Consent-Related Outcomes in the Alteplase Compared to Tenecteplase Trial

Abstract

Background and Objectives

In recent years, researchers have sought to address the challenges of obtaining informed consent for participation in acute stroke trials. We studied outcomes related to the use of deferral of consent in the phase 3 Alteplase Compared to Tenecteplase (AcT) trial.

Methods

As part of our protocol, we captured methods of consent, participant withdrawals, door-to-randomization times, and door-to-needle times. Participants at 3 sites were invited to complete a survey of attitudes regarding consent for AcT and for acute stroke trials generally.

Results

The AcT trial enrolled 1,600 participants from 22 centers across Canada of whom 1,537 were enrolled through deferral of consent (96.0%) and 63 (4.0%) were enrolled by prospective verbal consent followed by written informed consent. Of those enrolled by deferral of consent, 95% (1,454/1,537) consented to ongoing participation. Door-to-randomization times were similar regardless of method of consent, with an overall median of 30 minutes (interquartile range [IQR] 22–42): 29 minutes (IQR 22–42) in the deferral of consent group vs 32 minutes (IQR 25–44) in the prospective consent group (p = 0.1602). Survey respondents overwhelming agreed or strongly agreed with the use of deferral of consent in AcT (86%) and in any acute stroke trial (76%).

Discussion

Deferral of consent was broadly acceptable to participants in the AcT trial as demonstrated by low rates of withdrawal and by survey results. Door-to-randomization times using deferral of consent in AcT were short, although a system of prospective verbal consent used at 1 center took only slightly longer. These results support the importance of innovation around consent for acute stroke trials.

Introduction

Obtaining informed consent is increasingly recognized as a challenge for acute stroke trials: patients with stroke are frequently incapacitated, legally authorized representatives are rarely available quickly enough to provide consent,¹ and complex treatments must be administered with great rapidity.²

The Alteplase Compared to Tenecteplase (AcT) trial was a pragmatic trial³ conducted across Canada that sought to assess the noninferiority of tenecteplase to alteplase for patients with acute ischemic stroke within 4.5 hours of symptom onset.⁴ The AcT trial used a default deferral of consent approach: all participants regardless of capacity were to be enrolled without first obtaining written informed consent, with the expectation that consent would be sought as soon as possible thereafter from the participant or from a legally authorized representative.⁵ Deferral of consent differs from waiver of consent in which there is no intention to obtain consent from enrolled participants.⁶

In the United States, a process known as “Exception from Informed Consent (EFIC)” allows participants to be enrolled without consent only if the trial has been allowed to do so by the Food and Drug Administration, on a site-by-site basis. Achieving EFIC status is a lengthy and inconsistent process.⁷ Moreover, EFIC only allows consent to be waived if a surrogate is not present, meaning that it offers little in the way of time savings but does permit research participation when it would not otherwise be possible. It is important to note that EFIC does not mandate postenrollment consent, which is a decision of the local institutional review board.

Deferral of consent has the potential to facilitate rapid trial enrollment and treatment, and to limit biases associated with excluding potential participants who cannot consent. However, deferral may lead to enrollment against participants' wishes, had they been able to express themselves. To assess this risk, we conducted a preplanned substudy of consent-related outcomes in the AcT Trial.

Methods

Study methods for the AcT trial are described elsewhere,³ including a specific protocol for the use of deferral of consent.⁵ We collected metrics related to mode of consent, time-to-randomization and time-to-treatment, as well as withdrawal of consent. As a preplanned substudy, we administered a survey to participants who had been recruited at 3 sites where participants were enrolled by deferral of consent. At the time that consent to ongoing participation was obtained, the person completing the consent documents was invited to complete the survey. Data were collected in person and entered into the Qualtrics online survey platform. The survey was designed and tested at the Ottawa Hospital with input from researchers and patient partners (see appendix 1). Responses were analyzed using descriptive statistics.

Standard Protocol Approvals, Registrations, and Patient Consent

Participation in the AcT trial was exclusively by deferral of consent at all but one site. Written informed consent was then gained for all participants except in cases where the participant died and a surrogate could not be contacted. At the remaining site, consent was obtained whenever possible by brief verbal consent before enrollment, followed by written informed consent postprocedure; otherwise, enrollment was by deferral of consent. Participation in the postenrollment survey occurred with written informed consent. Ethics approval for the AcT trial was provided initially by the Conjoint Health Research Ethics Board at the University of Calgary (ID: REB18-1101) and then by site-specific ethics boards for the remaining recruiting centers. Consent for the survey was approved by the research ethics boards at the Calgary, Toronto-Sunnybrook, and Ottawa sites. The AcT trial is registered on ClinicalTrials.gov, with the number NCT03889249.

Data Availability

Data will be made available to researchers on reasonable request.

