Among the ways that DSMBs help safeguard the interests of study participants is by monitoring for serious adverse events, so as to ensure that a study intervention does not cause more serious adverse events as compared with the control group.

While the general obligation to safeguard the interests of individuals enrolled in a clinical trial exists regardless of whether a trial is explanatory or pragmatic in nature, several design features of pragmatic trials can influence the procedures for monitoring for serious adverse events (Simon et al 2019). First, unlike explanatory trials involving new therapies, pragmatic trials often involve interventions with previously established safety profiles. Second, pragmatic trials may involve differential contact with study staff by study arm, which in turn may lead to differential reporting of serious adverse events. Third, serious adverse events are often ascertained from healthcare system records, which may delay their identification.

These special circumstances may require negotiation with the DSMB to work out a more practical monitoring plan. Because interventions—even those that are presumed helpful and/or considered standard of care by some—may have unintended consequences in real-world settings, investigators should plan for how any unintended consequences will be detected (Curtis et al 2025). For example, the LIRE study tested the insertion of benchmark prevalence data into routine radiology reports for people who had received lumbar spine imaging (Jarvik et al 2020; Jarvik et al 2015). It was determined that the intervention could reduce detection of rare but potentially dangerous causes of back symptoms, so the team planned for these unintended consequences by having external safety officers monitor for emergency room visits within 90 days and deaths within six months of index imaging (Jarvik et al 2015, Curtis et al 2025).