Purpose: During the years 2000-09 in the state of Texas, NMR dropped from 4.37 to 3.77, and IMR dropped from 6.58 to 5.96. Texas ranked 22nd in USA in 2008 for IMR. To develop preventive strategies and further decrease NMR/IMR, it is critical to understand the demographic factors contributing to high NMR/IMR. We studied the trends in NMR/IMR during 2000-09 in areas served by Scott and White Hospital (SWH), a tertiary care center in Bell County. We compared NMR/IMR trends with demographic factors in surrounding 7 counties: (See Fig 1). Our hypothesis was: IMR/NMR in Bell County was lower than surrounding counties and the variations between the counties can be explained by ethnicity, income and hospital resources.
Figure 1 Map of Bell and surrounding counties
Methods: Births and death linked mortality data for Bell and surrounding 7 counties were obtained from Texas DHHS. Deaths were analyzed by county of residence. Median income (www.kidscount.org) data were compared with NMR/IMR. For study purpose, five small counties, Burnet, Coryell, Falls, Lampasas, and Milam were merged into one group labeled as ‘other'. The data, of NMR/IMR, year of birth/death, race, median income of four groups were compared. Multivariable Negative Binomial regression analysis was utilized to estimate the effects of county, ethnicity, median income, and year on NMR/ IMR. Total number of births was adjusted in the model as an offset variable.
Results: During study period, births steadily increased in Bell and Williamson counties. A major increase was among Whites and Hispanics. Births in Blacks remained the same. NMR/IMR were lower in Williamson (2.79/3.97) and Bell county (4.34/6.94) (P<.001) both with level III NICUs. Proportion of Black babies born was lowest in Williamson county (5.13 %) and highest in Bell county (23.2%). NMR/IMR remained significantly higher (7.93/12.3) among Black babies (P<.001) during the study period. Median annual income was highest ($64454) in Williamson county, followed by Bell ($ 41115), Other counties ($37636), and McLennan county ($36344) (p<.001). After adjusting ethnicity, median income and years in multivariable models, county effect became insignificant. Ethnicity effect was still significant (p<.001), and the NMR/IMR of Black was significantly higher than White, Hispanic, and other ethnicity even after adjusting county and median income.
Conclusion: Though both NMR/IMR decreased during the study period among the 8 counties served by SWH, significant decrease was seen mainly in Whites and Hispanics. Decrease was associated with higher income and greater access to tertiary care services in the counties. Our study provides a model to analyze NMR/IMR data county by county and available health services to better understand the factors contributing to high NMR/IMR. Higher mortality rates among the Black population underscore the need for improved maternal and neonatal health care services for this population in all counties .