with all other racial/ethnic groups included in the total). The gastroschisis case definition was based on the British Pediatric Association Classification of Diseases code (756.71) or the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for gastroschisis (756.73, or before 10/1/2009, 756.79, with verification to confirm cases of gastroschisis, because the previous code was shared with omphalocele). Gastroschisis cases included live births, fetal deaths, † and elective terminations. § Data were pooled at CDC, and gastroschisis prevalence was calculated for each year, maternal age group, and race/ethnicity. Prevalence was calculated as number of gastroschisis cases among all birth outcomes divided by the total number of live births. The denominators of total number of live births in the same catchment area as the birth defects surveillance program were reported by states or obtained from public use data files. Poisson exact methods were used to calculate 95% CIs for each prevalence estimate. Prevalence ratios were calculated by dividing the prevalence during 2006-2012 by the prevalence during 1995-2005, and CIs for the prevalence ratios were calculated using Poisson regression.Because the comparison of prevalence between the two study periods involved an artificial breakpoint during the 18-year data span and only examined pooled prevalence within those periods, joinpoint regression analysis was used to identify statistically significant changes in the annual prevalence of gastroschisis over the course of the entire study period (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012). Joinpoint regression initially models annual trend data by fitting a straight line (i.e., zero joinpoints). Then, joinpoints are added, one at a time, and a Monte Carlo permutation test is used to determine the optimal number of joinpoints. Each joinpoint in the final model corresponds to a significant change in the trend, and an AAPC and its 95% CI are calculated to describe how the rate changes within each time interval (3). The estimated overall percent change was calculated by first converting the AAPC to the projected single year change in prevalence and then exponentiating to the number of years studied minus one to estimate the total increase throughout the 18 years. This gives the magnitude of the increase, which