Inconsistency and institutional review boards.

To assess the adequacy of peer review for research on human subjects, identical research protocols in oncology and anesthesiology were submitted to 32 institutional review boards (IRBs) at major universities with medical colleges. Each of the protocols posed serious ethical issues, contained flaws in scientific design, and provided an incomplete consent form. Twenty-two IRBs participated in the investigation, which revealed (1) consistency in the nonapproval of the three protocols, (2) substantial inconsistency among IRBs in the reasons offered in support of similar decisions, and (3) substantial inconsistency in the application of ethical, methodological, and informed-consent standards for individual review boards. This evidence suggests that revision of the protocols to satisfy particular objections would result in approval of flawed investigations.