Standard methods based on last menstrual period dates misclassify and overestimate US preterm births

To compare number of US preterm births based on obstetric versus last menstrual period (LMP) estimates and evaluate their correlations with clinical risk indicators associated with prematurity.

Preterm births were assessed from LMP, per standard practice, and, separately, from obstetric estimates using the 2012 Natality Public Use File. Percentages of infants with neonatal intensive care unit (NICU) admission and low birth weight (LBW) were calculated.

More births were <37 weeks gestational age (GA) by reported LMP (11.4%) versus obstetric estimates (9.5%). Among infants preterm by LMP, but born at 37-41 weeks by obstetric estimates, there were 5.7% NICU admission and 7.7% LBW rates versus 25.2% and 35.4%, respectively, of those preterm by obstetric estimates but born 37-41 weeks by LMP assessments.

Obstetric estimates provide the most clinically relevant estimates of US preterm births. Assessments calculated from LMP alone may overestimate prematurity incidence by ~20%.