Comparing Race and Ethnicity as Reported on Infant

Comparing Race and Ethnicity as Reported on Infant Death and Matching Live Birth Certificates, North Carolina 2002 -2006 Paul A. Buescher, Ph. D. State Center for Health Statistics NC Division of Public Health June 3, 2008

Introduction During 2002 -2006, 99% of NC resident infant death certificates were successfully matched to a live birth certificate. n This matched file allows joint analysis of the variables from each file. n 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Introduction n The purpose of this project is to compare race and ethnicity as reported on infant death and matching birth certificates, and to compute infant death rates calculated using race/ethnicity at birth vs. race/ethnicity at death. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Introduction n NC has not yet adopted the new national model live birth or death certificates. NC certificates currently collect data on race via open-ended, fill-in-the-blank boxes. On the birth certificate, both mother’s and father’s race are collected; the data are usually tabulated by mother’s race. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Introduction In a study of 2002 NC birth certificates, we found that mothers wrote in more than 600 different text versions of mother’s “race. ” n Extreme examples of the race written in by the mother were “Son of God” and “Dominant White. ” “Human Race” was a common entry. n 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Introduction n These many open-ended responses to race on birth and death certificates are coded by Vital Records staff into ten standard racial categories according to rules provided by the NCHS: white, black, American Indian, Chinese, Japanese, Hawaiian, Filipino, Other Asian or Pacific Islander, other race, and unknown race. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Introduction n n On both the NC birth and death certificates, Hispanic origin is ascertained by a yes/no question. The data on race and ethnicity were combined for this project into 11 categories: white, non-Hispanic; black, non-Hispanic; American Indian, non. Hispanic; etc. for the other 7 race groups; and Hispanic of any race. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Results There were 5, 024 matched infant deaths during 2002 -2006 (defined by NC residence at birth) and for 4, 627 (92%) of these infants deaths the detailed race/ethnicity category was the same on both the infant death and the live birth certificate. n Conversely, the rate of disagreement was 8%. n 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Results n n Race-specific infant death rates are often published by states using race as recorded on the death certificate to specify the numerator. The following chart compares infant death rates when race/ethnicity from the death certificate is used versus when mother’s race/ethnicity from the matching birth certificate is used. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Results n Due to the small numbers of infant deaths in some of the racial categories, the data for Chinese, Japanese, Hawaiian, Filipino, and Other Asian are combined into one “Asian” category in the chart. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Results 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Results n The infant death rates for Hispanics and Asians are one-fifth and onefourth larger, respectively, when using the mother’s race/ethnicity from the matching birth certificate rather than the infant’s race/ethnicity from the death certificate. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Conclusions n n Mother’s race/ethnicity from the birth certificate is presumably more accurate because it is self-reported. Race/ethnicity on the death certificate is reported by a funeral director, ideally after asking the family. Race as collected on vital records is imprecise and subject to the vagaries of reporting by many people. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Conclusions n Tabulations of infant death rates by race and ethnicity (especially for smaller racial/ethnic groups) should use the mother’s race and ethnicity from the matching birth certificate for the numerator, which are likely to be more accurate. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

References n Buescher PA. Congruence between race and ethnicity reported on infant death and the matching live birth certificates: North Carolina, 20022006. SCHS Studies, No. 156. State Center for Health Statistics, February 2008. Available at: www. schs. state. nc. us/SCHS/pdf/SCHS 156. pdf 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

References n Buescher PA, Gizlice Z, Jones-Vessey KA. Discrepancies between published data on racial classification and selfreported race: evidence from the 2002 North Carolina live birth records. Public Health Reports. 2005; 120: 393 -398. 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008

Contact Information Paul A. Buescher, Ph. D. Director State Center for Health Statistics NC Division of Public Health 1908 Mail Service Center Raleigh, NC 27699 -1908 (919) 715 -4478 paul. buescher@ncmail. net www. schs. state. nc. us/SCHS 2008 NAPHSIS Annual Meeting Celebrating 75 Years of Excellence Orlando, FL June 1 st – 5 th, 2008