This section concerns how to think about an important issue that can arise when studying clinicians' ordinary work.

Traditional explanatory clinical trials are often conducted in specialized research facilities with clinical staff who are trained in research methods. The clinical staff have chosen to be involved in the trial, and often they see themselves as researchers. In contrast, pragmatic clinical trials study the effects of implementing interventions in the real world, and they must be conducted in more or less natural clinical settings using typical clinical personnel. When trials move into the ordinary venues of medical care, one of the issues that arises concerns the duties of the clinical staff to participate (Garland et al 2023).

Do clinicians generally have a duty to participate in pragmatic clinical trials about their usual work? If so, why? And on what grounds may they be excused from this duty?

First, clinicians generally do have a defeasible duty to participate in pragmatic clinical trials about their own work. In this context, "defeasible" means that clinicians may be excused from this duty, but it is their responsibility to provide reasons for this excuse, and possibly to try to mitigate the consequences of their refusal.

The duty to participate has 2 components. First, clinicians have a fiduciary duty to their patients to provide high-quality, evidence-based care. High-quality care requires research about which interventions are best. Because pragmatic clinical trials occur in "naturalistic" clinical settings, the trials will not occur if "ordinary" clinicians (as opposed to specialized research staff) do not participate. Thus, if high-quality care requires pragmatic research, and pragmatic research requires ordinary clinicians to participate, then the duty to provide high-quality care implies a duty to participate in pragmatic research.

This argument goes only so far. It shows there is an imperfect duty to contribute to pragmatic research (one in which there is some discretion about how it can be satisfied), but it does not specify that clinicians ought to participate in pragmatic trials. They could contribute in other ways. Thus, a second step is required. Clinicians have a duty of fairness to each other to participate in pragmatic trials. Because some clinicians must participate in order for there to be a good trial, clinicians who consistently refuse to participate and yet enjoy the benefits of the research are not bearing their fair share of the burdens of pragmatic research. Of course, not every clinician will have an opportunity to participate, but those who can participate ought to. Their circumstances sharpen the duty into more specific requirements, much in the same way that a duty of rescue applies to everyone in general but requires more specific actions of someone in a position to help.

Second, given the duty to participate, a clinician must offer a reasoned excuse to refuse to participate, and not merely express a preference against being involved in the research. However, several reasons to refuse can defeat the general duty in particular cases. Whether the reason to refuse implies an objection to the trial as such is an important distinction.

For example, a clinician may judge that something about the pragmatic trial in question will make patients worse off or otherwise fail to produce the evidence it aims to produce. In general, this objection would be hard to sustain, since clinicians are often not experts on trial design and conduct. Other stakeholders’ opinions on the study may carry more weight. Moreover, trialists seek stakeholder input in the design phase in part because they recognize that local clinicians can offer valuable critiques and context for a proposed intervention. This kind of objection to a study is, at least in principle, the right kind of objection and, if it cannot be resolved, implies that the trial should not occur in its current form. The trial sponsors and ethics committees may still judge that the trial is permissible. But a clinician who sincerely believes the trial is bad might be excused from participating in it.

More common are objections that do not generalize to other clinicians. They are specific to a particular clinician and, by offering them, the clinician does not imply that the trial should not occur. These objections can be either moral or nonmoral. Moral objections are often labeled “conscientious” objections. They should be protected because clinicians do not cease to be moral individuals when they put on their scrubs. Clinicians might be wrong about the moral permissibility of a trial, but if they cannot be persuaded otherwise, it would be inappropriate to require them to participate. For example, a senior clinician who has a strict idea of what informed consent requires might find a pragmatic trial to be too lax in its protections for participants. The trial team might give him several articles making arguments for the permissibility of modified consent practices, and he admits that he cannot refute these arguments. Yet he remains uncomfortable with the modified practice and feels uncertain about whether he can act correctly within the trial as designed. Requiring him to participate anyway would need a strong justification. Conscience is not a trump card, but it does matter.

It is also possible that a clinician may judge that something about their specific situation makes the trial a poor fit in the moment. Perhaps the trial equipment or techniques are too difficult for the clinician to use safely, and these features of the trial cannot be modified. This limitation might be a useful data point for the investigators, in that it may suggest that their intervention cannot be integrated into usual care. However, a clinician who judges that they cannot safely deliver the intervention has at least some reason to refuse to participate in the trial. It is also possible that the additional burdens of a trial would be excessive for a clinician’s current situation. Since many pragmatic trials do not substantially alter ordinary practices, it is less likely that this exception would apply. But it could, and it would not necessarily be wrong for a clinician to point out their own personal limitations in justifying their refusal to participate. Fairness may require that the clinician be open to various accommodations that would ease their burdens and free them to participate in the trial.

Nevertheless, in general, clinicians should participate in pragmatic clinical trials about their own work. The burden of justification rests on those who would refuse. It is possible to satisfy this burden, but clinicians should offer reasons for their refusal that others can accept.For a fuller discussion of clinicians’ duty to participate in pragmatic clinical trials, see Garland et al (2023).