Ethical approval among studies using the National Health and Nutrition and Examination Survey (NHANES): a cross-sectional analysis

Background: Institutional Review Boards (IRBs) were born out of the unethical treatment of individuals in several studies— the Tuskegee syphilis experiment, the Milgram experiment, and the Stanford prison study, among others. While IRBs are essential for protecting individuals and the integrity of research, local IRBs vary in what they perceive warrants ethical review which can lead to confusion among researchers— particularly about what constitutes human subjects research (HSR). The National Health and Nutritional Examination Survey (NHANES) provides a nationally representative dataset that includes many psychological and medical components used for public research. All components with any individually identifiable information have been removed, which classifies their use as NON-HSR. Given that IRB guidelines are not standardized across institutions and there is a lack of literature concerning ethics in the use of NHANES data, our objectives were to evaluate rates of IRB submissions among cross-sectional studies of publicly available NHANES data and associations with other study characteristics.

Methods: PubMed was queried for cross-sectional studies of NHANES data published after 2011. Next, the articles were randomized and 500 articles were exported to an excel sheet. Data screening and extraction were performed simultaneously by two authors until they independently retained 200 articles. Extracted data included IRB statements, country of the primary author, and if the study was funded. Descriptive statistics were recorded for extracted characteristics and chi-square tests were used to measure associations between IBR submission, study characteristics, and publications occurring before and after the 2019 revisions to HHS45-CFR46, although no changes were made to the definition of non-HSR.

Results: Among 203 included studies, 22 (11.1%) had declared being submitted to an IRB, 26 (13.1%) reported not having undergone IRB review, and 150 (75.8%) studies made no mention of ethics review. Of the 22 studies submitted for review, 6 (27.3%) were published in journals requiring an ethics determination, 7 (31.8%) of the author’s institutions required all studies to be submitted for review, and 6 (27.3%) studies were from universities which had exempted NHANES analysis from IRB submission. We found no significant associations between IRB submission and study funding, whether it originated within or outside the United States, nor among publications occurring before or after 2019.

Discussion: Our findings show the majority of studies using NHANES data did not include an ethics statement, and there was a nearly equal split among declarations of IRB submission and nonsubmission. Journal and institutional requirements were infrequent among studies that were submitted to an IRB, which may constitute the need for national standards for IRBs concerning federal datasets, such as NHANES. Accessibility to these datasets can provide researchers cost-effective avenues to test hypotheses which can provide critical preliminary data to support future clinical research and reduce research waste. Thus, public accessibility to NHANES, without unnecessary institutional or journal requirements is essential to maximize their benefit. Moreover, institutions must evaluate their policies that prove to be barriers to secondary data analysis.