Patient Notification About Pragmatic Clinical Trials Conducted with a Waiver of Consent: A Qualitative Study

Abstract

Background:

Some scholars have proposed that investigators and health systems should notify patients about their enrollment in pragmatic clinical trials (PCTs) conducted with a waiver of consent. However, others argue that decision-making about notification requires judgment, and reports suggest considerable heterogeneity about whether, when, and how individuals enrolled in PCTs with a waiver of consent are notified about that enrollment. Empirical data can inform this decision-making.

Methods:

We conducted semi-structured interviews with knowledgeable stakeholders involved in conducting and/or overseeing PCT(s) conducted with waivers of consent, including investigators, those charged with the oversight of human subjects research, and operational leadership. Interviews were conducted via video conference from September to December 2024 and were audio-recorded and professionally transcribed. Data were qualitatively analyzed using an integrated approach, including both a priori codes drawn from the interview guide and emergent, inductive codes.

Results:

Twenty-three of 28 experts invited to participate completed interviews. Respondents described rationales both for and against notification. Rationales for notification included both appeals to moral values (respect for persons, respect for autonomy, and transparency), as well as instrumental goals (promoting understanding of and/or support for research, avoiding downstream surprise, and supporting buy-in). Rationales against notification included preserving scientific validity, perceiving notification to lack value, and concerns that notification might be burdensome for patient-subjects or undermine trust and/or clinical or public health goals. Decision-making about notification was context-specific, and reflected features related to the study design, the health system setting, the patient population, the clinical condition, and the intervention(s) being evaluated. While some factors were consistently described as weighing against notification, including scientific validity or decisions for which a patient would not be offered a choice outside the research context, other factors resulted in divergent decisions across different PCTs (or even across different sites for the same trial).

Conclusions:

While several rationales support notification about enrollment in PCTs conducted with waivers of consent, the relative value and practicability of notification is context-dependent. Some features, such as the need to preserve scientific validity, may appropriately weigh in favor of forgoing notification. However, evidence of divergent decision-making for similar trials suggests the need for a framework to guide future notification decisions. These data can be an important input to inform future framework development.

Introduction

Pragmatic clinical trials (PCTs) aim to efficiently generate evidence to inform real-world treatment decisions. Through embedding research into settings where patients usually receive care, obtaining representative populations through broad eligibility criteria, and randomizing between interventions that reflect treatments that patients would receive in usual care, PCTs can close well-known evidentiary gaps—and, ideally, ultimately improve care and patient outcomes.

Most PCTs—like clinical trials more generally—require researchers to obtain informed consent from research participants. However, a sizable number of PCTs (>8%) are conducted with a waiver of informed consent,^1,2 and the number of PCTs in which consent is not obtained appears to be increasing.² Examples of large-scale PCTs conducted with a waiver of consent include trials comparing different decolonization strategies to reduce healthcare associated infections,^3,4 the effectiveness of recombinant versus standard-dose influenza vaccines,⁵ and antisepsis solutions before surgical repair of fractures, among others.⁶

In the United States, federal regulations permit waivers or alterations of consent when a study meets five criteria, among them being that the research: involves no more than minimal risk, could not practicably be carried out without the waiver or alteration, will not adversely affect subjects’ rights and welfare, and, that “whenever appropriate,” subjects be provided with “additional pertinent information after participation.”⁷ Criteria for waivers or modifications in ethical guidelines from the Council for International Organizations of Medical Sciences (CIOMS) include similar provisions related to impracticability and minimal risk, but also require that the research have “important social value.”⁸

