Objective: Our purpose was to characterize the clinical presentation or laboratory variables predictive of either abruptio placentae or eclampsia in women with severe preeclampsia.
Study design: Prospective collection of perinatal data from 445 consecutively managed women with severe preeclampsia and eclampsia. Univariate analysis was used to determine which of the independent variables were significantly different between the groups (abruptio placentae vs no abruptio placentae; eclampsia vs no eclampsia). Those with significant differences were then entered into multiple logistic regression analysis to determine those characteristics that were independently related to the outcome variable (abruptio placentae or eclampsia). Before multivariate analysis, the independent variables with an interval scale of measurement were converted to a dichotomous scale, with the receiver-operator characteristic curve used to determine a cutoff level.
Results: Univariate analysis revealed statistical significance for the following variables associated with eclampsia: uric acid concentration, > 8.1 mg/dL; proteinuria (>3+); headache; visual symptoms; deep tendon reflexes >3+; serum albumin concentration, <3 mg/dL; and serum creatinine concentration, >1.3 mg/dL. However, with subsequent multivariate analysis, only headache and deep tendon reflexes >3+ remained significant. Univariate analysis for variables associated with abruptio placentae revealed an association between bleeding and platelet count <60,000/mm3. There was no association between abruptio placentae and eclampsia and systolic, diastolic, or mean arterial pressure, quantitative proteinuria, epigastric pain, bleeding, gestational age at delivery, history of preeclampsia, or chronic hypertension.
Conclusion: Quantitative proteinuria and degree of blood pressure elevation were not predictive of either abruptio placentae or eclampsia, as has previously been suggested. The greatest morbidity associated with eclampsia occurred in women with preterm gestations not receiving medical attention.