Results

Enrollment Outcomes

The AcT trial was conducted at 22 sites across Canada, 21 of which approved default deferral of consent. At one site, the research ethics board mandated that prospective verbal consent be sought from participants who appeared to be capable, or from legally authorized representatives if they were present at the time of enrollment, but permitted deferral of consent otherwise.

Of the 1,600 participants enrolled in the AcT Trial, 1537 were enrolled through deferral of consent (96.0%) and 63 (4.0%) were enrolled by prospective verbal consent. At the one site using both verbal consent and deferred consent, 113 participants (64%) were enrolled by deferral of consent and 63 (36%) were enrolled by prospective verbal consent from the participant (33/63, 52%) or from a legally authorized representative (30/63, 48%). Two patients who were deemed eligible and capable were approached but declined participation in the trial (2/65, 3.1%).

Twenty-three participants (1.5%) withdrew their consent, all of whom had been enrolled by deferral of consent. In addition, 19 participants (1.2%) who had been enrolled through deferral of consent opted out of the 90 assessment but did not withdraw from the trial. A total of 41 participants (2.5%) enrolled by deferral of consent died before consent could be obtained. Thus, consent to full study participation was obtained in 1,454 of 1,537 participants (95%) who had been enrolled by deferral of consent.

Time-Based Outcomes

The median door-to-randomization time was 30 minutes overall (interquartile range [IQR] 22–42): 29 minutes (IQR 22–42) for those enrolled by deferral and 32 minutes for those enrolled by verbal consent (IQR 25–44), which was not a statistically significant difference (p = 0.1602 by the Wilcoxon rank-sum test). The median door-to-needle time was 36 minutes overall (IQR 28–49): 36 minutes in the deferral of consent group (IQR 28–49) vs 39 minutes (IQR 28–52) in the prospective verbal consent group, again with no statistically significant difference (p = 0.2707 by the Wilcoxon rank-sum test).

Survey Results

The survey was administered to 78 participants across 3 sites: 44 at University of Ottawa, 21 at University of Calgary, and 13 at University of Toronto. See Table for participant demographics.

Table.

Demographics

Total participants	78
Sex
Female	41
Male	37
Age (in y)
Median	63
Range	21–100
Education
College diploma or university degree	51
Some college or university	8
High school diploma	18
Declined to answer	1
Survey respondent
Participant	51
Surrogate decision maker	27

The vast majority of respondents felt it was “very important” (75/78, 96%) to find new ways to treat stroke. When asked how important it was to obtain a family member's consent before enrollment, a plurality of respondents 33 of 78 (42%) stated that it was very important, although 25 of 78 (33%) said it was only somewhat or a little important, and 20 of 78 (26%) reported that it was not important at all.

The vast majority of respondents (58/78, 74%) stated that they would find it acceptable to be enrolled into an acute stroke trial even if a family member were unavailable to consent on their behalf. Factors that would make enrollment without prospective consent more appropriate included: if the condition being treated carried a high risk of death or disability (59/78, 76%), if data supported similar safety for the experimental intervention and standard care (51/78, 65%), and if there were no approved treatment for the condition (48/78, 62%).

When asked about deferral of consent specifically, 18 of 21 participants (86%) either agreed (6/21, 29%) or strongly agreed (12/21, 57%) that “deferral of consent is acceptable for the AcT study.” Only 3 of 21 respondents (14%) neither agreed nor disagreed with that statement, and none disagreed. When asked whether deferral of consent was acceptable for any acute stroke study, 16 of 21 respondents (76%) agreed or strongly agreed that it was (see Figure).

Figure. Participant Agreement With Deferral of Consent.

When asked to describe their reaction to having been enrolled into the AcT trial without first having provided consent, a strong majority of participants used positive terms (57/78, 73%) such as “satisfied,” “honored,” and “grateful,” while 6 of 78 (8%) used neutral or negative terms such as “do not love it,” “upset,” or “disturbed.”

The vast majority of respondents expressed interest in advance consent to research participation (58/78, 74%) with a plurality expressing strong support (31/78, 40%). Most of the participants expressed interest in providing advance consent for any stroke trial to which they might be eligible (56/78, 72%). There were no significant differences between men and women across all responses.

Discussion

First, deferral of consent was feasible as the default method of enrollment in the AcT trial. It was approved as the default method of consent by 21 of 22 research ethics boards across Canada, and postenrollment consent was successfully obtained in 95% of cases.

Second, deferral of consent seems to have been broadly acceptable to participants, as demonstrated by the very low rate of participant withdrawal (1.5%) and by the results of the participant survey. These survey results align with data from the SPOTLIGHT and ULTRA trials in which 90% or more of respondents agreed with the use of deferral of consent.^6,8 Survey responses from participants in the ESCAPE trial differed,⁹ which may relate to the fact that participants in the ESCAPE trial could have been randomized away from receiving endovascular thrombectomy, which was widely considered to be an efficacious intervention despite limited RCT-level evidence at the time.