Considerable scholarship has explored the ethical appropriateness of and attitudes towards waivers of consent for PCTs.^9–21 Yet even when consent is waived, a decision must be made about whether to notify those enrolled about the research activity, and, if so, whether to provide any additional information, including: the study rationale, the use of randomization, the specific intervention to which a given individual was randomized, and the use of data and corresponding data protections. Example notification approaches for PCTs include patient information sheets, leaflets or fliers placed in patient care areas, or individual discussions with clinicians.^22,23 Several scholars have proposed that investigators and health systems should routinely notify patients about PCTs and related research activities underway within their health systems—including, and perhaps especially, those conducted with a waiver of consent.^23,24 Yet, as Asch and colleagues have observed, “although a presumption of disclosure is reasonable, calibrating the extent and form of disclosure to the nature of the activity...requires judgment and merits further study.”²³ This manuscript informs this calibration, reporting on the results of interviews with key stakeholders regarding decision-making about whether, why, and how to notify patients of their participation in PCTs conducted with a waiver of consent.

Methods

We conducted in-depth interviews with knowledgeable stakeholders involved in conducting and/or overseeing PCT(s) operating with waivers or alterations of consent. We used a purposive sampling approach to recruit individuals with diversity in role type (e.g., investigators, institutional review board (IRB) or human research protection program (HRPP) leaders, and operational leaders), and health system type (e.g., academic, integrated delivery system, and community clinics). We also sought to reflect diversity in the clinical conditions or contexts studied in the PCTs (e.g., acute v routine care; adult versus neonatal). Individuals were identified from two sources: (a) a review of PCTs conducted under a waiver of consent published within two leading medical journals,¹ and (b) PCTs funded by the NIH Pragmatic Trials Collaboratory and the Patient Centered Outcomes Research Institute. Participants were offered a $100 gift card. We report the study using the SRQR reporting guidelines (Supplementary material).²⁵

We conducted and recorded interviews via video conferencing from September to December 2024. We used a semi-structured format, combining scripted questions developed through review of the PCT literature and spontaneous follow-up probes (see supplementary materials for interview guide). Topics focused upon two main themes: (1) notification about PCTs conducted with a waiver of consent (the focus of this manuscript) and (2) sharing aggregate results from PCTs. All interviews were conducted by SM, a bioethics scholar with expertise in qualitative methods. The Johns Hopkins Bloomberg School of Public Health IRB determined these interviews did not constitute human subjects research. Individuals provided verbal consent to the interview and to recording. Interviews were conducted until reaching code saturation, or when additional interviews no longer presented new themes.²⁶

Interviews were professionally transcribed, and coding was managed using Dedoose software.²⁷ We used an integrative approach to coding, including both a priori codes drawn from the main domains of our interview guide and emergent codes derived from the transcript text. One author (SM) developed the initial codebook. Two authors (SM and AB) then independently coded the same subset of transcripts and met to compare their coding, resolve any discrepancies, and add themes not previously captured. The remaining transcripts were coded by a single author (AB) and then reviewed by a second (SM). Two authors (AB and SM) then developed analytical memos for codes most pertinent to our research question, identifying relevant subthemes and illustrative quotations. Quotations were edited to improve clarity and to deidentify studies or institutions.

Results

We invited 28 experts to participate; five did not respond to our invitations, and 23 completed interviews. Interviews averaged 47 minutes. Investigators were the most common role type (11 respondents), followed by IRB/HRPP experts (7) and health system/operational leaders (4). Respondents often served in multiple roles (e.g., physician-investigator and operational leader); we report here based on their primary role. A total of 22 respondents were based in the United States and one in Canada.

Expert respondents offered insights related to (a) the rationales for and against notification, (b) factors influencing notification decisions, (c) the influence of logistical factors on notification mode, and (d) notification decision-makers.

