Extended culture and the risk of preterm delivery in singletons: confounding by indication?

Sir,Wereadwithinteresttherecentarticle(Daretal.,2013)suggestingthatsingletonsbornafterextendedculturemayhaveahigherriskofpretermdelivery than those born after Day 3 transfer. While the importance ofoptimizing embryo culture and improving obstetrical outcomes after invitro fertilization should not be underestimated, we believe this studysuffersfromaseriousmethodologicalaw,namely‘confoundingbyindi-cation’. Due to this bias, blastocyst-stage transfer—the most effectivemeansofimprovingselectionandreducingmultiplegestation(Papaniko-laouetal.,2006)—maybeincorrectlyblamedforthissmalldifferenceinpreterm delivery risk (17.2 versus 14.1%).Theauthorshypothesizethat‘extendedculturemaycausedifferencesinimplantationandplacentation’thatpresumablypredisposetopretermdelivery. However, the differences that they attribute to extendedculture presuppose that the two populations (Day 3 transfer and Day5/6 transfer recipients) have the same a priori risk of preterm delivery.In fact,this is unlikely to be the case.Itiswellknownthatelectivesingleembryotransfer(eSET)isunderu-tilized.WhileeSETaccountedforonly4.0%ofcyclesinCanadain2006(Gunbyet al.,2011), 20.2%of theDay5/6deliveriesin this study wereafter SET, compared with only 5.8% in the Day 3 group. Patients whohaveacontraindicationtomultiplegestationduetoapriorpretermde-livery, cervical incompetence, uterine anomalyor medical complicationalmost universally receive eSET and prior research has demonstratedthat, likely due to these underlying factors, SET recipients may be atincreased risk for preterm delivery (Grady et al., 2012). Even adjustingfor the number of embryos transferred, as was performed by Daret al., maynot correct for this sourceof bias.Furthermore, by not including any data on the number of embryostransferred or implantation rates, it is not possible to determine theprobabilityofatermdeliverypercycleinitiated.Patientsengageininfer-tility treatment with the goal of achieving a healthy term delivery. Sincethere are signicantly more deliveries per blastocyst transfer than perDay3 transfer, patients have a higher likelihood of aterm deliveryafterblastocyst transfer. Though the authors of this study and a recentstudy looking at IVF singleton deliveries in the USA (Kalra et al., 2012)take the ‘glass half-empty’ interpretation, an optimist could look atblastocyst transfer as a way to increase term deliveries, especiallywhen single blastocyst transferis employed.In order to properly assess the potential impact of extended cultureon gestational age, future studies must control for the underlying riskof preterm delivery. Otherwise we fear that the concern raised by thisstudy may have the unintended consequence of shifting more patientsto cleavage-stage, multiple embryo transfers with resulting increasedriskof preterm delivery due to multiple gestation.