Analysis of Changes in Birthweight Pattern in Taiwan

• Main Authors: Li-Hsing Ho, 何黎星
• Other Authors: Jia-Yuh Chen
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/51937414134163343509

碩士 === 中山醫學大學 === 醫學研究所 === 90 === Background： Changes in pattern of birthweight is an important reference for population health and birth certificate computerized file is the most often used resource. Nevertheless, there was no study use this dataset to reveal this important topic. Objectives：There were three objectives in this study. First, to evaluate the completeness and implausibility of items in Taiwan birth certificate computerized file by years; second, to examine the different methods for excluding the implausible gestational age and birthweigh match and determine the suitable method for Taiwan; third, to illustrate the trends of changes in pattern of birthweight in Taiwan from 1978 to 1997. Materials & methods: Taiwan birth certificate computerized files of year 1978 to 1997 were provided by Ministry of Internal Affair. For analysis of implausible gestational age (GA) and birthweight (BW) matches we used every year. For other analysis we used only the year of 1978,1983,1988,1993 and1997. Methods for excluding implausible GA and BW matches included: two inter 75 and 25 percentiles, two standard deviation, expert opinion and deviation from normal probability plot. BW <2500 gm was defined as low birthweight (LBW), 1500 gm <= BW <= 2499 gm was defined as moderate low birthweight (MLBW), BW < 1500 gm was defined as very low birthweight (VLBW). GA < 37 weeks was defined as preterm births and GA >= 37 weeks was defined as term births. Results： 1. The quality of birth certificate computerized files improved yearly. The missing and implausible numbers of two most important two variables (GA and BW) were less than 2000 which were trivial as compared to total number of more than three hundred thousands. Missing rates of some variables (para, order, birth date, education of mother) increased among high risk groups (high parity, multipara, mother’s age less than 14, years of education less than 7). 2. The inter 75and 25 percentile and SD methods were not very useful in Taiwan data. Expert opinion method recommended by Alexander was used for further analysis. The number and rate of implausible GA and BW matches according to Alexander method did not changed across years. The fetal growth charts depicted according to the data cleaned by Alexander method seemed more acceptable. 3. LBW rate increased by years, median BW also decreased, percentage of lower BW categories (≦2250 gm) i ncreased by years and decreased for higer BW categories (≧3500gm). There were no obvious increasing or decreasing trends of LBW rate by term and preterm births. Because the preterm LBW rates were almost twenty times higher than term LBW rates yearly and the incidence of preterm births also increased. The combination of the two factors resulted in the increasing of total LBW rate. Conclusions: To explain the changes in pattern of birthweight across years, we should first consider the changes in the quality data itself and than understand the changes in the proportion of preterm with term births. Factor standardized adjusted LBW rate was needed for trends analysis. Increasing of incidence of preterm births was the main cause of increasing of LBW rate. Multi-factorial analysis was needed in the future to better understand this problem.