Rationales for (not) notifying

Respondents offered a range of reasons for notification, from which we identified at least six broad types (Table 1). First, respondents described notification as consistent with the principle of respect for persons, including respondents’ belief that individuals would prefer to be notified. Occasionally this included respect for autonomy, to enable individuals to opt-out of the intervention and/or the use of their data for research. A second rationale was an appeal to (other) moral values, including both statements specifically presenting transparency as an ethical good, and more general sentiments about moral value, such as one health system leader’s characterization of a decision about notification as motivated by “gut check” that notifying was the morally correct option. Third was to promote understanding of (and ideally, support for) research. Fourth was to promote trust in researchers and the health system (and the trustworthiness of those actors). Fifth, and relatedly, was to “avoid downstream surprise,” lest individuals learn about their research enrollment through other means, such as a media report, and corresponding concerns for reputational or other harms. Sixth, specific to notification of clinicians as participants, was the rationale that notification was necessary to support buy-in for and/or implementation of research, such as via adherence to randomization.

Table 1.

Reasons to Notify

Respect for Persons	There’s a difference between disclosure and permission. Just knowing about it is really a matter of respect... It [notifying] gives people agency...to say “I’m not happy that you did this and I don’t want to be a part of this.”
Appeal to Moral Values	A lot of researchers feel icky about doing [research] without telling people about it. If they can tell people about it in some way, they perhaps feel there’s some obligation of transparency or that it’s better to be transparent. I think that more than anything it was just a gut check...it just seemed somehow not completely kosher that this might be going on and a patient just being completely unaware...
Promote Understanding of (and Support for) Research	I think that there is value to patients understanding that when they come into a place that’s an academic medical center, this is a setting in which we’re constantly learning. It’s a learning healthcare system. We’re always looking at how to improve things, what works better, what doesn’t work better. I think there is value in talking to the community, so that they understand this is the process.
Promote Trust in Research(ers)	There’s a whole host of reasons why we should be [providing notification]. Trust for me is the big one. It’s not the self-determination. It’s trust in research and healthcare systems and the clinician in front of me. I think that’s critical here.
Avoid Downstream Surprise	I wouldn’t want people to pick up the newspaper and say, “Oh, there’s a wonderful new study that’s found this thing, and it was done in my hospital, and I didn’t know anything about it.” You’d much rather tell somebody up front than have something hit a newspaper and have them find out that way.
Support Buy-in for or Implementation of Research	[Notification of clinicians] was mostly for adoption. Clinicians have to choose whether or not to trust this prompt [in the EHR]...it was mainly part of the messaging of the trial to try to get engagement as much as it was to ‘notify.’

When considering reasons not to notify (Table 2), the most common offered by respondents was scientific validity, such as if notification were to change behavior and undermine the study. A second set of reasons related to a perceived lack of value, due to (a) lack of actionability (such as for cluster-level interventions for which individuals could not readily decline participation); (b) minimal opt-out decisions (interpreted by individuals as reflecting negligible autonomy gains from notification despite potentially substantial logistical burden); or (c) failure to meaningfully inform (either because notifications are “routinely ignored” or involve information about clinical decisions perceived as overly complex for a non-specialty audience). A third reason not to notify was that notification could be burdensome for patient-subjects (and/or families), particularly for trials in acute or critical care settings. Fourth, notification might undermine trust, including either in clinical competency, should information about the motivation for research undermine confidence in the knowledge or skills of the clinical team, or in research(ers). Finally, notification might undermine clinical or public health goals, were individuals to be so put off by unconsented research as to avoid future receipt of healthcare services.

Table 2.

