back Author Index Title Index Track Index forward
back   fwd

Abstract


Using GIS to Track IMR in a Predominantly Rural Health District
Track: Health and Human Services
Author(s): Simon Choi, Claire Morin

Introduction:According to the 1995 Women’s Health Survey conducted by the Georgia Department of Human Services, Division of Public Health, groups at risk for starting prenatal care late in pregnancy include women under 25, those whose families live below the poverty line and those with less than a high school education, among others.The population-based Georgia Pregnancy Risk Assessment Monitoring System Report shows that over the period from 1993 to 1997, nearly one quarter of Georgia mothers did not seek prenatal care by the end of the first trimester of pregnancy.

Methods:

Data for infant births and deaths in the ZIP code areas of the North Central Health District was obtained from the Georgia Linked Birth and Infant Death File for years 1996 through 1998.This information was classified by maternal age group and race.These are the most recent years for which ZIP code data is currently available in Georgia, the 1998 data being preliminary.Any ZIP code areas that overlapped into other health districts were eliminated from this study.


Infant mortality rates (IMRs) were calculated for all district ZIP code areas then standardized.For each year, the district crude IMR was employed as the standard in calculating standardized infant mortality ratios (SMRs) for ZIP code areas.Subsequently, ArcView GIS Version 3.2 was used to map the IMRs and SMRs of the health district (see fig. 1 through 6).An additional map was created to show the IMR trends between 1996 and 1998 (see fig. 7).


Risk factors were identified, and variables were selected accordingly: median household income (1997), percent of families living below poverty level (1990), percent rural population (1990), percent of population with high school diploma/equivalency or higher (1990), percent of population with bachelor’s degree or higher (1990), proportion of mothers who were of black race (1996, 1997, 1998), proportion of mothers age 40 and over (1996, 1997, 1998), and proportion of mothers ages 19 and under (1996, 1997, 1998).


For each of the years 1996, 1997 and 1998, the proportion of births by mothers of black race was calculated. In addition, the proportion of teen mothers and mothers of advanced age (40 and over) were calculated.These numbers, along with the other selected variables, were entered into the SPSS Version 7.5 statistical software.Simple Linear Regression and Spearman Rank Correlation tests were performed to determine associations between risk factors and SMRs.

The 1996, 1997 and 1998 crude IMRs as well as SMRs calculated for ZIP code areas within the Georgia North Central Health District (5-2) have been mapped using ArcView GIS, grouped by quantile. The IMR maps and SMR are nearly identical for each year.The Zip code areas have been mapped according to quantile to limit the effect of outliers.For 1997, the following outliers were excluded from statistical analyses:.

Simon Choi
Mercer University School of Medicine
1550 College St
Macon, GA 31207
USA

Phone: 912-301-5569
Fax: 912-301-4143
E-mail: choi.s@gain.mercer.edu