key: cord-1044504-tmdxxg9v authors: Brody, Abraham A.; Convery, Kimberly A.; Kline, Danielle M.; Fink, Regina M.; Fischer, Stacy M. title: Transitioning to Remote Recruitment and Intervention: A Tale of Two Palliative Care Research Studies Enrolling Underserved Populations During COVID-19 date: 2021-06-20 journal: J Pain Symptom Manage DOI: 10.1016/j.jpainsymman.2021.06.017 sha: a792b39f0a3d6c07797c7bafda8c422541dc55cb doc_id: 1044504 cord_uid: tmdxxg9v CONTEXT: During the COVID-19 pandemic, community-based research studies experienced prolonged shutdowns unless able to pivot to remote study procedures. OBJECTIVES: To describe the revision of two National Institutes of Health funded community-based palliative-focused clinical trials serving underserved populations to accommodate remote subject enrollment and examine its impact. METHODS: Transitioning to remote processes required multiple protocol and procedural changes including: altering informed consent processes; reducing the number of surveys administered; adding internet access as an inclusion criterion. To understand technological challenges, a screening tool was developed for one study to identify potentially eligible subjects’ technology abilities and accessibility. RESULTS: Subjects’ limited access to the internet and internet-enabled devices and discomfort with technology led to changes in recruitment patterns. Lack of familiarity with technology increased the amount of time it took research team members and subjects to connect remotely. Patients with significant cognitive and/or sensory deficits were at higher risk of experiencing fatigue during remote study visits leading to streamlining of data collection. A researcher-developed technology screening tool found that potential subjects were not comfortable with videoconferencing through Zoom expressing a preference for phone visits. Reduced travel time made scheduling remote study visits more efficient. CONCLUSION: Future community- and home-based palliative care trials must consider the best way to utilize remote recruitment, enrollment, and data collection processes to increase efficiency and reduce costs. Researchers should consider technology accessibility and train staff to ensure the greatest possible opportunity to recruit underserved populations who have traditionally been underrepresented in research studies. The World Health Organization first reported the SARS-CoV2 outbreak causing COVID-19 infection in January 2020. The pandemic's global spread had significant impact on research programs housed at academic centers across a broad continuum. Determining schedules to support social distancing in lab environments and prioritizing therapeutic trials with judicious use of Personal Protective Equipment (PPE) enabled cautious resumption of some research. Community-based research, including interventional palliative care trials, often without access to limited PPE supplies, faced prolonged shutdowns unless able to pivot to virtual study procedures. This manuscript's purpose describes the revision of two National Institutes of Health (NIH) funded community-based palliative-focused clinical trials to accommodate remote subject enrollment. Both experienced an initial pause at pandemic's start and initiated protocol modifications to facilitate study resumption. We outline eligibility alterations, technology transitions, and staff training processes each study adapted to this changing environment, ensuring safety for both study team and participants while maintaining protocol integrity. We also examine the impact of these adaptions to research recruitment, enrollment, and discuss lessons learned. With arrival of effective vaccines, this is an ideal moment to reflect upon innovations both research teams developed throughout the pandemic and sustain those highly effective procedures to run more equitable, accessible, and efficient communitybased palliative clinical trials in the future. A brief overview of purpose, aims, recruitment, and enrollment procedures of each clinical trial is presented followed by challenges faced and protocol modifications made due to the pandemic. Case Study 1 -The Dementia Symptom Management at Home Program (DSM-H) The DSM-H trial (NCT03255967) is a National Institute on Aging hybrid stage 3-4 embedded pragmatic cluster randomized control trial (RCT). 1 The multicomponent dyadic quality improvement intervention (Aliviado Dementia Care) provides home health clinicians with training, mentoring, and a toolbox of evidence-based resources to embed within the electronic health record (EHR) and use in daily practice to ensure person and caregiver-centered care, and includes specific components on addressing needs of specific underserved groups. Aims focus on whether the intervention, provided to persons living with dementia (PLWD) and their informal caregivers, improves dyadic quality of life, reduces symptoms, and decreases hospitalization and emergency room visits. Subject dyads are recruited within three days of admission across urban, suburban, and rural settings to one of three home health agencies specifically selected for their diverse populations. Overall enrollment target is 345 dyads, each completing four home research visits (days 1, 15, 30, and 60). Eligible participants are: 1) ≥65 years; 2) admitted to one of three participating home health agencies; 3) English and/or Spanish speaking; 4) scored at least 6+ on the Quick Dementia Rating Scale (at least mild dementia); and 5) with a caregiver ≥18 years who provides in person care to the PLWD for at least 8 hours/week. 2 Participants are excluded if they: 1) have a separate serious mental illness (bipolar disorder, schizophrenia) other than forms of dementia, depression, or anxiety; 2) are solely receiving infusion or home health aide services; 3) reside in assisted living facilities or board and care homes. A point of contact, at each agency, either reviewed patient information at intake and admission or developed an automated report. Eligible subjects were submitted into a REDCap database. 