Reasons to Not Notify

Preserve Scientific Validity	I think in terms of the upfront reasons, when disclosure could change the science is the slam dunk case [for not disclosing].
Information Lacks Actionability	We’re doing cluster randomized trials...the alternative [to study participation] is, “go to a different hospital,” which is rarely an option that’s available to individuals, particularly [for someone in emergent or acute contexts]. In fact, for the notification I showed you, there was an early draft that said, “if you don’t agree with this, you could go to a different hospital.” That was wisely struck from the text.
Few Individuals Opt Out	...almost nobody opts out. It’s not that we want to deny people information, it’s just that it’s not so easy to get it to people reliably and confirm that you’ve done it and that they received it...It adds one more layer to the 200 things you have to do to get a trial started. My enthusiasm is not very high because almost nobody opts out, and I don’t think it’s actually very meaningful or helpful.
Information Fails to Meaningfully Inform	Look, I’ll confess to not being absolutely confident. I don’t think those flyers made any difference to anyone. I think they were routinely ignored. I don’t really think they serve a useful purpose. Going to the community so that they understand the trial is pointless...that’s asking to confuse patients...
Information as Burdensome for Patients/Families	I am opposed to notification in situations where it adds to what is already an overwhelming burden of information on patients and families, particularly in ICU settings.
Avoid Undermining Trust (in clinical care)	I think it’s very hard and often disturbing if you walked into a hospital and the 17 banners on the front, instead of saying, “Top 10 hospital, come get your care here,” said, “We don’t know 90% of what we’re doing. If you come here with a heart attack, 97% of the things we do to you have been unproven. If you come here for diabetes, 67% of the things that we do to you are unknown.” Is that what we want to say when you walk into a hospital?
Avoid Undermining Trust (in research/ers)	When we first proposed this to our IRB, our IRB explicitly came back and said, “Absolutely not. Don’t tell people that we’ve given a waiver, that we put them in research,” because they thought people would be mad and upset and would go to the press and say, “[institution name] is enrolling people in research,” and that sort of stuff.
Avoid Undermining Clinical or Public Health Goals	If we scare patients off by overwhelming them with technicalities that they may not necessarily understand, or something that may not--never even apply to them, the potential downside is that you scare them off from the whole process. It makes them paranoid about being in the hospital that they’re going to be experimented on.

Several respondents expressed ambivalence about notification, either in general or for particular studies, characterizing notification decision-making as “hard” and involving “competing factors;” a few suggested that others might reasonably have made different decisions.

Factors influencing notification decisions

Decision-making about notification was context-specific, and reflected features related to the study design, the health system setting, the patient population, the clinical condition, and the intervention(s) being evaluated (Table 3).

“Every study is going to be different and every patient population is going to be different or perhaps have different expectations for how they are interacted with in both the medical and research setting.” (IRB/HRPP)

“I think it varies across studies also, but even in the same study, we have different ways that we’re notifying or not notifying at all participants based off of the site, which ...seems very arbitrary and weird to me... the way we notify participants and/or what that alteration is varies among the sites, even within the same trial approved by the same IRB, single IRB.” (IRB/HRPP)

Table 3.

