Standard Methods Based on Last Menstrual Period Dates Misclassify and Overestimate US Preterm Births

摘要

The US Centers for Disease Control and Prevention uses the last menstrual period (LMP) date to calculate preterm birth (PTB), defined as less than 37 weeks' gestational age (GA). However, previous studies have shown LMP to be inaccurate at classifying GAs at 35 weeks or less, thus increasing economic and clinical burden of prematurity on the population. The obstetric estimate, which is now available on all US birth certificates, is based on multiple prenatal factors, including but not limited to ultrasound measurements, fundal height, first pregnancy symptoms, and fetal maturation, may be a more accurate indicator. The authors of this study were interested in the number of PTBs in the United States based on LMP versus best obstetric estimate. They also sought to determine birth weight and neonatal intensive care unit (NICU) admission data, 2 important outcomes of PTB, from US 2012 Natality Public Use File records and then to correlate these data with recorded obstetric and reported LMP-calculated estimates. Clinical relevance and validity of each method were determined by looking at the risks of NICU admission and low birth weight (LBW) in preterm infants relative to term infants (born at 37-41 weeks), estimated through logistic regression models adjusted for maternal race and age. The authors analyzed the records of 3,736,345 live births. Last menstrual period-calculated estimates showed a greater percentage of births to be preterm than did the obstetric estimates (11.4% vs 9.5%), with the largest difference in births at 32 to 35 weeks' GA (5.4% vs 4.2%). In addition, not only did the LMP dating predict a larger percentage of PTBs, but it also did not correlate as well as the obstetric estimate with NICU admission and LBW. The relative risk of NICU admission in preterm versus term infants identified by obstetric estimate was significantly higher (17.2 vs 10.2). In addition, among those identified as preterm by LMP but term by obstetric estimate, 5.7% were admitted to the NICU, whereas among those identified as preterm by the obstetric estimate but term according to LMP, 25.2% were admitted to the NICU. In terms of LBW (<2500 g), the same trend was seen. Obstetric estimate correlated better than LMP, with relative risk of LBW in preterm versus term infants identified by obstetric estimate being significantly higher (49.3 vs 24.1). Among those preterm by LMP but term by obstetric estimate, 7.7% were LBW. Among those preterm by obstetric estimate but term by LMP, 35.4% were LBW. Overall, based on NICU admission and LBW data, the obstetric GA estimate is a more accurate and clinically relevant indicator of PTB than the LMP dating, which appears to overestimate the incidence of PTBs by approximate to 20%. Sole reliance on LMP dating should not be used.