Benefits and Burdens of Participation For Caregivers of Cancer Patients in Hospice Clinical Trials: A Pilot Study

Research in hospice is relatively new, and clinical trials in the hospice setting are rare.¹ A small body of research has examined hospice patients’ participation in research. LeBlanc and colleagues write that multiple ethical obstacles have slowed efforts to create a strong evidence base in hospice, including arguments that 1) goals of research conflict with hospice care, 2) research unduly burdens patients and families at a difficult time, and 3) research is too difficult in the hospice setting. However, they note that these barriers can be overcome, and that we have a responsibility—as noted in the Belmont Report of 1979—for rigorous study of beneficial clinical interventions, respect for persons, and the application of justice or fairness in recruiting human participants, such as those in hospice care.²

Ethical guidelines for research in hospice and palliative care have been recommended, however, such recommendations do not reflect consensus among all stakeholders.^3,4 Moreover, research in the field has primarily been observational or retrospective in nature. A recent review found very few clinical trials in the hospice setting.¹ As a result, there are remarkably few studies published about ethical challenges in hospice clinical trials, including the recruitment and retention of participants in these types of trials.

Williams found that 46% of hospice patients were interested in participating in research involving surveys or interviews, and 45% were interested in therapeutic intervention research.⁵ There were a variety of reasons hospice patients reported wanting to be involved in research. Their reasons included wanting to help others (including the healthcare team), thinking research would improve their symptoms, the desire to contribute to science, and a sense of purpose or meaning.⁵ Phipps recorded reasons hospice patients did not want to participate in hospice research. Forty-two percent of hospice caregivers who declined to participate noted other priorities and spending time with family as the reason for not participating. Having a negative experience with research in the past, health concerns, and emotional distress were cited by approximately a third of declining caregivers. General disinterest was the most common reason for non-participation, cited by 61% of those declining participation. When comparing research participants with non-participants, it was found that participants felt they had more to gain from participating in research.⁶ While it is likely that some of these motivators and concerns are viewed similarly by hospice family caregivers, little is known regarding their perspectives on research participation.

ACCESS: A Hospice Clinical Trial

This study explores the measured and expressed benefits and burdens of participation in an ongoing clinical trial of an intervention named ACCESS (Access for Cancer Caregivers to Education and Support for Shared decision-making). The trial, funded by the National Cancer Institute, is a five year, multi-site, cluster-crossover, pragmatic, randomized clinical trial testing an intervention that provides education and emotional support to family caregivers of hospice cancer patients using online support groups and web conferencing. The ACCESS intervention uses a Facebook group as support for caregivers as they use web conferencing to join the hospice staff in plan of care meetings every other week. Details of the intervention can be found elsewhere.¹²

This pilot study used the ACCESS referral base and sought to answer the following research questions: 1) In what way do demographics impact the decision by caregivers to participate in hospice clinical trials? 2) How does depression, quality of life, and caregiver burden compare between those choosing not to participate in the clinical trial and those who do participate? 3) In what ways do the physical, psychological, economic, familial, and social dimensions of caregivers’ lives influence their decision to participate in hospice clinical trials?

Conceptual Model

This pilot study was supported by a conceptual model previously developed by Ulrich et al noting the benefits and burdens of research participation in cancer clinical trials (Figure 1).¹³ The model notes that the decision to participate in research involves the perceptions of benefits versus burdens related to several factors: physical, psychological, economic, familial, and social. In this case, physical dimensions related to participation in research can relate not only to physical symptoms but also to added logistical difficulties. Physical dimensions are those things that one physically feels or is physically required to do. For example in this trial, answering survey questions or interacting with Facebook or participating in a hospice team meeting. Psychological dimensions involve the attitude or feelings that come with participating in research. For example, participants might feel good about giving something back to society; on the other hand, they might encounter stress, fatigue, and anxiety from participation. Economic dimensions involve the financial cost (actual or perceived) from participating in research. For example, the inability to participate because one is working at a job might be particularly burdensome. Familial dimensions involve the effect of participation on the family system. For example, the benefit of learning something from the research that can be shared with family or that impacts one’s children is often perceived as beneficial. Family can also be perceived in a burdensome way, especially when there are pressures that might influence decisions. Finally, the social dimensions involve the impact of the research on someone’s social networks. For instance, the interaction with Facebook would increase one’s social network, which could be seen as a benefit or a burden. The decision to participate is the outcome of weighing the burdens against the benefits.¹³ Figure 1 illustrates the dimensions of benefit and burden and how their interpretation leads to a decision to participate or not participate in the parent trial.