Factors Shaping Notification Decisions

Impact on Scientific Validity	[Notification] wouldn’t be reflective of what we’re actually trying to understand, which is what [outcome would we get from providing this intervention in a real-world context]... It’s hard, maybe not impossible to do that with some sort of notification.
Actionability of Information	Our argument was that these are critically ill patients who aren’t choosing, not only aren’t choosing which ICU they get admitted to because they’re typically brought in by ambulance to the local facility, admitted through the ER, but no one ever gets to choose [study intervention]. Our perspective was that in this context, it was not helpful and potentially actually harmful to well-being of critically ill people if they had capacity and/or their family members... because even under usual care, they don’t have options of how their care is staffed. Anyway, the IRB agreed with that and there was no notification whatsoever.
Whether Intervention Involved a Choice that Patients Would Be Offered Outside the Research Context	The context was one in which patients and family members would, under usual circumstances, have some say in what would happen. Is this something that a patient normally has agency about? Do they usually have oversight? Do they have choice? ... They don’t usually have that type of choice, nor necessarily should they, to the general operations of how these things go.
Research Involving Vulnerable or Historically Marginalized Populations	When you talk about children being participants of research, that’s the most vulnerable of vulnerable populations. Then you get into premature neonates, that takes it up another level...from my perspective, it’s really important to be transparent with the parents of those children about how their child is being treated and cared for. The population we serve is overly representative of historically marginalized people and people in whom experimentation without consent or without prior knowledge has occurred... If this were a very affluent, empowered group of patients, a privileged group of patients, would we have felt the same? I don’t know, but that’s not the reality of who we treat.
Studies in Critical Care or other “High Stakes” Clinical Contexts	More generally, what I hear from people just having family members who are critically ill is that they aren’t really in receive mode. People talk to them, and they understand the words, and they don’t sink in. In those situations, I think we want to be careful about just adding to the overload of information that just may not be something they really need to hear. If the intervention is treating something that is high-risk, medically complex, then the IRB would typically push as hard as possible for including an information sheet about it, just because the thought there is that since patients were dealing with a very complicated medical decision, then they wouldn’t want to feel left out of at least being aware that they were part included in a research study.
Logistical Burden of Notification for Researchers or Clinical/Operational Systems	I probably understated the pragmatic barriers...there’s quite a lot of barriers to what we do [that would be exacerbated by adding notification]...all of these communications that go out to patients all have to be vetted by marketing for brand and voice. They have to be vetted by the security office to make sure we’re not communicating inappropriate information like PHI...everything is closely scrutinized and then even legal has to look at these sometimes...if on top of that we want to start considering additional notifications that they have been part of a trial, that has to go through the whole vetting system for that, too...none of this is insurmountable as far as those practical hurdles, but [health system partners] just want to be very careful and they’d probably ask us “what are you hoping to achieve?”
Extent to Which Study Introduces Risk or Otherwise Alters What Would Be Experienced in Clinical Care	It depends on the impact to the patient....another project we have [compares two common approaches for a condition]. The patients in both arms are receiving care. It might be a little bit different, but they’re both going to get two approved cares. We’re looking at the differences. There we’re not telling the patients because we don’t feel like it’s impacting their outcomes or them significantly, where it could be something that they might have an issue with.
Patient Expectations	The culture at all of these clinics is research, research, research. Patients are quite aware of it... It’s expected. It’s also pretty explicit in their forms. When you become a patient [within that health system], they talk it through. I felt okay that putting a flyer was sufficient given the culture of the institution.

This heterogeneity notwithstanding, some factors were viewed as consistently weighing against notification. Among these included scientific validity and perceived lack of actionability, described earlier as rationales against notification. An additional factor was interventions for which patients (or their proxies) would not routinely be offered a decision outside the research context, such as the choice between different strategies for infection prevention in an emergency department, or between different types of barrier precautions worn by healthcare personnel in intensive care units.

Some factors, while raised infrequently, were consistently presented as weighing in favor of notification. These were generally patient-specific factors related to populations requiring special ethical consideration, including research with neonates or with historically marginalized populations.

Conversely, other factors resulted in divergent decisions across different PCTs. When considering studies in critical care or other “high stakes” clinical contexts, some respondents described the clinical context as being so burdensome for patients and families as to weigh against notification. For these respondents, information about research was at best an unnecessary distraction, adding to an already overwhelming volume of information during a high-stress event. Some even expressed concern that such notifications could be harmful, causing emotional distress and raising (inappropriate) concern about the motivations of the clinical team. However, other respondents described similar contexts as motivating the importance of notification, given the “higher stakes” nature, such as for studies involving mortality or severe morbidity as an endpoint. As an IRB/HRPP leader explained, “the high-risk situation is something that sways us more towards informing, because the risk of something bad happening, in general, for the patient is higher.”

Another factor resulting in divergent decisions was the burden of notification for researchers or clinical/operational systems. Some respondents described these burdens as motivating against notification, given opportunity costs that could reduce the feasibility of generating socially valuable data. Others, however, viewed logistical burdens as insufficient grounds to justify forgoing notification. Additionally, studies described as presenting little incremental risk or involving minimal alteration from usual clinical care were sometimes, but not consistently, described as not requiring notification.

