause of death, but suggest that reversing the sequencing order has the potential to elicit more information for multiple cause of death analyses. This final study evaluated the effect of the medical certification format on the quality of death statistics.

AGENCY SPONSOR: National Center for Health Statistics

FEDERAL CONTACT: James Weed, Ph.D.

PHONE NUMBER: 301-436-8951

PIC ID: 7098

PERFORMER ORGANIZATION: Research Foundation of the State University of New York Binghamton, NY