Classification of twins and neonatal morbidity.

Prematurity is the most significant determinant of neonatal morbidity among twins, but the relative importance of other factors is unclear. We compared the prognostic value of intrauterine growth retardation, discordancy, and other classifications in 131 consecutive sets of surviving twins. A four-factor model for neonatal morbidity was developed from 161 potential outcome measures. Individual evaluation of each twin for intrauterine growth retardation using singleton growth curves was more effective than discordancy and other classifications in predicting neonatal morbidity. Our findings suggest that evaluation of twin growth should concentrate on individual twin growth rather than discordancy. These results also speak against interpreting growth curves based on twins in the same way as growth curves based on singletons; it is apparent that the frequency of growth retardation with associated neonatal morbidity is increased among twins.