Real-time Perspectives of Surrogate Decision-Makers Regarding Critical Illness Research

Abstract

Objective:

We undertook the current investigation to explore how the pressures of serving as a surrogate decision-maker (SDM) for an acutely ill family member influence attitudes regarding clinical investigation.

Methods:

We conducted a prospective study involving SDMs for critically ill patients cared for in the ICUs of two urban hospitals. Measurements included participation in focus groups designed to explore perceptions of ICU care and clinical research. Audiotapes were transcribed and analyzed to identify common patterns and themes using grounded theory. Demographic and clinical data were summarized using standard statistical methods.

Results:

Seventy-four SDMs (corresponding to 24% of eligible patients) participated. Most SDMs were women and described long-term relationships with the patients represented. SDMs described their role as “overwhelming,” their emotions were accentuated by the fatigue of the ICU experience, and they relied on family members, social contacts, and religion as sources of support. Altruism was reported as a common motivation for potential study participation, a sentiment often strengthened by the critical illness episode. Although research was viewed as optional, some SDMs perceived invitation for research participation as tacit acknowledgment of therapeutic failure. SDMs expressed a preference for observational studies (perceived as low risk) over interventional designs (perceived as higher risk). Trust in the ICU team and the research enterprise seemed tightly linked.

Conclusions:

Despite significant emotional duress, SDMs expressed interest in investigation and described multiple factors motivating participation. Consent processes that minimize the effects of anxiety may be one strategy to enhance recruitment.

Clinical investigations enrolling acutely ill patients are essential to advancing critical care medicine.¹ Critical illness investigations are associated with challenges not encountered in chronic disease studies.^1,2 Conditions prompting ICU admission are precipitous and life threatening, and the care provided is technological.² Neither patients nor family members have had an opportunity for education regarding the nature of the disease process, expected outcome, or treatment alternatives. Further, patients in ICUs are frequently incapacitated and unable to provide informed consent permitting medical intervention or research participation.^1,3‐6 Finally, many therapies must be administered quickly following diagnosis. To enroll critically ill patients in clinical studies, permission must be obtained expediently from surrogate decision-makers (SDMs), many of whom are being confronted with complex and serious medical issues for the first time.^1,7‐9

Investigations involving cognitively impaired individuals are perceived as carrying greater risk and requiring additional safeguards compared with studies enrolling those who are cognitively intact.^10‐12 Previous investigations have examined the manner in which institutional review boards (IRBs) apply such protections, explored the views of individuals regarding substituted judgment for research of chronic diseases, and looked at the emotional impact of functioning as an SDM in the ICU setting.^7,10,13‐19 We undertook the current investigation to further study how the pressures and concerns of serving as SDM for an acutely ill family member might influence attitudes regarding clinical research and how this information might be used to enhance the consent process.

Materials and Methods

We have provided a concise description of the methods used in this study; greater detail is provided in e-Appendix 1 ^{(1.1MB, pdf)} . This study was approved by the Washington University Human Studies Committee (HRPO 06-0637).

SDM Sample and Recruitment

Participants were recruited from the surgical and medical ICUs of two tertiary-care institutions: Barnes Jewish Hospital (BJH) in St. Louis, Missouri, and Parkland Hospital (PH) in Dallas, Texas. These study sites were chosen because of the demographic and ethnic diversity they afforded. Eligible participants were SDMs for adult patients in the ICU (aged ≥ 18 years) who were intubated and mechanically ventilated for ≥ 48 h, and who were expected to require ventilatory support for ≥ 24 additional hours.

Data Collection

Data collection occurred during eight 2-week periods, which were scheduled approximately every 3 to 4 months to allow for complete turnover in ICU census. Focus group (FG) discussion guides were developed based on themes related to critical illness research. Hypothetical cases, vignettes, and samples of IRB-approved consent language were used to generate discussion regarding receptivity to enrolling critically ill patients in research. Participants were English-speaking adult SDMs representing adult patients. Two facilitators (A. C. and E. S.) moderated all FG discussions.

Analysis

Analysis of FG transcripts was based on grounded theory, which entails the simultaneous process of data collection, analysis, and theory construction.²⁰ Data are analyzed immediately following collection to identify dominant patterns, themes, and constructs of interest.^20‐22 ATLAS.ti software (ATLAS.ti Scientific Software Development GmbH) was used for coding and analysis. Findings reported here are based on analysis of narratives corresponding to the following codes: clinical care, clinical research, experience as an SDM, coping mechanisms, support networks, communication with the ICU care providers, medical research, perceptions of risks and benefits of research, motivations to enroll a loved one in research, and concerns about research participation. Verbatim quotations from the FG sessions have been included in the Results section, both to provide detail and to illustrate major themes that arose from these discussions. These quotations were edited slightly for purposes of clarity. For readability, brief quotations are included in this text; longer quotations are provided in e-Tables 1-5 ^{(1.1MB, pdf)} . Demographic and clinical data were analyzed using standard statistical software (Sigmastat; Systat Software).

