Perinatal death in a term fetal growth restriction randomized controlled trial: the paradox of prior risk and consent

BACKGROUND: The disproportionate intrauterine growth intervention trial at term was an intention to treat analysis and compared labor induction with expectant monitoring in pregnancies complicated by fetal growth restriction at term and showed equivalence for neonatal outcomes.OBJECTIVE: To evaluate trial participation bias and to examine the generalizability of the results of an obstetrical randomized trial.STUDY DESIGN: We used data from participants and nonparticipants of a randomized controlled trial-the disproportionate intrauterine growth intervention trial at term (n=1116)-to perform a secondary analysis. This study compared induction of labor and expectant management in women with term growth restriction. Data were collected in the same manner for both groups. Baseline characteristics and neonatal and maternal outcomes were compared. The primary outcome was a composite measure of adverse neonatal outcome. Secondary outcomes were delivery by cesarean delivery and instrumental vaginal delivery; length of stay in the neonatal intensive care, neonatal ward, and the maternal hospital; and maternal morbidity.RESULTS: Nonparticipants were older, had a lower body mass index, had a higher level of education, smoked less, and preferred expectant management. The time between study inclusion and labor onset was shorter in participants than in nonparticipants. Notably, 4 perinatal deaths occurred among nonparticipants and none among participants. Among nonparticipants, there were more children born with a birthweight below the third centile. The nonparticipants who had expectant management were monitored less frequently than the participants in both the intervention and the expectant arm.CONCLUSION: We found less favorable outcomes and more perinatal deaths in nonparticipants. Protocol-driven management, differences between participants and nonparticipants, or the fact that nonparticipants had a preference for expectant management might explain the findings.