Definitions of high risk in pregnancy and evaluation of their predictive validity.

The predictive validity of Hobel's criterion (score of 10 or more) for high prenatal and intrapartum risk and two alternative definitions was evaluated in two sociodemographically similar samples from two different institutions. At one institution, intrapartum risk was associated with neonatal morbidity independently of prenatal risk; at the second institution, both were related to neonatal morbidity, depending on the definition of high risk. Definition of high risk as a score of 15 or more improved the prognostic ability of Hobel's method at the first institution, but not at the second. Grouping prenatal and intrapartum risk as low (0 to 9), medium (10 to 19), and high (20 or more) provided a clearer understanding of the association between prenatal/intrapartum risk and neonatal morbidity, although differences between the two institutions persisted. Modifications and validation of risk assessment methods are necessary when used in populations other than the one for which they were originally developed.