| Sub-title | |
| Authors | V Zammit A Buhagiar V Grech |
| Abstract | Aim: The development of national weight for gestational age centile charts with pathological subgroup analysis and comparison to other populations. Method: Anonymised birth weight for gestational age data with relevant maternal and neonatal observations over the period 1995-2005 were obtained from national statistics. The formats were standardized and imported into an SQL database that enabled filtration for single live births and grouping by sex. The data was scrutinized manually for obvious keying errors. The best estimate of gestational age from last menstrual period (LMP) and expected date of delivery (EDD) was selected using established guidelines. Erroneous outliers were detected by comparing the gestational age with the binned median and standard deviation. A Box-Cox transform was used to fit an LMS regression and generate separate centile charts within the R package. ICD 9 coding of the original data enabled selection of maternal and infant pathological subgroups. AutoRegressive Integrated Moving Average (ARIMA) was used to check for seasonal variation in birth weight. Results: Processing of the initially filtered 22,396 males and 20,936 females revealed the expected paucity of births below 35 weeks but still enabled the fitting of charts starting at 24 weeks. Infants of diabetic mothers, those with pregnancy induced hypertension, plus other infant pathology groups were compared to the fitted model. Conclusion: Centile charts for Maltese neonates are presented. Median Maltese term birth weight was just under 100g of standard Caucasian models. Possible reasons for this are proposed |
Published in: | |
| Journal | Malta Medical Journal |
| Volume | 18 Supplement |
| Page | |
| Date | 01/12/2006 |
| Link to journal | |
| Key word | gestational age centiles |