We have examined the relationship between the rate of very low-birth-weight deliveries in a population and the neonatal mortality of that population on three ecologic levels: in one hospital over a 12-year span; among the 50 states and the District of Columbia; and among 13 industrialized nations. In each of the three sets of populations the VLBW rate is an excellent predictor of neonatal mortality, accounting for about three-quarters of the variance in the outcome in all of the populations studied. The relatively high neonatal mortality of the United States as compared to that in some other industrialized nations is primarily attributable to its disadvantageous birth-weight distribution. Holding the adverse birth-weight distribution constant, the United States appears to do better than most of these nations in neonatal mortality. The weight distribution of live births in any population is closely linked to indices of social class. Survival of infants at a given birth weight, however, might well be a function of perinatal care. Since weight-specific mortality rates for populations are not widely available, examination of the variance in neonatal mortality rates once the VLBW rate is held constant might be a first step in comparing the quality of medical care for newborn infants among different populations.
PIP: The authors examine the relationship between the rate of very low-birth-weight deliveries in a population and neonatal mortality, using data for one New York hospital over a 12-year period, U.S. data by state, and data for 18 industrialized countries. The results show that the very low-birth-weight rate is an excellent predictor of neonatal mortality, accounting for about three-quarters of the variance in the populations studied. The reasons for variations in neonatal mortality between the United States and other countries and within the United States are discussed