Introduction
For decades, antenatal care has been universally considered as beneficial and strongly supported by governments and the medical profession. It has been referred to as "the perfect example of preventive medicine." During the last decade, there has been growing skepticism, especially among epidemiologists, about the effectiveness of routine antenatal care. The appropriateness of procedures, tests and technological innovations for all pregnancies, regardless of risk, has also come into question. It is important to have one place for information about the measurements of the babies at birth and the birth outcome in each city. In the Birth Registry Office in Riyadh, these statistics are lacking, although the information needed to measure birth outcome and the effect of antenatal care are all registered in the mother's and babies files. We decided to conduct a study to measure birth weight in the population studied, compare the mean birth weight in the population studied with the standardized mean and assess the effect of antenatal care on birth outcome in the form of birth weight and admission to the neonatal intensive care unit (NICU).
Methodology
Initially, a pilot study was conducted. We selected 40 random days in a two-year period (October 1995-October 1997) and compared the birthweight distribution of the babies born on those days to the WHO standardized distribution. We enrolled all Saudi women in the post-natal wards following delivery at Riyadh Military Hospital for a one-month period (27 Oct.-27 Nov. 1997). We designed a questionnaire to cover the mother's socio-demographic data and medical, obstetrical and drug histories, antenatal care received, mode of delivery and birth outcome. The principal investigator interviewed all women on the post-natal ward during the day of delivery and reviewed the mothers' medical files. We used this Z-score and the need for NICU admission as the outcome measurements. Parents' educational level, consanguinity, mother's age, parity, gestational age, antenatal care and other variables were considered as explanatory variables. The T-test and ANOVA were used to assess differences in means. Simple linear regression was used to screen all possible explanatory variables for trend. Multiple linear regression was used to eliminate confounders and to explore the causal risk factors for deviated mean Z-score of birth weight and length.
Results
Six hundred and three (603) mothers were enrolled in this study. Their mean age was 27 years (range 16-45). The mean gestation at delivery was 39.6 weeks (range 27- 43), and the mean number of total ANC visits was 8.2 (range 0-27). The babies weights ranged from 600 to 4740 grams, mean weight was 3155 grams, mean length was 48.02 cm, and mean head circumference was 34.3 cm. The sex distribution showed a female: male ratio of 1.05:1. The mean Z-weight and length were -0.75 and -1.25 respectively. The distribution of the birth weights was bimodal, with deviation of the mean birth weight to the left compared with the standard. This is similar to the findings when we looked at the birth weight of 750 babies in the pilot study. The length curve was also deviated to the left of the standard. Among the 15 explanatory variables examined, gestational age at birth was the strongest predictor of weight and length of the babies. However, all gestational age groups had mean lengths below the reference standard. After gestational age, the level of education of the baby's father was the next strongest predictor of birth weight but had no statistically significant effect on length. Statistically significant differences were found in weight and length between babies who required NICU admission and those who did not, and in weight and length of babies born to mothers who suffered from medical or obstetrical problems during pregnancy. The number of ANC visits bore no relationship to the birth weight or length outcome. In multiple linear regression, parous versus nulliparous mothers and the number of visits to the obstetrical clinic were statistically significant predictors for BWZ but not for BLZ.
Conclusion
Despite the mothers' regular antenatal visits, the babies' birth weights and lengths deviated to the left of the standardized curves. It seems that the numbers of ANC visits are beneficial to a certain point, after which additional visits do not necessary yield improved outcome. The bimodal pattern of the birth weight and length curves may represent two different populations of exposures, or it may represent one group with missing data in between. Missing data (artifact) is unlikely because of the persistence of this bimodal pattern in the birth weight curve of 750 babies born over two years. Our findings showed that gestational age had the strongest effect on both birth weight and birth length. The striking finding in our results is that the BWZ was normal for babies born at 39 and 40 weeks of gestation. Although the husband's education was found to be a constant important predictor affecting the baby's size, it must act as a marker for other factors associated with the mother's daily environment and nutrition. Accurate birth weight measurements should be used as the principal indicator for birth outcome in all health services and should be available in the birth registry office to facilitate both follow-up of changes in birth weight and regular assessment of the effect of ANC and educational programs. ANC allows the health provider to detect and handle problems when they arise. However, in order to conserve health resources for higher-risk patients, uncomplicated pregnancies need not be seen as often as complicated cases. Nulliparous mothers should be given special attention during their antenatal visits. All efforts should be made to prolong the length of gestation. Medical and obstetrical problems should be strictly controlled during pregnancy.