Thyroxin Threshold Linked to Impaired Outcome in Preterm Infants : A Retrospective Cohort Study

s of the 26th European Workshop on Neonatology September 2–5, 2018 Cappadocia, Turkey Thyroxin Threshold Linked to Impaired Outcome in Preterm Infants: A Retrospective Cohort Study Jean Michel Hascoet1, Stephanie Coquelet1, Helene Deforge2 1Department of Neonatology, Regional Maternity, CHRU Nancy, France 2EA3450DevAH, University of Lorraine, France Introduction: It remains controversial whether transient hypothyroxinemia of prematurity infl uences shortand long-term outcomes. Thus, we aimed to defi ne a thyroxin threshold associated with neonatal clinical impairment and outcome at 3 years of age. Methods: We retrospectively analyzed medical records of infants born at a gestational age (GA) of < 29 weeks. A thyroxin threshold value was defi ned by ROC curve analysis in a cohort of infants born from 10/2008–12/2012, and validated in a second cohort of infants born from 01/2014– 12/2016 in our institution. Results: Our analysis included 460 patients (mean GA, 26.7 ± 1.3 weeks; mean birth weight, 935 ± 206 g). Thyroxin (FT4) measurements were available for 196/274 infants from the early time period, among whom 35 exhibited neonatal clinical impairment. ROC curve analysis indicated an FT4 threshold of 10 pmol/L, with a sensitivity of 85.7%, and a specifi city of 49.1%. FT4 measurements were available for 176/186 infants from the second time period: neonatal clinical impairment occurred in 20/78 infants with FT4 ≤ 10 pmol/L versus 3/98 with FT4 > 10 pmol/L (P <.001). Three-year follow-up data were available for 147/196 eligible infants. Poor outcome occurred in 65% (58/89) with FT4 ≤ 10 pmol/L versus 34% (20/58) with FT4 > 10 pmol/L (OR, 3.555; 95% CI, 1.774–7.128; P <.001). Conclusion: We defi ned and validated a FT4 threshold of 10 pmol/L as a signifi cant risk factor for neonatal clinical impairment, and a good predictor of poor outcome at 3 years of age.