All NIH-funded clinical trials must include appropriate plans for data safety and monitoring that describes research oversight to ensure the safety of research participants and the validity and integrity of the data. The level of monitoring should be commensurate with the risks, size, and complexity of the clinical study. If applicable, the detailed data safety and monitoring (DSM) plan must be approved by both the institutional review board (IRB) that is overseeing the trial and by the funding agency.  An observational study, or a study that does not test an intervention, is not a clinical trial and does not require a DSM plan. In addition to the overall NIH policy for clinical trials monitoring, many of the NIH Institutes and Centers (ICs) have additional IC-specific clinical trials policies.

A common misconception is that an independent board known as a data safety and monitoring board (DSMB) always manages DSM. In fact, monitoring can take many forms. For example, a clinical trial that is low risk and involves only a small number of human subjects might require only close monitoring by the study investigator, whereas a multisite or large clinical trial might necessitate reporting to a central entity, such as the DSMB or an independent safety monitor. For more details, see NIH’s Further Guidance on a Data and Safety Monitoring Plan for Phase I and Phase II trials. A Phase III clinical trial requires a DSMB. The National Institute of Arthritis and Musculoskeletal and Skin Diseases offers examples of appropriate levels of monitoring for different phases and presents details about the composition of a DSMB.

One well-known example of the role of a DSMB involves a Women’s Health Initiative clinical trial of hormone replacement therapy in healthy menopausal women. During the trial, the DSMB concluded that the harms of taking estrogen plus progestin exceeded the benefits. The DSMB then promptly stopped the study. Read more about this incident in a press release from 2002.