Although the primary responsibility of a DSMB is to help ensure that study participants are not unduly harmed, DSMBs also have a secondary responsibility to help ensure the integrity of the trial. Ensuring a trial's integrity helps promote the likelihood that it will yield valid data that can inform clinical practice and policy decision making (Ellenberg et al 2015; Friedman and Schron 2011).

One key issue for integrity relates to trial adherence, or the degree to which the behavior of trial participants corresponds to their assigned intervention. Issues with protocol adherence can present a range of challenges for a trial's integrity, complicating statistical analyses and confounding the ability to derive scientific conclusions (Robiner 2005).

How, if at all, do these considerations change when applied to a pragmatic clinical trial? Recall that pragmatic trials permit greater flexibility in adherence than do traditional explanatory trials, as their objective is to evaluate the effects of a treatment in a real-world context. Nevertheless, monitoring for adherence is critical in a pragmatic trial for at least 3 reasons. First, data about adherence can inform assessments of treatment effects, including whether a trial that finds no difference between study interventions in the primary outcome is more likely to indicate true equivalence rather than high rates of nonadherence (Ellenberg et al 2015). Second, information regarding adherence to treatment may itself be considered a relevant finding, providing information about the likely use under real-world conditions—such as, for example, if nonadherence results from participants perceiving the cost, convenience, or other aspects of their assigned treatment to be problematic (Ellenberg et al 2015). Third, it is important to understand the impact of potential changes in clinical care within the participating healthcare delivery systems during the conduct of the clinical trial.