Finally, a few respondents cited patient expectations as shaping notification decisions. A repeated theme was the belief that patients at academic medical centers expect (or should expect) that research is constantly underway within their health systems, rendering study-specific notification unnecessary. Notably, the belief that patients seeking care from other systems, such as community hospitals, would not expect research to be underway was also used as a justification against notification, out of concern that notification might undermine patient trust. When asked directly about patient expectations, several respondents expressed uncertainty, with a few calling out the potential value of future empirical work exploring patient preferences regarding notification. As one investigator noted, “I think we’re making a lot of assumptions without knowing.”

Logistical factors shape mode of notification

Respondents identified diverse logistical considerations that shaped decisions about the notification approach, including appropriately targeting the notification to both promote awareness among those enrolled and avoid unnecessary messaging to those ineligible; viewing technology as a means to support (or impede) accessibility of information, depending on the patient population; health system policies regarding patient communications; and the need to minimize burden on ongoing clinical and operational activities (Table 4). These considerations ultimately led to notification being provided via diverse modalities reflecting study- and health-system-specific considerations. Example modalities for patient notification included individual handouts, signage in relevant clinical units, patient portal messages, and mailed letters.

Table 4.

Logistical Considerations Inform Decision-Making about the Mode of Notification

Appropriately Targeting Information to (Only) Those Enrolled	The IRB’s first suggestion was to use posters for notification. I came back and said, I don’t feel comfortable with that. I prefer to [use individually targeted information sheets] because then I can make sure that families at least got the information. I can’t ever verify that [families of those enrolled] ever walked by the poster and stopped and looked at it. Also, this burden of the other 130 [non-enrolled patients] who weren’t in the study at any given time.
Technology as Supporting (or Impeding) Notification	I think that the technology allows to do this [provide notification] in a better way now. It is much more useful...I think that we’re talking about what we did back then and where we are now, it’s just completely different as far as some of these things in notification... I’m 37 years old. I know exactly what [a QR code] means. I scan it with my phone and I know that I’m going to learn something. An 80-year-old is going to look at that and be like, “I don’t really care what that is.” Is that then an appropriate way to inform that individual?
Health System Policies Regarding Patient Communications	Our institution doesn’t allow flyers anywhere in the building, any building, so it makes it a little bit difficult to do that. I think it comes, actually, from corporate compliance. There are also guidelines developed there for, if you [notify via] MyChart, how long should the MyChart message be? It cannot be more than four paragraphs. It needs to include this text, this text, that text...that MyChart message gets reviewed along with every other document provided for that research study, and then they enforce what the compliance committee had recommended...
Minimizing Burden on Ongoing Clinical and Operational Activities	You only have so much wall space, you only have so much brain space. You can’t have signs everywhere. The most important sign is the direction sign—”how do I get where I’m going in a hospital?”

Decision-makers regarding notification

The relevant decision-makers regarding notification varied across PCTs. For some trials, the initial impetus for notification was initiated by investigators, whereas in others it was encouraged by the IRB or via an external ethics consultation process. While several respondents mentioned IRBs as generally supportive of notification, IRB support for notification was not universal. One respondent described their IRB as initially resistant towards an investigator-led notification proposal, fearing that notification about unconsented research might foster patient anger and distrust. Another investigator described an IRB as initially supporting the investigators’ plan to offer notification, but then agreed instead to forgo notification after the participating health systems rejected the proposal to notify based on concern that patients would object to unconsented research.

For studies that used some form of notification, decisions about notification mode typically involved collaboration or deliberation across several groups, including investigators, IRBs, clinical and operational leaders, legal and/or compliance, and patient advocates. Sometimes this was as straightforward as investigators proposing a flier that was signed off on by the IRB. Other PCTs required a more complex process, involving knowledge-sharing between investigators and diverse clinical partners about operational processes that could influence the feasibility of notification and the likelihood of information reaching the target audience, such as health system restrictions on whether or where fliers could be posted.