Results

Patient and SDM Characteristics

Data collection took place during eight 2-week periods (February 2008-July 2009), during which time 524 ICU admissions were screened to identify 224 patients (43%) meeting inclusion criteria (Table 1). We were unable to contact SDMs representing 40% of eligible patients; this occurred more commonly at PH than BJH. A minority of SDMs (11%) declined participation, while SDMs representing 24% of eligible patients participated. Most patients were men, and nearly one-half either lacked medical insurance or indicated Medicaid as a payer source. Compared with PH, patients at BJH were older, more commonly white, and more likely to have respiratory failure as the admitting diagnosis. Severity of illness was reflected in prolonged ICU stay and a mortality rate of 22% (Table 2). We recruited 16 FG cohorts (4-8 SDMs per cohort, 74 SDMs total) and conducted 21 FG sessions. Most SDMs were women, described long-term relationships with the patients represented, and characterized themselves as moderately to very spiritual or religious. SDMs recruited from BJH were more commonly white and had completed higher levels of formal education (Table 3). We detected no interinstitutional differences in views expressed by FG participants.

Table 1.

—Patient and SDM Screening Summary

Category	Barnes Jewish Hospital	Parkland Hospital	Combined	P Valuea
Patients screened	244	280	524	NA
Eligible patients	115	109	224	NS
Patients represented by SDMs consenting to participateb	31 (27)	22 (20)	53 (24)	NS
SDM reasons for not consenting				.008
Unable to comply	16 (14)	11 (10)	27 (12)
Patient condition	13 (11)	5 (5)	18 (8)
Declined to participate	14 (12)	11 (10)	25 (11)
Unable to contact SDMc	39 (34)	50 (46)	89 (40)
Other	2 (2)	10 (9)	12 (5)

All data given as No. or No. (%). NA = not applicable; NS = not significant; SDM = surrogate decision-maker.

^aComparison between Barnes Jewish and Parkland Hospitals.

^bPercentages reflect proportions of eligible patients.

^cSDM not available by phone or in person during recruitment efforts.

Table 2.

—Patient Characteristics

Characteristics	Barnes Jewish Hospital (n = 32)	Parkland Hospital (n = 21)	Combined (n = 53)	P Valuea
Age, mean (± SE), y	53.7 (2.8)	45.7 (3.1)	50.1 (2.6)	.025
Sex, male (female)	13 (41)	7 (33)	20 (38)	NS
Ethnicity				<.001
Black	12 (38)	5 (22)	17 (32)
White	19 (59)	6 (28)	25 (48)
Latino and other	1 (3)	10 (50)	11 (20)
Payer source				NS
Commercial	10 (31)	5 (22)	17 (28)
Medicare/Medicare HMO	11 (34)	4 (17)	15 (28)
Medicaid	8 (25)	6 (28)	14 (26)
Uninsured	3 (9)	7 (33)	10 (18)
Reason for ICU admission				.006
Trauma/burns	6 (19)	8 (39)	14 (26)
Postoperative	9 (28)	1 (6)	10 (20)
Respiratory	13 (41)	6 (28)	19 (35)
Other categoriesb	4 (12)	6 (28)	10 (19)
ICU length of stay, mean (± SE), d	17 (2)	31 (7)	20 (2)	NS
ICU mortality rate	6 of 32 (19)	6 of 21 (28)	12 of 53 (22)	NS

Data given as No. (%) unless otherwise indicated. HMO = health maintenance organization. See Table 1 legend for expansion of other abbreviation.

^aComparison between Barnes Jewish and Parkland Hospitals.

^bSepsis, CNS, miscellaneous.

Table 3.

—SDM Characteristics

Characteristics	Barnes Jewish Hospital (n = 44)	Parkland Hospital (n = 30)	Combined (n = 74)	P Valuea
Age, mean (± SE), y	50.7 (1.9)	43.9 (3.1)	48.0 (1.7)	NS
Female sex	33 (75)	21 (69)	54 (73)	NS
Ethnicity				<.001
Black	18 (41)	7 (24)	25 (34)
White	24 (54)	7 (24)	31 (43)
Latino and other	2 (5)	16 (52)	18 (23)
SDM relationship to patient				NS
Spouse	13 (29)	4 (12)	17 (23)
Child	12 (27)	8 (28)	20 (28)
Parent	9 (21)	5 (16)	14 (19)
Otherb	10 (23)	13 (44)	23 (30)
Duration of SDM-patient relationship, mean (± SE), y	35.0 (1.7)	31.4 (2.8)	33.6 (1.5)	NS
Formal education, mean (± SE), y	14.0 (0.4)	10.8 (0.8)	12.8 (0.4)	<.001
Religiosity				NS
Not religious or spiritual	1 (3)	1 (4)	2 (4)
Slightly religious or spiritual	1 (3)	7 (24)	8 (12)
Moderately religious or spiritual	22 (50)	7 (24)	29 (39)
Very religious or spiritual	20 (44)	14 (48)	44 (46)
Prior employment in health care	20 (46)	7 (24)	27 (38)	NS
Prior experience with research	7 (16)	4 (12)	11 (14)	NS

Data given as No. (%) unless otherwise indicated. See Table 1 legend for expansion of abbreviations.