3 Research staff contacted patients and caregivers to introduce them to the study and complete eligibility screening. Once eligible, an in-person home visit was scheduled within three days of the patient's start-of-care date to complete subject informed consent and enrollment. On March 10, 2020 all in-person recruitment and enrollment activities were suspended. The five-year (54 month) multisite study was in its 31st month with 109 subjects enrolled. Following discussion with the Data Safety Monitoring Board (DSMB) and IRB, follow-up visits with previously enrolled subjects were conducted by phone, and when possible, data were collected via Zoom Ò teleconferencing. As the pandemic's severity became clearer, study protocol modification from in-person to fully remote data collection was presented to the DSMB in June 2020. The revised protocol was approved by the DSMB in June and IRB in July for the New Jersey site and October for the Utah site. The third site in Florida was about to implement at the start of the pandemic, and as of May 2021, had not restarted due to a large proportion of staff still unable to be vaccinated and staff burnout due to the pandemic. Case Study 2 -Apoyo con Cariño (Support with Caring): Improving Palliative Care Outcomes for Latinos with Advanced Medical Illness Apoyo is a National Institute of Nursing Researchfunded RCT of a culturally tailored, lay patient navigator intervention versus enhanced usual care (NCT03181750) aimed at improving palliative care outcomes for Latinos with serious non-cancer illness. 6 Conducted across Colorado in urban, rural, mountain, and frontier communities, all enrolled patients, recruited from five academic-affiliated sites and five community sites, receive culturally tailored, linguistically appropriate materials discussing advance care planning (ACP), pain/symptoms, and hospice care written at a 5th grade reading level. Intervention patients receive five home visits from bicultural, bilingual lay community navigators who address palliative care barriers. Visits follow a prescribed visit manual and aim to improve quality of life, ACP, pain/symptom experience, and hospice utilization. The enrollment target is 240 patients with up to 210 family caregivers. Eligible participants are: 1) ≥18 years, 2) self-identified Latinx; 2) English and/or Spanish speaking; 3) capable of providing informed consent; 4) diagnosed with advanced medical illness having limited life expectancy as identified by either a) meeting one of the noncancer CARING 7 (Cancer, Admissions ≥ 2, Residence in a nursing home, Intensive care unit admission with multi-organ failure, ≥ 2 Non-cancer hospice Guidelines) criteria, a prognostic index to identify patients who may benefit from a palliative approach; or b) patient's primary or specialty care provider answers "no" to the following question: "Would you be surprised if this patient died within the next year?" 8, 9 Patients may invite a patient-defined family caregiver (≥18 years and have decisional capacity to provide informed consent), although their presence and participation are not required for enrollment. Participants are excluded if they: 1) lack decisional capacity; or 2) are already enrolled in hospice care. The research team identified a healthcare provider (e.g., nurse, physician, clinical trials coordinator, medical assistant) as the onsite coordinator. Patients were invited to participate by their primary or specialty provider either in person at an appointment or via letter generated by the research team who screened Latinx patients for serious illness eligibility criteria through EHRs, then messaged or emailed the provider to confirm participants' eligibility and to electronically sign a letter inviting patient participation. Ten days post letter mailing to the patient's home, the navigator telephoned the patient to schedule an in-person consenting visit and baseline data collection in the home or place of their choosing. Navigators contacted patients within the week to inform them of intervention or control arm assignment. On March 16, 2020 the academic center's campus closed and halted all research activities. Two days prior, the study team had paused all in-person interactions with patients due to the vulnerable nature of their underlying serious illnesses and the lack of PPE for home use. The four-year (48 months) study was in its 41st month with 195 participants enrolled, 178 had completed participation, 17 were in active study period, and 10 were active in the intervention. Enrollment was in the final quarter and due to funding constraints, further extension for subject enrollments was impossible. Case Study 1. In transitioning to an all-remote data collection process, several changes were made to both protocol and standard operation procedures. First, because of social distance restrictions in senior housing, and concerns among caregivers of reducing risk even in those not in senior housing, some family caregivers had shifted much of the direct care they were providing to management of care. Because of this, caregiver inclusion criteria which initially required 8+ hours of "in-person" care were changed to "managing care". Second, the number of surveys administered was decreased because we found that Zoom or telephone interviews took longer than in-person surveying, causing greater burden. PLWD, suffering from moderate to severe impairment (QDRS 2 score of >12), were no longer questioned on pain, quality of life, and confusion. Caregivers or direct observation became the sole source of this information. Additionally, we learned there were frequent webcam problems during visits with already enrolled subjects at the beginning of the pandemic. Therefore, the 3D-CAM, 4 a delirium assessment that required observing the PLWD, was replaced by the FAM-CAM, 5 another instrument by the same researchers that uses a family informant. Third, informed consent was moved to an electronic format utilizing REDCap. 