Discussion

This interview study explored the views of those responsible for the conduct or oversight of PCTs conducted with a waiver of consent about decision-making to offer notification about the study to those enrolled. We identified a few areas of consensus and several differences across experts regarding both the ethical desirability and the potential outcomes of notification. Respondents generally agreed that preserving scientific validity may justify forgoing study-specific notification of participation. Concern for scientific validity is arguably most relevant for studies in which the intervention is behavioral, or the outcome is subjective and patient-reported. However, perspectives about the relevance of other factors for decisions about notification varied across respondents, as did the process for deciding whether notification was required—and if so, in what form.

Some reasons identified by respondents as supporting notification reflect support for viewing informed consent as not merely a means to respect autonomous choice about research participation, but as a process that serves a diverse range of functions.²⁸ Functions of consent that remain relevant for decision-making about notification of research conducted with a waiver of consent include providing transparency, promoting trust, and promoting the integrity of research and researchers.²⁸ Offering notification reflects a belief that transparency offers ethical value independent of any decision, demonstrating respect and helping individuals understand and perhaps even appreciate the nature of their involvement in an activity.²⁸ Similarly, notification can support researcher integrity, encouraging investigators to design trials whose purpose and features can be justified to those enrolled.²⁹

Our data suggest considerable heterogeneity in decision-making about notification. Some areas of divergence are consistent with a view that the importance of transparency is context-dependent.²⁸ From this perspective, it is appropriate that trials with different morally relevant features take different approaches regarding both whether and how to notify. Notably, several factors identified by our respondents as relevant for assessing the appropriateness of notification for a given PCT parallel those presented within a recent proposal for how to make research ethics oversight for clinical trials more “fit for purpose,”³⁰ including both the incremental risk presented by a research activity and whether the intervention under study involves a decision for which patients would (or should) be engaged in outside the research context.

Yet our data also suggest that the same features of a given PCT may sometimes lead different decision-makers to very different decisions about notification. These differences may lack appropriate justification. Importantly, several respondents referenced making decisions based upon empirically testable—but largely untested—claims. Examples include: whether patients and families in acute care contexts appreciate notification about research or instead perceive it as unduly burdensome, whether notification undermines trust, or whether patients within academic medical centers expect research to occur as part of the routine delivery of clinical care. While prior studies in related areas may suggest some insights,^17,31 empirical research to test these claims could inform future decision-making.

Additional empirical work might explore patient and family preferences about notification, including how and when such notification might be most appropriate. While these preferences do not yield determinative ethical guidance, they nevertheless can be informative. As one of us has argued elsewhere, when considering how best to fulfill the principles of beneficence and respect for persons, those we aim to benefit and respect through and during research presumably should have at least some say in how “benefit” and “respect” should be understood.¹⁶

Admittedly, empirical research is resource intensive and presents opportunity costs. These activities are likely to have differential value for different PCTs. For example, understanding the impact of notification may be more important for a PCT comparing two interventions in the context of a “high stakes” clinical event than one randomizing between two different approaches to encourage uptake of annual recommended screenings. Relatedly, as others have thoughtfully explored elsewhere, decisions about who to engage can have important consequences for which perspectives are elicited.³² For example, patients and families who have experienced the clinical condition under study may have different preferences than those without similar lived experience. Engagement with patients and community members during the design of a previous trial in emergency tracheal intubation³³ found that community members who had not experienced critical illness were more likely to express the desire to be informed about participation in minimal risk research than patients who had experienced critical illness; the latter group were more likely to express concern about the added cognitive and emotional burden of receiving notification about minimal risk research when facing high-stakes, time-sensitive decisions about clinical care.

These considerations notwithstanding, our finding of considerable heterogeneity in notification decisions suggests the value of future framework development to guide decisions about notification, such as through a Delphi process to generate consensus guidance. Our data, along with additional empirical work exploring patient attitudes and testing consequentialist claims about the impact of notification, could be important inputs for this process.