^aComparison between Barnes Jewish and Parkland hospitals.

^bGrandchild, grandparent, sibling, domestic partner.

Focus Group Findings

Challenge of the SDM Role:

Many individuals described the SDM role as “overwhelming” or “stressful” (e-Table 1 ^{(1.1MB, pdf)} ). Several factors contributed to emotional distress: observing a loved one in a critically ill state, grappling with uncertainty regarding recovery, and contemplating whether the loved one’s views were being accurately reflected. Fatigue contributed to this emotional burden. Comments such as the following were commonplace: “She’s been here almost 2 weeks. And since she’s been here, I’ve been here every day. I haven’t missed a day. And I’m not going to miss a day.”

Most participants reported relying on family members and social contacts (close friends, coworkers) for support and described developing close bonds with other SDMs encountered in ICU waiting rooms (e-Table 2 ^{(1.1MB, pdf)} ). Many SDMs reported relying heavily on their religious faith, frequently praying over decisions, and finding strength through scripture.

We explored the manner in which SDMs perceived that they made decisions (e-Table 3 ^{(1.1MB, pdf)} ). While engaged in the decision-making process, SDMs ultimately appeared to rely on the judgment and recommendations of the critical care team.

Motivations for Research Participation:

Most SDMs reported that altruism would motivate a decision to participate in research (e-Table 4 ^{(1.1MB, pdf)} ). For example, one SDM responded: “If it was to help someone, I think my grandmother would agree to it...because she would think ‘Well, maybe if somebody would’ve done it for me, I wouldn’t be in the situation I’m in right now.’” Many acknowledged that present-day medical innovations resulted from individuals participating in past studies.

For several respondents, this altruistic perspective was one that was gained through their ICU experience, which they described as personally transforming. While research participation was generally viewed as optional, a minority of respondents assumed that research was commonly performed without explicit permission. This perception appeared to result from the practice of clinicians performing interventions without consents. Several individuals stated that the fact that they would be approached for a research study implied that participation may impact the care their loved one received. “Anytime a person has a relative that’s in critical condition, they’re more apt to agree to do these kinds of things because they’re thinking it might benefit their loved one.” A minority of respondents reported that they would agree to research participation only if it provided benefit.

SDMs expressed a desire to learn the results of a study in which a loved one might participate. These individuals reviewed return of results as a tangible outcome of study participation. “Well, it’s going to make you feel more apt to do this again if you see that there was a good result from what you did. Where if you never get any feedback at all, I don’t know that anything I did was worth doing.” When potential obstacles to the return of results were discussed, specifically, the necessity of maintaining some of form of identification to allow follow-up, interest in the return of results waned.

Receptivity to research participation was balanced with consideration of potential risks. When presented with a spectrum of possible study designs (ie, observational studies, which were described as low risk, vs interventional designs, which were described as entailing greater risk), participants were generally receptive only to low-risk study designs.

Trust:

We explored how perceptions of trust and confidence in the medical (and research) establishment that SDMs held preceding the current episode of illness may be reinforced, altered, or impacted by their ICU experience (e-Table 5 ^{(1.1MB, pdf)} ). While respondents viewed ICU care providers favorably, some described a lack of trust. Mistrust seemed most pronounced in the setting of complications suffered by their loved ones. Individuals expressing a lack of trust stated that they would be unwilling to enroll their loved one in a clinical investigation. Discussants provided contrasting perspectives on the importance of the relationship with the care provider influencing willingness to participate in a clinical study. For some respondents, it appeared important to have a personal relationship with the individual requesting study participation. In contrast, others expressed explicit trust in a research team that appeared affiliated with the institution where care was being provided.

Finally, we examined whether attitudes pertaining to the study sponsor influenced the willingness of SDMs to permit participation. Many respondents expressed the view that a study’s sponsor was important. Hypothetical research supported by foundations and universities was, in general, viewed more favorably than research supported by the federal government or the pharmaceutical industry. More than one participant expressed negative or hostile views toward the government, while views toward for-profit companies were more nuanced. Ultimately, it did not appear that sponsor preference was strongly linked to willingness to permit participation.

