Analysis on decomposed time series of Congenital Heart Defects among Children Born to New York State Residents, 1983-2005

• Activity Number: 581
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Gang Liu*+ and Ying Wang and Charlotte Druschel and Igor Zurbenko
• Companies: New York State Department of Health and NYS DOH and NYS DOH and SUNY at Albany
• Keywords: Kolmogorov-Zurbenko filter ; Seasonality ; Weekly pattern ; approximation ; congenital heart defect

Abstract

Background: Several studies have been conducted on examining the seasonality in congenital heart defects (CHDs), but the results were not consistent. The objective of this study was to explain the prevalence of nine selected CHDs using New York State Congenital Malformation Registry (CMR) data from 1983 to 2005. Methods: Time series of daily prevalence was created on children with nine selected CHDs. Kolmogorov-Zurbenko (KZ) filter was used to decompose the time series. Linear regression was applied to explain the long term trend. Graphical analysis was applied to examine seasonal and weekly patterns. Walter & Elwood test was also applied on the seasonality. Results: The long term trend could be modeled by two long term trends of the percentage of older maternal age (>35) and the percentage of Hispanic mother. Neither KZ filter nor Walter & Elwood find seasonal pattern in CHDs prevalence, while weekly pattern was found as decreased prevalence in Sunday.