Listed below are links to the tabulations of birthweight-specific fetal and infant mortality rates by health district and race for the 1994-1996 birth cohorts. These data update the information that was provided to you by perinatal region for the 1991-1993 birth cohorts as part of the World Health Organization (WHO) Collaborating Center report. In addition, tabulations are provided describing the gap in infant mortality rates between each district and a referent group with low infant mortality.

• fetal deaths
• neonatal deaths (birth up to 28 days of age)
• postneonatal deaths (28 days to 1 year).

Please note that these data are cohort mortality rates prepared using linked birth and infant death certificates. Thus, the numbers of deaths and rates will differ somewhat from tabulations based on standard vital statistics reports that we have previously disseminated. There are two reasons for this difference. First, infant deaths are attributed to the year of birth in the cohort method and to the year of death in the standard method. Second, county of residence for infants who died is attributed to the mother's county of residence at birth in the cohort method and to the infant's county of residence at death in the standard method. Using the cohort method, 1996 is the most recent year available, since infant deaths among the 1996 birth cohort may occur in 1996 or 1997. We have previously reported on infant mortality using the standard method for 1997, which is based on births and infant deaths occurring in 1997. Although the cohort approach is less timely, its advantage is that it allows calculation of mortality rates by birthweight due to the linkage of information on birth and death certificates.

For each health district, we have provided an assessment of the gap in infant mortality between that observed for the district and that observed among a population with a low rate. We have adopted the criteria established by the WHO Collaborating Center to define this referent population: college-educated white women >20 years of age residing in the metropolitan Atlanta area (this is a slight modification of the WHO definition which used the Atlanta perinatal region). These tabulations show the difference in both infant and late fetal death rates between the district and the referent population. For the difference in infant mortality, this gap is partitioned into three components, including the percentage of the gap that is attributable to differences in:

• the birthweight distribution (e.g., the frequency of low birthweight), corresponding to the segment of the gap attributed in the WHO report to maternal health,
• birthweight-specific neonatal mortality rates, analogous to the WHO perinatal care category, and
• birthweight-specific postneonatal death rates, corresponding to the WHO infant care category.

There are two differences between the method used by WHO and the method we used to define these gaps. First, we partitioned only the infant death rates, not the total fetal-infant death rates, given differences in strategies for preventing a late fetal death versus preventing premature live birth or an infant death. Second, we do not attribute all excess infant deaths in the <1,500g birthweight category to "maternal health" but instead have identified the portion of these excess deaths that are due to differences in birthweight-specific mortality rates. This reflects the substantial variations observed among districts in death rates for this group, and the potential for further reductions through improving access to regional perinatal care services. As highlighted in the WHO report and as apparent from these tables, the greatest potential benefit remains in the prevention of premature delivery, as evidenced by the much larger role that birthweight distribution plays in contributing to the observed gaps.

Lastly, we are preparing updates of the CD-ROMs provided to health districts that will include the 1994-1996 cohort data sets, enabling you to replicate and extend these tabulations.

These tabulations were prepared by Maternal and Child Health Epidemiology Unit. If you have questions about these data, please feel free to contact us at 404-657-2558.

Data for Georgia
Birthweight-Specific Fetal and Infant Mortality,
1994-1996 Birth Cohorts

All of these charts are in pdf format.