Surrogate Consent for Genetic Testing, the Reconsent Process, and Consent for Long-Term Outcomes in Acute Respiratory Distress Syndrome Trials

To the Editor: Advancing critical care research is necessary to improve patient outcomes and has been defined as a priority for our healthcare system (1). However, most critically ill patients are initially incapacitated due to their acute illness, and are unable to participate in informed consent for research participation decisions (2). Therefore, surrogates make decisions for patients and often do so without a priori knowledge of the patients’ wishes. The surrogate consent process to enroll critically ill patients into research studies is complex. During the initial consent for a clinical trial, surrogates may also be asked to consent for the collection of biospecimens from the patient, including genetic material. Though consent rates for most genetic studies are generally high, individuals who are able to consent for themselves often have concerns regarding the use of their genetic material (3). In addition, racial and ethnic disparities have been reported in the willingness of individuals to consent to their own participation in genetic studies (4–6). However, whether surrogates are willing to consent for the collection of genetic material from critically ill patients has not been previously determined. When a surrogate provides consent for a research study, surviving patients who regain decisional capacity should be reconsented for their prior and continued participation. This reconsent process is unique to critical care research, as other incapacitated research participants, such as those with dementia, usually do not regain consent capacity. A better understanding of this reconsent process may provide insight into the patient’s perception of the burden of participating in clinical research. Finally, multicenter clinical trials of critically ill patients are recommended to include assessment of long-term outcomes (LTO) (7). However, it is presently unclear whether critical care survivors are willing to participate in LTO assessments. Therefore, using 1,164 patients enrolled into three Acute Respiratory Distress Syndrome Network trials (ALTA, OMEGA, and EDEN) (8–10), we sought to better understand the surrogate consent for genetic studies, the reconsent process, and the willingness of critical care survivors to participate in subsequent LTO studies. At the time of consent for these three clinical trials, surrogates were asked to provide consent for the collection of the patient’s genetic material for three types of ancillary studies: (1) genetic studies related to the parent study only (n = 1164), (2) future genetic studies for any acute respiratory distress syndrome (ARDS)-related research (n = 1164), and (3) future genetic studies for non–ARDS-related research (n = 1059). Patient race was categorized as white, African American, other, or not reported. Patient ethnicity was defined as Hispanic or not Hispanic; thus, study patients could be coded as being any race and also Hispanic. When they regained decisional capacity sufficient to provide informed consent, surviving patients underwent reconsent for their study participation. In regard to LTO, surrogates were initially consented for subject participation in assessments at 6 and 12 months after ARDS onset. Patients meeting eligibility criteria and not reconsented by hospital discharge were reconsented for LTO participation when subsequently contacted by telephone. Some of the results of these studies have been previously reported in the form of abstracts (11, 12). Overall, surrogates were generally willing to consent to the collection of the patient’s genetic material for all three types of ancillary studies (type 1, 92.0%, 95% CI = 90.3–93.4%; type 2, 90.5%, 95% CI = 88.7–92.1%; and type 3, 84.6%, 95% CI = 82.3–86.7%). However, surrogates were statistically less likely to provide consent for genetic studies when the future use of the material was not related to the parent study or ARDS research in general (P < 0.05). In univariate and multivariate analyses, surrogates of African Americans and other races were less likely to consent for each of the three different genetic studies when compared with surrogates of white patients (Tables 1 and ​and2).2). Surrogates of Hispanic patients were less likely to consent for genetic testing related to the parent study and genetic testing for future ARDS research not related to the parent study (Tables 1 and ​and22). TABLE 1. SURROGATE CONSENT FOR ANCILLARY GENETIC SUBSTUDIES BY PATIENT RACE/ETHNICITY TABLE 2. SURROGATE CONSENT FOR ANCILLARY GENETIC SUBSTUDIES BY PATIENT RACE/ETHNICITY: MULTIVARIABLE ANALYSES Of the 946 surviving patients, 407 (43%, 95% CI = 40–46%) were not reconsented due to either not being assessed for regaining consent capacity (n = 165) or a perceived lack of decisional capacity upon assessment (n = 242) (Figure 1). Of patients who survived and regained decisional capacity sufficient to provide reconsent, 522 of 539 (97%, 95% CI = 96–98%) affirmed their study participation. A total of 659 surviving patients met eligibility criteria for LTO assessments. The majority, 440 (67%, 95% CI = 63–70%), had provided reconsent for participation prior to hospital discharge. The remaining 219 (33%, 95% CI = 29–37%) were either not assessed for reconsent or lacked reconsent capacity in the hospital. Subsequently, they were consented for LTO assessment at the time of the initial follow-up telephone call conducted as part of the LTO assessment protocol. Overall, 211 of 219 (96%, 95% CI = 93–99%) were willing to consent to ongoing LTO study participation. Figure 1. Diagram of the reconsent process. Optimizing the surrogate consent process for critical care research is imperative to both protect the rights of vulnerable patients and increase study enrollment. To our knowledge, this is the first investigation examining the willingness of surrogates to provide informed consent for the collection of biospecimen samples from critically ill patients. In our study, surrogates were less willing to provide consent for future non–ARDS-related genetic research studies. Patients are generally willing to consent broadly to the use of biospecimens, but desire information regarding the type of research performed on their specimens before providing consent (3, 4). Similarly, our results demonstrate that surrogates are also less willing to provide consent for the collection of genetic material from patients when there is uncertainty regarding the use of the genetic material. Higher rates of study participation from surrogates may occur with enhanced communication concerning the actual use of the biospecimen material. In general, individuals of racial and ethnic minorities are less willing to agree to participate in clinical research studies (3–6). The lower consent rates for genetic studies in surrogates of underrepresented minorities highlights potential concerns regarding cultural differences and disparities in medical research (13, 14). Future prospective studies should examine the role of racial and ethnic disparities of surrogates in providing consent for a critically ill patient’s participation in research. In 2008, the Office for Human Research Participations Subcommittee for the Inclusion of Individuals with Impaired Decision Making in Research recommended that incapacitated research participants who are anticipated to regain consent capacity be evaluated for reconsent (15). Our high rates of reconsent may indicate that subjects agreed with their surrogates’ consent decision; however, this would be an oversimplification of a complex consent process. Previous studies have shown that significant discrepancies exist between critically ill patients and their surrogates regarding willingness to participate in hypothetical critical care research studies (16). A complete understanding of the reconsent process is also inherently hampered by the inability to include patients who died before they could be reconsented (i.e., survivorship bias). Furthermore, reconsent rates may be influenced by the magnitude of burden from continued study participation at the time of reconsent. As 43% of the surviving patients were not able to be reconsented, our results raise important concerns about the feasibility of conducting these assessments. To improve the conduct of the reconsent process, specific tools to assess decision-making capacity exist and should be used, and research personnel should be properly trained to reliably conduct competency assessments (17–19). Although obtaining LTO assessments of critical care survivors is important, concerns have been raised regarding feasibility of these studies and cohort retention (20). Our study demonstrates that subjects are willing to be contacted for LTO assessments, and therefore, high rates of cohort retention are possible in studies of critical care survivors. In conclusion, our study begins to examine the nuances of the surrogate consent and reconsent process, and demonstrates the need for future investigation in this area.