Clinical assessment of gestational age in the newborn infant. An evaluation of two methods.

The scoring systems of Dubowitz et al. and Parkin et al. were evaluated in two selected materials of newborn infants referred to a neonatal unit. Our estimation of gestational age by Dubowitz scores tended to be too high for the extreme prematures. "Small-for-dates" were also overestimated, whilst "appropriate-for-dates" and infants with respiratory difficulties were underestimated. "Large-for-dates" fell close to the standard curve. Ninety-five per cent confidence limits were up to +/- 5 weeks for Dubowitz scores and nearly +/- 6 weeks for Parkin scores. Further statistical analysis displayed some limitations in the use of linear regression formulas for scoring systems based on external and/or neurological characteristics. Thus, the results obtained with these methods must be used with caution in some selected newborn materials.