Figure 1:

Conceptual Model

Methods

This study used a mixed method analysis to address the research questions. ^14,15 Statistical and qualitative data analysis compared those who did not consent to participate in the parent trial with those who consented to participate in only the pilot study. In the final analysis, the statistical data were mixed with the qualitative data to help interpret the findings.

Participants

Family caregivers from seven hospice agencies across one Midwestern state were recruited for the ACCESS trial. Hospice agencies served both rural and urban populations. Hospice agencies referred patients into the study following their initial hospice admission visit. Participants were required to: 1) be an informal caregiver for a hospice patient with cancer at one of our study sites, 2) be 18 years of age or older. This pilot study was approved by the university institutional review board. The parent trial is registered on clinical trial.gov (NCT02929108). With consent to record, all recruitment calls for the ACCESS and pilot study were audio-recorded for 9 months (June 2019-Feburary 2020). Once a referral declined to participate in the ACCESS trial, the research staff member asked if they would be interested in this pilot study as another research opportunity. Thus, all participants consented to participate in research, however, not all participants consented to participate in the clinical trial.

Recruitment calls were recorded with permission and used to identify benefit and burden statements explaining the decision to consent or not consent to the ACCESS study. The following script was approved by the University Institutional Review Board and used to gain permission to record the call.

Hello, this is [name of research staff] from the University of [Name]. May I speak with [Name of caregiver]? Your hospice has let us know that you’re willing to learn more about the research we are conducting with family caregivers. You received a flyer about the study during admission but we know that visit is pretty overwhelming and you may not remember seeing or hearing about the research. It takes about fifteen minutes to talk about the study- can I share a bit more about the work we’re doing and how you could help by being involved, or would another time be better for you? May I please audio record our call for quality review by our team?

Instruments

Measures in both groups were given at baseline following consent into each group. Caregivers were asked demographic questions including: age, relationship to patient, employment status, and past research study experience. To minimize survey burden and keep within a 10-minute interview timeframe, we chose to utilize the short version of the ACCESS trial scales. Instruments included the Generalized Anxiety Disorder- 2 item measure (GAD-2),¹⁶ the Patient Health Questionnaire-2 item measure (PHQ-2) ¹⁷, Caregiver Quality of Life-Revised (CQLI-R) ¹⁸, and Caregiver Burden Interview (Zarit-4)¹⁹. The brief scales have been found to be reliable and valid when compared with the longer forms.^17–19

Data Analysis

Pilot study only participants were compared with ACCESS participants consented during the same timeframe. Chi-square analysis was used to compare differences among caregiver demographics between the two groups. For the brief instrument scores in both groups, t-tests were used to examine differences between clinical trial and pilot study only participants on measures of anxiety (GAD), depression (PHQ), caregiver quality of life (CQLI-R) and burden (Zarit) separately, and a combined measure for the anxiety/depression was examined as well.

