Deliveries at 34 ^0/7 to 36 ^6/7 weeks, by definition, are late preterm births (LPB). In 2007, of 4,317,119 births in the United States, about 9% (383,792) were LPB; of all births before 37 weeks, almost three-quarters were LPB. Between 1996 and 2006, there has been a 20% increase in the rate of LPB. This increase is alarming because compared with term newborn infants, LPB neonates are at increased risk of morbidity during their perinatal hospitalization (respiratory distress syndrome [RDS], temperature instability), of readmission (hyperbilirubinemia, feeding difficulties) and, most importantly, of neonatal and infant mortality. Thus, it is understandable why there is an impetus to decrease the rate of LPB.

It might be convenient to accept and encourage a reduction in LPB but it is prudent to acknowledge that current evidence-based knowledge does not allow a safe reduction. There are at least 4 reasons why the rate of LPB is irreducible. First, there are no ACOG practice bulletins or committee opinions on how to decrease the rate. Even the report from the workshop on this topic, did not provide details on how to lower the rate. The lack of recommendations means there is insufficient knowledge. Second, E = mc or, stated differently, the Etiology of LPB has Multiple Complex Causes. According to McIntire and Leveno approximately 80% of LPB were attributed to idiopathic LPB or ruptured membranes and 20% to obstetric complications like hypertension, abruption, previa, or fetal compromise. The heterogeneity of the causes precludes a simple preventive strategy or randomized clinical trial (RCT). Third, the sample size for a RCT to decrease 1 complication is daunting. As noted by McIntire and Leveno, about 16,000 women need randomization to show a 33% reduction in RDS, if this complication occurs in 1.4%. Fourth, though the rate of preterm birth and LPB has increased, the rate of stillbirth and perinatal mortality for these newborn infants has actually decreased.

In summary, when the cause has multiple complex causes, the rate is difficult to influence. In addition, there is no objective evidence that a decrease is possible, without concomitantly increasing the mortality rate.