On the Use of Multiple Imputation Approach in Pediatric Research

In a recent study from Salmon et al reported in The Journal, the investigators assessed the association between clinical chorioamnionitis and neurodevelopmental disorders at 5 years of age in children born preterm. 1 Salmon F. Kayem G. Maisonneuve E. Foix-L'Hélias L. Benhammou V. Kaminski M. Marchand-Martin L. Kana G. Subtil D. Lorthe E. Ancel P.Y. Clinical Chorioamnionitis and Neurodevelopment at 5 Years of Age in Children Born Preterm: The EPIPAGE-2 Cohort Study. The Journal of Pediatrics. 2024 Apr 1; 267113921 Abstract Full Text Full Text PDF Scopus (0) Google Scholar The authors did not find a significant association between the two variables based on EPIPAGE 2, a national, population-based cohort study of children born before 35 weeks of gestation in France in 2011. 1 Salmon F. Kayem G. Maisonneuve E. Foix-L'Hélias L. Benhammou V. Kaminski M. Marchand-Martin L. Kana G. Subtil D. Lorthe E. Ancel P.Y. Clinical Chorioamnionitis and Neurodevelopment at 5 Years of Age in Children Born Preterm: The EPIPAGE-2 Cohort Study. The Journal of Pediatrics. 2024 Apr 1; 267113921 Abstract Full Text Full Text PDF Scopus (0) Google Scholar In the database used by the authors, there is a high missing rate (41%) of the primary outcome variable 'neurodevelopmental disorders' and high missing rates (between 31.9% and 49.4%) of the secondary outcome variables (eg 'cerebral palsy at 5 years,' 'coordination disorders,' 'cognitive impairments,' and 'behavioral Difficulties'). Rather than ignore the incomplete records the authors conducted statistical analyses after multiple imputation (MI) procedures were used, which has been regarded as an effective method to reduce nonresponse bias.