Recruitment calls were recorded and analyzed for those who agreed to participate in ACCESS and the pilot study only. A code book based upon the conceptual model dimensions of burdens and benefits was developed. The narratives of the recruitment call were read and coded to identify benefit and burden statements by participants, using a framework analysis. ²⁰ The first two authors tested the code book on an initial sample of the data and made revisions. Two staff (OL, CL) co-coded all recruitment call transcripts together, discussing each utterance that referred to a decision regarding participation in the trial. The coders met and jointly agreed on the final code for each statement. The final coding was reviewed by the first two authors and discrepancies were discussed by peer debriefing. Trustworthiness was assured with ongoing peer debriefings, an audit trail, and prolonged engagement in the research setting. Finally, findings were reviewed by the clinical trial interventionists as a member check on validity.

Results

Between June 1 2019 and February 29, 2020 the hospice agencies referred 400 caregiver’s to the clinical trial. Of those 400 referrals nearly half (n=197) had patients that were too ill to contact or we were unable to contact. Of the remaining 203 referrals contacted, 74 refused to consent to either study. The stated reason for not consenting was primarily due to the patient dying before they were contacted (n=20), the caregiver stating they were too busy (n=27), patient was actively dying, and several miscellaneous reasons (n=21) such as patient leaving hospice, caregiver contacted was not a primary caregiver, or a caregiver was unwilling to use technology (n=6), A total of 129 referrals consented to either the clinical trial (n=87) or the pilot study (n=42). See Figure 2.

Figure 2:

Summary of Recruitment and Qualitative Data on Benefits and Burdens of Trial

Overall, trial participants were more often adult children to the patient (55% vs. 21%, p = .005), younger (56 vs. 63 years, p = .04), and employed (47% vs. 24%, p = .02) as compared with pilot study participants (See Table 1). There were no statistical differences between the groups on previous experience with research as only approximately 15% of both groups had such experience.

Table 1.

Differences between pilot study participants and ACCESS consents

	Overall		Pilot Study		ACCESS
	N	%	n	%	n	%	p-value
Overall	129	100	42	32.50	87	67.50	--
Age, mean (sd) *	58.5	(14.1)	62.7	(17.4)	56.5	(11.7)	.04
<= 40 yrs	17	13.18	7	16.67	10	11.49
41–64 yrs	64	49.61	12	28.57	52	59.77
>= 65 yrs	47	36.43	23	54.76	24	27.59
Relationship to patient							.005
Spouse or partner	36	27.91	19	45.24	17	19.54
Parent	9	6.98	3	7.14	6	6.90
Adult child	57	44.19	9	21.43	48	55.17
Sibling	15	11.63	6	14.29	9	10.34
Other	12	9.30	5	11.90	7	8.05
Employment *							.02
Not employed/retired	61	47.29	27	64.29	34	39.08
Employed part-time	15	11.63	5	11.90	10	11.49
Employed full-time	51	39.53	10	23.81	41	47.13
Participate in any other research study?							.38
Yes	20	15.50	6	14.29	14	16.09
No	89	68.99	32	76.19	57	65.52
Unknown	20	15.50	4	9.52	16	18.39
Outcomes, mean (sd)
Anxiety (GAD)†	2.47	(1.87)	2.40	(2.06)	2.51	(1.78)	.77
Depression (PHQ) †	1.57	(1.76)	1.38	(1.81)	1.66	(1.74)	.40
Anxiety/Depression Combined	4.04	(3.25)	3.79	(3.28)	4.16	(3.24)	.54
Burden (Zarit) †	6.15	(3.76)	4.05	(2.93)	7.16	(3.72)	<.0001

^*There was one participant that did not provide their birthdate to calculate age, and two participants that did not answer the employment item.

^†GAD/PHQ measures reduced to 2 items each, Zarit reduced to 4 items.

Reported levels of anxiety and depression (separately, or combined) were not found to differ between the clinical trial and pilot study only participants. However, clinical trial participants reported significantly more burden than pilot study only participants (4.05 vs 7.16, p <.0001), see table 1.