Moreover, these data also highlight how decision-making about notification is shaped—and arguably hampered—by asymmetries in the norms and expected practices for communicating uncertainty in the research context as compared to clinical care. In clinical care, clinicians frequently must choose between multiple treatment options for a given condition with insufficient evidence about the comparative effectiveness of those options, and little discussion occurs with patients about the clinical decision or the lack of evidence to inform it. Yet, when that same clinical decision is shifted to the research context, clinician-investigators are expected to communicate the rationale for the research (including the insufficient evidence base) and the relative risks and burdens of alternative treatments. As others have noted, this asymmetry may lead individuals to mistakenly believe they can avoid the risks presented by clinical uncertainties by refusing research participation.³⁴ This asymmetry arguably creates perverse incentives, setting higher standards for disclosure about (the lack of adequate) evidence for research—which generates evidence to improve future care—than for clinical care—which exposes patients to the same treatments without generating this evidence. This challenge may offer support for proposals, elaborated elsewhere,^24,35 for the importance of more routinely communicating the rationale for PCTs and related research—and, ideally, implementing systems to ensure that the results of such research are incorporated into ongoing clinical practices to improve future clinical quality.

Our study has several limitations. We spoke with a relatively limited number of experts and prioritized respondents with experience conducting or overseeing multiple PCTs. Our findings may not reflect experiences and approaches in other settings, particularly those with less PCT experience. Further, while we purposively stratified our sample by key professional characteristics, health system leaders were underrepresented relative to investigators and IRB/HRPP professionals. Finally, given differences in regulations and norms across countries, we focused our interviews on notification in the US context.

Conclusion

While considerable scholarship has explored the use of waivers or alteration of consent for PCTs, decision-making about notification when consent is waived has received little attention. Qualitative exploration of the experiences and judgments of those tasked with decision-making about notification can inform ethical practices regarding future notification decisions. Our results suggest that decisions about notification are context-dependent and reflect both normative and practical considerations. However, evidence that the very same factors may lead different decision-makers to very different conclusion about the appropriateness of notification suggests the potential need for a framework to better inform these decisions.

Table 5.

Who Decides?

Heterogeneity in IRB Support for Notification

it’s mostly an institutional review board decision as to how much a patient gets to know about studies where the full process for informed consent is not happening...The investigators usually want to gravitate towards the opt-out process. The IRB usually wants to gravitate towards the opt-in process. That is typically the push and pull while a study is being reviewed. When we first proposed this to our IRB, our IRB explicitly came back and said, “Absolutely not. Don’t tell people that we’ve given a waiver, that we put them in research,” because they thought people would be mad and upset and would go to the press and say, “[Institution’s] enrolling people in research,” and that sort of stuff. We had a couple of meetings with them and convinced them that we thought this was the right thing from a transparency standpoint and from an autonomy standpoint with patients to let them know this is what we’re doing and this is what we’re interested in learning. Our IRB then said, “Okay.” I think they were anxious about it initially, but they said, “Okay, we’ll try this.” Honestly, it’s gone fine.

Notification Often—but Not Always—a Collaborative Process Involving Diverse Operational and Clinical Leaders

At our institution, we had the IRB, obviously... We have patient privacy at the table... There’s some legal considerations. We have legal representation on the team. The clinicians and the researchers, I think that’s important, that caveat that all of us as researchers have a clinical background, but that’s different than clinicians who are at the bedside. We made sure to have representation at that table to talk through, like I mentioned, the logistics... What we found is that sometimes environmental services takes that flyer down every single time they clean. Who’s on the ground, making sure that that’s still available? Then to the point of like, if clinicians- they need to point it out. Families might not know the flyer’s there. When we do an orientation to the room, it’s one thing to potentially point out as an FYI, so talking through how we’re actually going to make sure families see it... It was quite a team to start drafting what approach we might take... I don’t think there was any institutional sign-off beyond the IRB and the investigators working through it together.