How are neonatal and maternal outcomes reported in randomised controlled trials (RCTs) in reproductive medicine?

How do randomised controlled trials (RCTs) in reproductive medicine report maternal and neonatal outcomes, specifically singleton live birth? Despite the widespread appeal to use singleton live birth as the outcome measure in subfertility trials, 80% of RCTs fail to do so, and fail to report on neonatal and maternal outcomes. The aim of reproductive medicine is to assist subfertile couples in their wish to have children. A decade ago it was proposed to use singleton live birth as the outcome measure. We assessed whether clinical research has followed this recommendation, and how neonatal/maternal outcomes are reported. A review of the published literature from 1 January 1966 to 31 December 2012 was performed using the Cochrane database. We compared the time periods before and after 2004; the year after ESHRE recommended the use of singleton live birth. We searched the Cochrane database for RCTs in reproductive medicine, and recorded the number of studies that used singleton live birth as the outcome measure. We also recorded the reporting neonatal and maternal outcomes. We identified 910 RCTs that reported on fertility treatments, of which 182 RCTs (20%) reported on singleton live birth [before 2004 96/518 (19%); after 2003 86/392 RCTs (22%)]. Singleton live birth was the primary outcome in 68 RCTs (7.4%). Only 44 RCTs (4.8%) reported on neonatal outcome, while 52 RCTs (5.7%) reported on maternal outcome. We only included Cochrane reviews, thus report here only on the higher quality studies. The actual reporting on maternal and neonatal outcome may even be lower when studies of lower quality are included. Although a decade ago singleton live birth was recommended as the outcome measure of reproductive medicine research, this has not been followed; currently most clinical research in reproductive medicine does not report beyond the occurrence of pregnancy. No funding was received for the study. The authors have no conflicts of interest to declare