Third, deferral of consent seems to have facilitated rapid randomization and treatment. Door-to-randomization times were fast across the AcT trial, both for those enrolled by deferral of consent and by prospective verbal consent. In the ESCAPE trial, a 2-physician model of consent was used to enroll incapable patients, which was not faster than prospective written consent, with door-to-randomization times about 20 minutes slower than in the AcT trial.⁸ In the IST-3 trial, enrollment through prospective written informed consent was associated with a delay of more than 20 minutes when compared with enrollment through deferral of consent.¹⁰

In the AcT trial, there was only a nonstatistically significant difference of 3 minutes between door-to-randomization times for participants enrolled by deferral of consent and those enrolled by prospective verbal consent. This result illustrates how brief the conversation about participation must have been, although none of the 63 participants enrolled in this manner subsequently withdrew from the study. Although 3 minutes is not a large amount of time, it could amount to a clinically meaningful difference in the setting of acute stroke. That being said, the time savings may be balanced by the additional value of obtaining pre-enrollment verbal consent. In circumstances where deferred consent is not permitted, abbreviated verbal consent before enrollment followed by written consent thereafter could balance the goals of respecting participant wishes while minimizing treatment delays. A similar 2-stage approach was used in the TICH2 trial, although participants were still required to sign a 1-page information sheet before enrollment. Despite this requirement, CT-to-randomization and onset-to-treatment times were significantly shorter for participants enrolled using an abbreviated consent than standard consent in the TICH2 trial.¹¹

We acknowledge several limitations. First, a small but important number of patients died before consent could be obtained, and therefore, the number of participants who withdrew from the AcT trial could possibly have been greater. It is appropriate to consider the advantages and disadvantages of having included these patients' data in the analysis of the trial given that consent was never obtained, although doing so limits the bias that might be encountered if data from deceased patients are excluded. Second, the small number of participants enrolled by prospective verbal consent at only 1 site may have limited our ability to detect a true difference in randomization times between that group and those participants enrolled by deferred consent. Third, the number of participants who completed the survey was a relatively small proportion of the overall number of patients enrolled in the AcT trial (5%), although the sample size is robust when compared with other surveys of participants in acute stroke trials.^8–10 Moreover, a technical issue led to the loss of responses to 1 survey question from 2 sites, and therefore, the sample size for that single question is smaller.

As operationalized in the AcT trial, deferral of consent was acceptable to the vast majority of study participants and facilitated rapid door-to-randomization and door-to-needle times. These results support the use of a well-justified and standardized deferral of consent process in acute stroke trials. Success at 1 site with verbal consent before randomization followed by written informed consent after randomization also holds promise and suggests the importance of innovation around consent methodologies for acute stroke trials.

Appendix. Authors

Name	Location	Contribution
Michel Christopher Frank Shamy, MA, MD	Department of Medicine, Ottawa Hospital Research Institute, University of Ottawa, Ontario	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Brian Dewar, MLIS	Ottawa Hospital Research Institute, Ontario	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Yan Deschaintre, MD	Centre Hospitalier de L'université de Montréal, Quebec	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Nishita Singh, MD	Department of Neurology, University of Manitoba, Winnipeg	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Carol Kenney, RN	University of Calgary, Alberta	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data
Mohammed A Almekhlafi MD, MSc	University of Calgary, Alberta	Analysis or interpretation of data; major role in the acquisition of data; drafting/revision of the manuscript for content, including medical writing for content
Ayoola Ademola, PhD	University of Calgary, Alberta	Major role in the acquisition of data; analysis or interpretation of data
Brian H. Buck, MD	Department of Neurology, University of Alberta, Edmonton	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Tolulope T. Sajobi, PhD	Department of Neurology, University of Calgary, Alberta	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Luciana Catanese, MD	McMaster University, Hamilton, Ontario	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Kayla D. Sage, MD	University of Calgary, Alberta	Major role in the acquisition of data; analysis or interpretation of data; additional contributions (in addition to 1 or more of the above criteria)
Dar Dowlatshahi, MD, PhD	Department of Medicine, Ottawa Hospital Research Institute, University of Ottawa, Ontario	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Laura C. Gioia, MD	Centre Hospitalier de L'Université de Montréal, Quebec	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Aleksander Tkach, MD	Kelowna General Hospital, British Columbia	Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data
Richard H. Swartz, MD, PhD	Sunnybrook Health Sciences Centre, Toronto, Ontario	Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data
Bijoy K. Menon, MD, MSc	University of Calgary, Alberta	Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data

Study Funding

The AcT Trial was supported by the Canadian Stroke Consortium, the Canadian Institutes of Health Research (grants 419722 and 450890), the Alberta Strategy for Patient-Oriented Research Support Unit, Alberta Innovates, the Heart and Stroke Foundation, and the University of Calgary.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures.