Because very low birth weight (VLBW) infants (< 1500 grams) are usually also very preterm (< 31 weeks), it is commonly assumed that reporting of outcome for high-risk infants by birth weight is comparable to reporting outcome by gestational age. To identify potential bias in such an assumption, mortality and neurodevelopmental outcome in VLBW infants was compared with very preterm infants. 100% of live born infants who were either VLBW or very preterm born between January 1, 1985 and December 31, 1993 were identified from a defined region (all Nova Scotia except Cumberland county and the 4 counties in Cape Breton) and were followed to at least 2 years in the Perinatal Follow-Up Program of Nova Scotia. Gestational age was obtained as the best estimate from maternal dates, antenatal ultrasound and/or examination of the newborn. Of the 635 infants identified, 397 (62%) overlapped because they were both VLBW and very preterm, 192 (30%) were VLBW but were ≥ 31 weeks and 46 (7%) were ≥ 1500 grams birth but very preterm. The comparison of the VLBW and very preterm infants is shown in the Table below. Outcome by birth weight shows that both mortality and disability rates are lower than that reported by gestational age. This is particularly noticeable at gestational ages < 25 weeks or for birth weights < 1000 grams. For example, the mortality at 24 weeks in this population is 67% but for infants between 700-800 grams (the size of the average 24 week infant in Nova Scotia), the mortality is only 36%.

Table 1 No caption available.
Full size table

Conclusion: By using birth weight specific outcomes rather than gestational age specific outcomes, the potential exists for providing obstetricians or parents misleading information upon which to make clinical decisions.