In The News: Many Perinatal and Neonatal Studies Have Incomplete Data, Review Finds

Healio reports on a study of outcome data from randomized controlled perinatal and neonatal studies. Researchers used PubMed for such trials published between January 1, 2020 and December 31, 2021. The search was limited to high profile journals along with a number of pediatric specialty journals. Of the 87 trials reviewed, they found nearly 90% report incomplete primary outcome data.

The study authors note their analysis looked at the missing data as potentially jeopardizing comparability between groups and compromising inferences about treatment effects. An accompanying editorial went further and referred to the missing data as a threat to the validity of trial results as well as the health and well-being of children. As for the reasons for the incomplete data, study authors included parents or guardians of participating infants not responding to questions, failure to adhere to trial protocols, requesting early withdrawal from protocols, or refusal to continue data collection.

Ultimately, the conclusion is that the handling of missing outcome data was inadequate. Researchers offered some strategies to address trials that also suffer from the lack of complete data sets. Those include appropriate estimation of sample size, different imputation methods, and ways of dealing with study withdrawal. In the editorial, authors note there must be greater advocacy for resources needed to complete randomized trials in pediatric populations.

Study authors noted whether a trial was missing data, method used for primary analysis, and whether sensitivity analysis was performed to account for missing data. Researchers found 89% of reviewed trials had incomplete primary data, with 12% not addressing the missing data in any way. In the studies, 79% restricted main analysis to participants with complete information. Sensitivity analysis was performed by 49% of researchers.

PIDS member and JPIDS Editor-in-Chief, Ravi Jhaveri, commented on the article. “These findings highlight the unique challenges of neonatal and pediatric research, where long-term health outcomes are likely more important but also more challenges to obtain.  While the investigators are right that other investigators should be aware, those reviewing and curating the medical literature should be prioritizing papers that include proper outcome measures and sensitivity analyses that include a broad range of variables that may be seen in real-world settings.”