circumstances that were not met upon chart review; iii. a diagnosis was reported in general terms (e.g., heart anomaly) when a more specific diagnosis was available; or iv. the hospital report was in error, and there was no evidence in the chart for the reported diagnoses. APORS case finding is an ongoing process; children with birth defects identified during the newborn stay are added for previous years whenever they are found. This report presents birth 4 defect rates among newborns and infants up to 2 years of age, born between 2002 and 2018. The rates are presented for two distinct time periods reflecting the evolution of data collection methodologies over time. The first time period spans 2002 to 2012, during which active case verification was first introduced and utilized. The second time period spans 2013 to 2018 during which several changes took place. In 2013, APORS fully implemented an electronic reporting system for hospitals which aides in reporting by systematically flagging certain reportable conditions, including birth defects, when noted on the birth certificate. In addition to ongoing active case verification, hospital discharge data were used to ascertain new cases and the International Classification of Diseases, Tenth Revision Clinical Modification (ICD-10-CM) coding was implemented to categorize birth defects. In cooperation with CDC, rapid case identification techniques were employed (2015 and 2016 births) to quickly ascertain certain birth defects potentially related to the Zika virus. Because of the new data collection activities implemented during the 2013 to 2018 time period, rates are not comparable to the 2002 to 2012 period when active case verification alone was used. In addition to trends for the two time periods, this report as also provides and compares current birth defects prevalence for the geographies of Illinois and Chicago alone for the year 2018. METHODS Calculation and Interpretation of Rates and Confidence Intervals Thirty-five categories of birth defects are included in this study. A listing of the International Classification of Diseases – Ninth and Tenth Revisions Clinical Modification (ICD-9-CM and ICD-10-CM) codes for the selected birth defects is provided in Appendix A, together with a brief description of each birth defect. Annual incidence rates (per 10,000 live births) for selected congenital anomalies identified during the newborn hospital stay up to 2 years of age or associated with a fetal death were calculated as: The numbers of live births were obtained from the IDPH’s master birth files. Occurrence of a specific birth defect is assumed to be a rare event, therefore following a Poisson distribution. Exact confidence intervals were calculated for each rate (Hardeo & Khurshid, 1993). Where there are a large number of birth defect cases, the confidence interval is narrow, indicating that Number of infants and fetuses with selected congenital anomaly number of live births × 10,000 5 the rate is stable. Where there are few birth defect cases, the confidence interval becomes very wide, indicating that the rate is not very stable and a small change in the number of infants born with the specific birth defect could result in a large change in the rate. To compare two rates, it is important to look not just at their value, but also their confidence intervals. As a conservative approximation, if two confidence intervals overlap, then there is no evidence that the two rates are really different. If two confidence intervals do not overlap, then the rates are said to be statistically different. In this report, 95% confidence intervals are used; where the confidence intervals do not overlap, the rates are statistically different at the 5% level (p < 0.05). Analysis of Trends Trends in Illinois birth defect rates for 2002-2018 were modeled using a log-linear regression model (which is appropriate for data following a Poisson distribution). Analyses were performed using the Joinpoint Regression Program (Version 4.8.0.1, April 2020, Statistical Research and Applications Branch, National Cancer Institute). This software compares a linear model with a single slope to linear models with different slopes joined by one or more join-points. The model tests whether the slope(s) are significantly different from 0 (whether there is a change over time) and whether any change in slope between two segments is statistically significant. Multiple Comparisons Because this report examines a large number of birth defects, the corresponding statistical tests are subject to the “multiple comparison problem.” In this report, no explicit corrections were made for multiple comparisons because the focus was to detect trends, not compare trends; instead, exact probabilities are reported when discussing trends. The smaller the reported probability, the more likely it is that the difference is not simply the result of chance. FINDINGS Rates of Birth Defects for Illinois and Chicago, 2018 Birth defect rates for selected categories among Illinois and Chicago newborns in 2018 are presented in tables 4 and 5. Rates for Chicago are similar to those for Illinois as a whole. Statistically significant differences were not seen overall or for individual defect categories. 6 Trend Analysis Table 3 shows statistically significant trends were found for 10 birth defects (See Table 3). Six defects showed significance from 2002 to 2012, and three defects displayed significance during