3 Fourth, internet access was added as an additional inclusion criterion due to consenting and data collection process changes. Finally, an additional algorithm-based dementia identification report was added at the New Jersey-based site in an attempt to increase subject referrals. Research coordinators were trained through mock remote visits prior to restarting. Case Study 2. The patient navigator team served as key informants in shifting to a remote enrollment process and intervention delivery. In response to navigators' concerns about barriers to technology, screening questions (SDC1) were developed for potentially eligible participants. The team also queried researchers with experience in remote enrollment for additional ideas related to training for consenting and delivering a palliative-focused intervention remotely. The Colorado Multiple IRB approved the revised protocol April 23, 2020. The research team conducted remote roleplaying with the navigators until the investigators determined navigators' comfort and ease in following virtual scripts. The research team continued to use pre-COVID-19 screening and patient identification methodologies. Potentially eligible patients were still contacted by their primary or specialty care provider via letter or at virtual visits. Unless patients opted out, one of the navigators attempted to contact the patient, conducted technology screening, and invited them to participate. Consenting visits and collection of baseline measures were completed remotely via the participant's preferred mode of communication. For participants randomized to the intervention arm, visits were conducted over the participant's preferred mode of communication. Follow-up data collection measures continued by telephone, unchanged from pre-pandemic study procedures. Simple descriptive statistical tests including frequencies, means, and medians were used to analyze the enrollment data in both trials. Changes made to both study protocols regarding participant eligibility, enrollment, technology access, and data collection are summarized in Table 1 . Results of these changes and impact on each study are described. Case Study 1. One of the key issues that arose was that the additional algorithm-based referral approach to recruit more individuals served as a barrier as it was not specific enough. This led to higher ineligibility rates due to a lack of dementia, sapping staff resource time (Table 2) . Additionally, we found it harder to make contact with caregivers to perform screenings (Table 3) , which we hypothesize relates to increased but unmeasurable caregiver burden during the pandemic. Case Study 2. The first challenge during this time was a decrease in potentially eligible patients based on screening procedures. This phenomenon was well documented during the first surge; the number of persons admitted for acute and chronic health conditions such as heart failure and chronic obstructive pulmonary disease decreased while the number of admissions related to COVID soared. 10 This led to a smaller eligible patient pool to approach for study participation. Case Study 1. Overall, between July and December 2020, 928 automated referrals were generated, of whom 221 dyads were contactable by telephone and agreed to be screened, 62 were found eligible, and 30 consented to participate in the study. Case Study 2. Overall, 14 participants consented during the remote enrollment and intervention delivery phase of the study from May 1 through July 31, 2020. At three-month follow-up survey completion, outcomes assessors unblind themselves to collect direct feedback on patient navigator visits. Seven patients were randomized to the intervention group with follow-up data collected. Case Study 1. Prior to receiving the Zoom link, subjects were screened for internet access; usage of videoconferencing software on their device(s) was confirmed. Some subjects failed enrollment due to poor internet connectivity or computer issues ( Fig. 1 and Table 3 ), e.g., signal was not strong enough to view documents online and/or receive control of the mouse to sign consent digitally. In other cases, subjects connected to Zoom using smartphones but were not enrolled as electronic signatures could not be completed due to platform limitations. Overall, eligibility, and in particular lack of internet (n =32, 26.8% of failed enrollments among eligible subjects), became a significant enrollment barrier to otherwise willing individuals. Among this group, 21 adult children, eight spousal caregivers, one sibling caregiver, and two PLWD who did not live with their caregiver did not have access. Another issue affecting enrollment was subject's familiarity with or access to technology. Some older adult caregivers did not have a computer and/or were not comfortable with downloading the Zoom app. If one part of the dyad was without access to internet and computer screen, the dyad would fail enrollment. Case Study 2. Access and comfort level with technology proved to be a significant barrier to virtual study conduct. All eligible persons were screened using the technology screening survey (SDC1) and the majority reported having either limited internet access or Zoom capabilities (Table 3 ). All screened patients (100%, N = 19) recruited during COVID-19 were not comfortable with videoconferencing through Zoom expressing a preference for phone visits. Case Study 1. For those who were contacted, research coordinators were trained to administer assessments by telephone and Zoom teleconference. To reconceptualize for Zoom administration, survey questions were shared with subjects as PowerPoint slides (PDF) prior to the visit to enable greater ease of administration. To complete informed consent electronically, research coordinators invited caregivers to attend a Zoom meeting. Prior to receiving the Zoom link, subjects were screened for internet access; usage of videoconferencing software on their device(s) was confirmed. When collecting data virtually, caregivers were heavily relied upon to provide reassurance, focus, and clarity to PLWD as research coordinators conducted assessments. If the PLWD did not have assistance from the primary or a secondary caregiver, we could not facilitate data collection. Additionally, primary caregivers often experienced challenges when trying to download Zoom software, run the application, and/or found they had insufficient internet or noncompatible device(s). In several cases, connectivity issues impacted study sessions by extending call duration, affecting assessment data completeness. Case Study 2. Patient navigators, pre and during the pandemic, administered survey assessments by telephone. The structured interview, previously performed in patient's homes, used paper and pen to administer surveys where the questions were both visible to the subjects and read aloud. To reconceptualize for telephone administration, informed consent documents were mailed to participants prior to the consenting visit enabling them to read along with the navigators and sign consent documents. Consent forms must be completed correctly and mailed back to the navigator prior to administering baseline surveys. If signatures were missing, the IRB required that the form was returned to patients for revision prior to any data collection. Case Study 1. The flexibility provided by Zoom visits allowed the dyad to be in different locations at the same time, which led to fewer scheduling conflicts causing failed enrollment (7.7% vs. 2.4%). Remote visits enabled more interviews in a day due to lack of travel time, and increased hours to perform enrollments facilitating access for caregivers working during daytime hours. Case Study 2. Patients reported high levels of satisfaction with remote navigator visits (100% rated visits very satisfying) and perceived them to be very helpful (86%). Overall, remote processes proved to be more time efficient without time required for travel. This was especially important if a patient missed a visit, as the navigator could shift to other study tasks. This manuscript provides information around technology use in elderly, their caregivers, and underserved populations. There are opportunities across all homebased and outpatient care to move to telephonic or virtual care − not only in research but from a clinical perspective. Case Study 1. The most important lessons learned in the remote study processes include: 1) caregivers who did not live with the PLWD were now providing more care remotely due to COVID-19 social distancing; 2) PLWD with significant cognitive and/or sensory deficits were at higher risk of experiencing subject fatigue during televisual study visits leading to a streamlining of data collection; 3) lack of access to internet or internet-enabled devices led to increased screening failures; 4) lack of familiarity with technology increased the amount of time it took for research team members and caregivers to connect remotely for data collection; 5) remote data collection required a reconsideration of what assessment instruments were required and/or feasible; 6) remote processes could increase efficiency and access for working family caregivers. Case Study 2. This study garnered similar and unique lessons from remote recruitment and enrollment processes: 1) both providers and patients were less likely to respond to outreach efforts compared with the time prior to the pandemic; 2) internet access was limited in the population, the majority of whom are socioeconomically underserved; 3) even for those with internet access, comfort and familiarity with technology was limited; 4) a simple technology screening tool helped the study team accommodate participants' preferences on how to interact with patient navigators; 5) satisfaction with the intervention remained high despite visits by phone rather than in person. This manuscript describes two methodologically different clinical trials, in separate areas of the country, recruiting and enrolling distinct yet similarly underserved populations. Both trials adapted to pandemic conditions by shifting in-person research activities to remote environments. Not surprisingly, both trials experienced a decrease in recruitment due to lower eligibility rates, whether due to inability to identify and contact potential participants or an increase in screening failures (Fig. 1) . Technology challenges proved to be important barriers across both trials. Even those with smartphone access but no computer had difficulty accessing systems such as REDCap as small screens were not optimized for consents. While data support the pervasiveness of household smartphones 11 and computers, this is not evenly distributed across populations. Older adults and households with lower median income are far less likely to own these devices compared to younger populations or those in higher socioeconomic brackets. We found that access was an important barrier although lack of comfort and proficiency with the technology proved to be significant barriers as well. This led to a preference for telephone over video technology in some circumstances. In other cases, using video technology demanded considerable additional time to conduct research tasks or visits, leading to increased burden and fatigue for participants and staff. These issues may raise substantial issues of equity and ability to recruit and enroll underrepresented groups into studies, particularly those of low socioeconomic status or living in rural areas. Given the already significant inequities in recruitment of underrepresented groups in palliative studies and generally worse outcomes, this issue merits particularly careful consideration. 