To assess variation in IRB review, we evaluated the response of 18 institutions to the same NICHD Neonatal Research Network protocol, after critique by external reviewers and approval by the Advisory Board and all neonatal divisions. This ongoing trial, subsequently approved at all sites without modification to the protocol, assesses whether vitamin A (vit A) supplementation in a dose higher than in prior trials prevents chronic lung disease and sepsis in extremely low birthweight infants. It involves i.m. or sham injection of vit A, vit A blood levels in 30% of the 800 subjects, and weekly exams by masked investigators. Each IRB's first written response was assessed. Fifteen IRBs requested consent form modifications. Although 9 IRBs had no concerns about the protocol or minor requests for clarification, 9 had major concerns related to trial design, all involving controversial questions about the proper conduct of clinical research, including:Table We conclude that IRB reviews vary considerably, in large part because of unresolved questions about the design of ethical clinical research.

Table 1