Very low birth weight births in Georgia, 1994-2005: trends and racial disparities.

OBJECTIVE: To investigate the nature of very low birth weight (VLBW) births in Georgia-a major contributor to the overall and the black-white disparity in infant mortality-as a step toward elucidating strategies for reducing VLBW births.
METHODS: This population-based retrospective cohort study utilized maternally linked vital records data from Georgia to examine the status of and contributors to the VLBW rate for non-Hispanic blacks and whites by comparing trends in the proportion represented by singleton versus multiple gestations, first versus recurrent events, and specific subtypes over three, consecutive 4-year periods (1994-1996 through 2003-2005); and logistic regression to model the risk of various subtypes of VLBW as a function of maternal and obstetrical characteristics.
RESULTS: Georgia's VLBW rate remained unchanged from 1994-1996 to 2003-2005, although there was a significant decrease in the rates of twin and first VLBW and a significant increase in recurrent VLBW. For both first and recurrent VLBW, there was a statistically significant increase for blacks and a decrease for whites. The strongest risk factor for a VLBW birth of any subtype for blacks and whites was a prior VLBW, with recurrent VLBW accounting for 4.8-16% of all VLBW depending upon the subtype.
CONCLUSION: From 1994-1996 to 2003-2005, the rate of recurrent VLBW increased while the rate of first VLBW decreased in Georgia. For both first and recurrent VLBW, the black-white disparity widened. Because the strongest risk factor for a VLBW birth is a previous one, there is a need to identify strategies to prevent a woman's first VLBW birth and to reduce recurrences.