Aim
To determine the rate of nonpublication and discontinuation of randomised controlled trials (RCTs) in newborns.
Methods
This was a retrospective, cross‐sectional study of RCTs registered with the Australian and New Zealand Clinical Trials Registry (ANZCTR) between 2008 and 2012.
Results
Fifty trials were identified, of which 23 (46%) were retrospectively registered. Thirty trials (60%) were published. After a median follow‐up of 8.0 (range 4.6–17.4) years from Research Ethics Committee approval, 15 of 41 completed trials (37%) remained unpublished, representing 5422 neonatal trial participants. Nine trials (18%) were discontinued, including four that were published. The most frequent reason for discontinuation was poor recruitment (n = 4). Sample size discrepancies between registration and publication were found in 17 (65%) of the 26 completed, published trials. In nine (35%) of these trials, the calculated sample size in the method section of the published article differed from the planned sample size in the trial registry (relative difference −20% to +33%).
Conclusion
Nonpublication and discontinuation of RCTs conducted in newborns is common. Additional efforts are needed to minimise the number of neonatal trial participants that are exposed to interventions without subsequent publication.