RECENT TRENDS IN CANADIAN INFANT-MORTALITY RATES - EFFECT OF CHANGES IN REGISTRATION OF LIVE NEWBORNS WEIGHING LESS-THAN 500 G

Objective: To ascertain whether the increase in the crude infant morta lity rate in Canada in 1993 was due to a recent increase in the regist ration of newborns weighing less than 500 g as live births. Design: Ec ological study, with Poisson regression analysis. Setting: Canada. Sub jects: All live births and infant deaths in Canada between 1987 and 19 93, as reported by Statistics Canada. Data from Newfoundland were excl uded because they were incomplete for 4 years. Outcome measures: Propo rtion of live births by low-birth-weight category; and annual crude an d adjusted infant mortality rates. Results: Over the study period the proportion of newborns weighing less than 500 g registered as live bir ths increased significantly (chi(2) for trend = 71.26, p < 0.01). This trend was an isolated phenomenon rather than a general increase in al l low-birth-weight categories (chi(2) for trend in the proportion of n ewborns weighing 500 to 2400 g registered as live births = 1.14, p = 0 .28). The crude infant mortality rate per 1000 live births decreased f rom 6.4 in 1991 to 6.1 in 1992 and then increased to 6.3 in 1993. Pois son regression analysis revealed that the apparent increase in the inf ant mortality rate was caused by the increased registration of infants weighing less than 500 g as live births. The adjusted infant mortalit y estimate for 1993 was lower than that for 1992. Conclusions: The inc reased infant mortality rate in Canada in 1993 appears to be due to in creased registration of infants weighing less than 500 g as live birth s, Comparisons of infant mortality rates by place and time should be a djusted for the proportion of such live births, especially if the comp arisons involve recent years.