Qualitative Findings

There were 139 individuals contacted to consent into the ACCESS clinical trial. We secured consent to record 89 total recruitment calls. Of those consenting to be recorded, 55 were participants consented to participate in the ACCESS trial, and 34 participants consented to participate in the pilot study only. As shown in Table 2, there were differences in the expressed benefits and burdens noted in the recruitment transcripts of ACCESS trial and pilot study only participants.

Table 2:

Coding summary of recruitment transcripts: Expressed benefits and burdens of participation

	Clinical Trial Participants n=55 transcripts; 31 expressed burdens and/or benefits			Pilot Study Only Participants n= 32 transcripts; 29 expressed burdens and/or benefits
	Benefit n=16	Burden n=15		Benefit n=7	Burden n=22
Physical (things that one physically feels or is physically required to do)	n=0 No comments	n=12 Finding the time to do it. The other thing is I get so much junk mail that my inbox is overflowing. Literally, I have 300 emails in my inbox...Because I just donť have time. I work on a computer all day long...When I get off work, the last thing I want to do is go look at my email.		0 identified No comments	n=16 Well in the first place, I donť do anything online or Facebook or any... I just donť participate in any of that stuff. I personally probably wouldnť be able to do it. I’m real busy with work and everything.
Psychological (attitude or feelings that come with participating in research)	n=13 Anything I can do to make any kind of this transition easier for anybody else is wonderful. I’m honored to have the opportunity.	n=3 I’ll warn you, I’m not the best at actually explaining my thoughts and whatnot, trying to find the words and all of that good stuff.		n=3 Anytime I can do something verbally like this, I will be delighted to.	n=2 I personally probably wouldnť be able to do it. I’m real busy with work and everything and I’m just kind of like so overwhelmed with everything.
Economic (involve the cost (actual or perceived) from participating in research)	n=1 Interventionist: I may be able to get an iPhone for you to be able to use. Caregiver: That would be fantastic.	n=1 Caregiver: The only thing is that right now I donť have my own phone. Interventionist: And you donť have Internet access? Caregiver: No mam.		n=1 No comments	n=2 No, I canť afford it [internet]
Familial (dimensions involve the effect of participation on the family system)	n=1 I mean, if my brothers donť have any problem with it, iťd be easier if you were just talking to one of us, [inaudible] getting all three of us involved	n=0 No comments		n=1 No comments	n=2 I’m going to have a whole house full of company tomorrow. This is like, yeah, anyway, all our people want to come and see her. And so, we’ve got four guests, I’ve got one coming tonight. One coming Sunday. This is like disaster.
Social (social dimensions involve the impact of the research to someone’s social networks	n=1 I’m a Mizzou grad...So, leťs get the University of Missouri some information.	n=1 I’m just kind of a very private type person.		n=2 I love our caseworker. Yeah, she’s wonderful. And when she called, and I’d do anything for her,	n=0 No comments

There were 21 pilot study only participants who expressed 28 statements for their decision not to participate in the clinical trial. Those 21 noted 22 statements classified as burden and only six statements classified as benefits to participation in the clinical trial. The most common burden identified was within the definition of the physical dimension (70%). Of the 21 benefit/burden statements for pilot study only participants, 16 (76%) included at least one burden related to the defined physical dimension. Physical dimension burdens were nearly evenly split between those who considered the technology or social media burdensome (55%) and those who noted caregiving burdens preventing participation. The second most reported dimension was psychological. Burdens were also noted in the familial and economic dimensions. Interestingly, more than one-third of pilot study only participants noted a psychological benefit, primarily of helping science, and 1 social benefit despite their decision not to participate in the clinical trial.