Discussion

We probed the attitudes and thought processes of individuals who were actively engaged in critical care decisions in real time. Severity of illness was substantial, as evidenced by prolonged lengths of stay and mortality rates > 20%. SDMs appeared emotionally vulnerable, commonly describing their role as “stressful,” and frequently reported physical fatigue due to disruption of their daily routine. Further, many SDMs expressed altruistic tendencies that were accentuated by their ICU experience, and a willingness to contribute to clinical research in some tangible fashion. Finally, there appeared to be an undercurrent of therapeutic misconception.²³ Some SDMs assumed being approached for research participation implied that their loved one’s situation was dire, and that participation in such research may represent a last opportunity for successful treatment.

Our findings complement recent investigations exploring the manner in which the stress of serving as an SDM might influence decision-making regarding clinical research. Mehta et al¹⁷ examined the attitudes of SDMs who had recently been approached regarding a clinical study. Those agreeing to participate did so out of altruism, to further medical progress, and because they felt that the patient would have agreed to participate. SDMs declining participation did so predominately because of anxiety. Similarly, Menon et al¹⁸ found that SDMs for pediatric patients frequently declined study participation because of emotional duress. Collectively, these findings suggest that anxiety is a common barrier to research participation and support the argument for excluding low-risk studies from need for informed consent.⁷ At the very least, these perspectives underscore the importance of clinical investigators being sensitive to the dynamics and motivations of SDMs in recruitment efforts.

We examined the level of trust individuals placed in potential sponsors of research. Despite being a significant source of investigative funding, several individuals expressed negative views toward the federal government.²⁴ We could not determine whether this antipathy resulted from a negative perception of government in general, or whether there was a specific concern about government’s role in research that was problematic, such as the perception that the government would have access to patient-level data. In contrast, SDMs’ views toward a sponsoring role of for-profit entities, such as pharmaceutical companies, were more nuanced. While many individuals resented the possibility of companies profiting from their loved one’s participation, they recognized the necessity of capitalism in enabling medical advances. For most respondents, the nature of the sponsor did not appear to substantially influence receptivity to study participation.

While many studies have used qualitative approaches in the ICU setting, these investigations present challenges.^25‐27 Our low recruitment rate may have been a consequence of our study design. We recruited during a series of 2-week blocks, interspersed with several weeks of nonrecruitment to allow for census turnover, and targeted SDMs present at the patient’s bedside during daylight hours. In 50% of cases, nonparticipation was due to our inability to contact an SDM during these recruitment attempts. Further, it was necessary for SDMs to return at a later time to attend FG sessions. This was often precluded by scheduling conflicts and lack of availability due to precipitous patient decline. Additionally, we solicited opinions about clinical research from those agreeing to participate in our study. Such attitudes may differ significantly from those declining our invitation. Finally, opinions expressed in FGs might be influenced by the perspectives and personalities of coparticipants. While these factors may limit the generalizability of our findings, our study was successful in capturing the diversity of viewpoints held.

Our study had additional limitations. We probed perceptions and concerns about clinical research in abstract terms, identifying factors that may influence participation. We do not know the extent to which these perceptions translated into an actual willingness or disinclination to permit study participation. While an SDM may be receptive to allowing participation in a hypothetical circumstance, when confronted with the realities of an actual study, interest in participation may diminish. In addition, while qualitative methods allow us to explore the range of views held by participants, this approach does not allow us to dissect the determinants or correlates of particular view points (eg, education level, religiosity, severity of illness, time elapsing between onset of critical illness and FG participation, and so forth). Application of quantitative methods to ascertain these relationships will be essential to understanding the views of demographic subsets. Finally, we interviewed SDMs who were actively involved in clinical decision-making, but did not interview patients, and do not know the extent to which the views of SDMs and patients align. Determining this alignment, as well as its determinants, is essential to understanding the extent to which patients’ views are represented by SDMs.^9,28

Conclusions

Our findings highlight opportunities for the conduct of critical illness research. Despite emotional burdens, SDMs expressed interest in learning about and participating in clinical investigation. In response, ICUs should create a culture of educating patients, family members, and visitors about the importance of clinical research. The placement of informational materials in high traffic areas of ICUs and adjoining waiting rooms may be one way of achieving these educational goals and recruiting individuals for specific studies. In addition, some form of tangible acknowledgment of study participation, such as return of aggregate results, may be an important means of further solidifying relationships between the investigative and patient communities. Informed consent requires that individuals are accurately apprised of the purpose, risks, and benefits of proposed research, and make a voluntary decision to participate.¹² In recruitment efforts, investigators must be sensitive to the emotional state of SDMs and potential motivations for participation, to ensure that these standards are upheld.

Supplementary Material

Online Supplement

Abbreviations

BJH

Barnes Jewish Hospital

FG

focus group

IRB

institutional review board

PH

Parkland Hospital

SDM

surrogate decision-maker