12 On the positive side however, the Federal Communications Commission has just approved a substantial new broadband subsidy program, which includes a device purchase subsidy that may address some access issues. 13 Our findings highlight the need for clinical trials utilizing technology-based interventions to ensure they have adequate resources (time, training, or technology itself) to support all participants. Additional resources may help avoid increasing inequities through barriers to participation in research using innovative technology-based interventions. Another approach that could increase representation given technology barriers is to apply for waiver of written authorization and documentation of consent so that telephone consent can be obtained. Especially for remote or telephonic studies that are minimal risk this may be a feasible option to increase access to participation without affecting the rights and welfare of participants. Funders, including NIH, may want to consider supplemental funding for clinical trials that include a substantial proportion of persons who are less likely to have access or proficiency with technology. Both studies focus on underserved and underrepresented populations and were designed for an in-person intervention delivered in a home-based setting. Other projects, e.g., ENABLE trials, 14, 15 have effectively utilized telephone-based interventions. Our findings raise concerns that the necessary shift to remote study recruitment and intervention delivery may have impacted both study teams' ability to reach the most vulnerable patients. The impact of study adaptations on respective trial outcomes will need to be addressed in the analyses. We detail the experiences of only two clinical trials and cannot generalize the impact COVID-19 pandemic had on research, in general. The intent was to provide an in-depth description of adaptation to the pandemic from two geographically, methodologically diverse studies that include vulnerable or underserved populations, exploring differing and similar lessons and themes from the research experience. With several highly effective vaccines now available, it is possible to consider a landscape beyond the COVID-19 pandemic. Converting future communitybased, in particular home-based, palliative trials to remote recruitment, enrollment, and data collection processes may increase efficiency and reduce costs. Yet, depending on the population and technologies used, using remote technologies may increase inequities in palliative research participation. Researchers should very intentionally consider, both the necessity and accessibility of various technologies and how research staff are trained to ensure the greatest possible opportunity to recruit groups who have traditionally been underrepresented from research studies. Protocol for an embedded pragmatic clinical trial to test the effectiveness of Aliviado Dementia Care in improving quality of life for persons living with dementia and their informal caregivers. Contemp The Quick Dementia Rating System (Qdrs): a rapid dementia staging tool Research electronic data capture (REDCap)−a metadata-driven methodology and workflow process for providing translational research informatics support 3D-CAM: derivation and validation of a 3-minute diagnostic interview for CAM-defined delirium: a cross-sectional diagnostic test study Screening for delirium using family caregivers: convergent validity of the Family Confusion Assessment Method and interviewer-rated Confusion Assessment Method Apoyo con Carino (support with caring): RCT protocol to improve palliative care outcomes for Latinos with advanced medical illness A practical tool to identify patients who may benefit from a palliative approach: the CARING criteria Utility of the "surprise" question to identify dialysis patients with high mortality Living long in fragile health: the new demographics shape end of life care Hospitalizations for chronic disease and acute conditions in the time of COVID A third of Americans live in a household with three of more smartphones. Pew Research Center Disparities in palliative and hospice care and completion of advance care planning and directives among non-Hispanic blacks: A scoping review of recent literature Approves a $50 monthly high-speed internet subsidy. The New York Times Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: the Project ENABLE II randomized controlled trial Early vs. delayed initiation of concurrent palliative oncology care: patient outcomes in the ENABLE III randomized controlled trial Funding for the DSM-H trial is from the National Institutes of Health, National Institute on Aging, R01AG056610.Funding for Apoyo research is from the National Institutes of Health, National Institute of Nursing Research, 1R01NR0164671.All authors have met authorship guidelines, and they do not have any conflicts of interest to report. Supplementary material associated with this article can be found in the online version at doi:10.1016/j. jpainsymman.2021.06.017.