Although 55 ACCESS trial participants agreed to have the recruitment call recorded, the majority of the recorded calls did not reveal differences between expressed benefits or burdens. Twenty-two ACCESS participants made 31 statements related to why they were participating in the trial, and the statements were evenly split between benefits and burdens. Nearly one-third of those providing statements expressed both burdens and benefits to their decision to participate. Similar to the pilot study only participants, the majority of statements made by ACCESS participants were categorized in the physical dimension, however, two were statements related to technology or social media. The primary physical dimension burdens were related to caregiving demands. There was no trial participant that noted a familial burden. Clinical trial participants made 13 statements coded as benefits in the psychological dimension. The majority stated a desire to help others as reason for participation. Social, familial, and economic benefits were each cited one time as the only other benefits to participation. Of note is one statement that expressed the economic burden of the internet and need for a cell phone. The burden was addressed by providing a phone, and the participant chose to accept and participate, turning the burden into a benefit to participation.

Discussion

To the best of our knowledge this is the first study focused on understanding the benefits and burdens of hospice clinical trials for family caregivers. This is only a pilot study and lays groundwork for further investigation. We acknowledge the study is limited in its generalizability given the qualitative nature and the limitations of the recruitment calls rather than specific interview questions. Additionally, given the limitations of the data, we were unable to probe deeper into specific burdens or benefits since we could not ask questions, only respond to what was said and recorded. Finally, given the need for participant consent for any research it is not possible to compare those consenting to a clinical trial with those who do not consent for any study. Therefore, we compared those consenting to a clinical trial with those who would not consent to a clinical trial but would consent to a short pilot study.

Significant demographic and other differences were found between clinical trial participants and those who chose to participate in the pilot study only. Those participating in the clinical trial scored higher on the Zarit Caregiver Burden Interview than those who did not participate. This was opposite of what we had anticipated however, it was logical once we compared the demographics between the two groups. Significant differences between trial participants and pilot study only participants included mean age, employment status, and care recipient relationship. The additional burden placed on younger individuals still employed may be the reason for greater burden reported on the Zarit Burden Inventory. Additionally, increasing age has previously been shown to be a protective factor against anxiety and depression,²¹ helping to explain the lower anxiety and depression by the older pilot study only participants. Although this was not statistically significant, it is consistent with the lower perceptions of burden. Likewise, the older age of pilot study only participants may be playing a role in the identification of physical dimension burdens related to technology and social media by the majority of pilot study only participants. The older pilot study only participants had a significant number (nearly 30%) who specifically noted the technology or social media concerns as their reason for non-participation in the parent clinical trial. Despite our propositions that differences in age and employment status between study groups may explain these unexpected findings, further investigation is needed to confirm whether or not burden to participate is mitigated by increasing age and less employment.

The fact that the majority of participants noted the psychological benefit of contributing to the knowledge base about caregivers and trusting the University (See Table 2) reinforces the importance of noting these benefits early in the recruitment process. By clearly identifying themselves with the specific academic institution and acknowledging the need for better understanding of the problems of hospice caregivers and the potential contributions the study would make to science, they can help potential participants to consider the psychological benefits to their participation.

Likewise, the physical dimensions of caregiving related to the demands and the time required are notable burdens that are barriers to participation in hospice clinical trials. Minimizing the time required by caregivers in the design of the interventions and clearly articulating the time commitment may help overcome this barrier, or at least acknowledge this burden early in the discussion. The transcripts also identified the time of the phone call to be problematic in some cases. Connecting with caregivers who are at their place of employment, driving, or actively engaged in meeting patient needs is not ideal, although also not controllable. Recognizing inappropriate timing for these conversations and negotiating a time to call back would be advisable.

Conclusion

The benefits and burdens model for clinical trial participation is applicable to the caregiver experience in the hospice setting. Understanding the perceptions and dimensions of benefits and burdens to potential study participants is critical to not only the intervention design but also the tailoring of recruitment contacts and informed consent process. With regard to this specific intervention, assessing the digital literacy of those referred, the discussion of the benefits of the study towards improvement in the caregiving experience for others, and the articulation of the required time commitments to participate are opportunities for improvement. Using these strategies, it may be possible to help referrals identify benefits in participation and overcome perceived burdens, providing more information for an improved informed consent process.