Conducting Biobehavioral Research in Patients With Advanced Cancer: Recruitment Challenges and Solutions

Abstract

Despite significant advances in cancer treatment and symptom management interventions over the last decade, patients continue to struggle with cancer-related symptoms. Adequate baseline and longitudinal data are crucial for designing interventions to improve patient quality of life and reduce symptom burden; however, recruitment of patients with advanced cancer in longitudinal research is difficult. Our purpose is to describe challenges and solutions to recruitment of patients with advanced cancer in two biobehavioral research studies examining cancer-related symptoms. Study 1: Symptom data and peripheral blood for markers of inflammation were collected from newly diagnosed patients receiving chemotherapy on the first day of therapy and every 3–4 weeks for up to 6 months. Study 2: Symptom data, blood, and skin biopsies were collected from cancer patients taking epidermal growth factor receptor inhibitors at specific time points over 4 months. Screening and recruitment results for both studies are summarized. Timing informed consent with baseline data collection prior to treatment initiation was a significant recruitment challenge for both the studies. Possible solutions include tailoring recruitment to fit clinic needs, increasing research staff availability during clinic hours, and adding recruitment sites. Identifying solutions to these challenges will permit the conduct of studies that may lead to identification of factors contributing to variability in symptoms and development of tailored patient interventions for patients with advanced cancer.

Despite significant advances in cancer treatment and symptom management interventions over the last decade, patients continue to struggle with cancer-related symptoms. Generally, individuals with advanced cancer receive some type of anticancer treatment for the rest of their lives; however, the same anticancer treatments that are meant to extend lives and decrease symptoms caused by the tumor(s) often have their own side effects that can interfere with patients’ physical and emotional health (Kroenke et al., 2010; Singer, Das-Munshi, & Brahler, 2010) and derail chronic treatment goals (Joshi et al., 2010; Tofthagen, Donovan, Morgan, Shibata, & Yeh, 2013). Furthermore, the type and intensity of symptoms that patients experience are highly variable and often depend on cancer histopathology, cancer treatment, stage of cancer, preexisting patient chronic conditions, or a combination of these factors (Reilly et al., 2013). Moreover, patients often experience multiple, concurrent symptoms (Dodd, Miaskowski, & Lee, 2004; Miaskowski, Dodd, & Lee, 2004) that can be highly distressing, can interfere with quality of life (QOL; Berger & Mitchell, 2008; Bower, 2014; Byar, Berger, Bakken, & Cetak, 2006), and may lead to increased health-care utilization (Barcenas et al., 2014). Finally, predicting symptom occurrence, severity, and distress for individual patients has proven to be extremely difficult (Koleck & Conley, 2016; Miaskowski et al., 2014).

Compelling evidence indicates that interactions among cognitive, biological, and behavioral responses to treatment may be a source of variability in patients’ symptom experiences (Cleeland, Fisch, & Dunn, 2011); therefore, biobehavioral research aimed at understanding links among psychological (cognitive), behavioral, and biological factors (Greenberger, Yucha, Janson, & Huss, 2007) and how those links influence symptom occurrence, severity, and distress is crucial for patients with cancer. Adequate baseline and longitudinal data are necessary components of the biobehavioral research required for designing interventions to improve patient QOL and reduce symptom burden; however, recruitment of patients with advanced cancer for longitudinal research is difficult. The purpose of this article is to review some of the challenges we have encountered in recruiting and enrolling patients with advanced cancer in biobehavioral research as well as possible solutions to these challenges.

Background

Patients with advanced cancer may experience significant symptom burden from the tumor(s), the cancer treatments, and the psychological stress of facing a serious medical diagnosis (Reilly et al., 2013). For example, tumors can put physical pressure on nerves causing pain and neuropathy as well as release inflammatory cytokines which are associated with increased fatigue, depression, and pain (Reyes-Gibby et al., 2008). Cancer treatments are associated with their own set of distressing symptoms. For example, chemotherapy can cause nausea and vomiting, taste changes, and loss of appetite (Brant et al., 2011). Radiation therapy is associated with increased fatigue and uncomfortable skin reactions (Bray, Simmons, Wolfson, & Nouri, 2016). Targeted therapies such as those that inhibit the epidermal growth factor receptor (EGFR) can cause an itchy, painful rash (Lacouture et al., 2011; Solomon & Jatoi, 2011; Wickersham et al., 2014) and diarrhea. Finally, the stress response associated with living with cancer can lead to chronic activation of the hypothalamic–pituitary–adrenal and sympathetic–adrenal–medullary axes, ultimately leading to a dysfunctional inflammatory response and elevated anxiety, depression, and pain (Subnis, Starkweather, McCain, & Brown, 2014).

Research using biobehavioral methods to describe cancer symptom mechanisms and to improve symptom management has exploded in the last 10 years. There is a growing understanding of how combining objective biological measures with subjective reports of cancer-related symptoms yields a more complete picture of the nature of cancer symptoms. Cleeland’s model of cancer symptom science (Figure 1) depicts the trajectory of cancer diagnosis, treatments, and the likely pattern of symptom development. In addition, this model includes recommendations for optimal timing to measure biomarkers and patient-reported data (Cleeland et al., 2011).

Figure 1.

Cleeland’s model of cancer symptom science depicts the trajectory of cancer diagnosis, treatments, and the likely pattern of symptom development. DNA = deoxyribonucleic acid; fMRI = functional magnetic resonance imaging; MDASI = MD Anderson Symptom Inventory; PET = positron emission tomography; RNA = ribonucleic acid. Reprinted from Cleeland et al. (2011).

For patients with advanced cancer, symptoms such as pain, fatigue, and skin and hair changes can be extremely upsetting because they hold significant cultural meaning in the context of having a potentially life-threatening diagnosis. For example, the rash patients can experience while receiving targeted therapy with an EGFR inhibitor is often not only painful but also socially isolating due to its visibility (Boone et al., 2007; Joshi et al., 2010; Wickersham et al., 2014), especially for individuals in the workforce. Others may interpret experiencing pain and/or fatigue as a sign of disease progression (Cohen et al., 2004; McCutchan, Wood, Edwards, Richards, & Brain, 2015). Furthermore, not all patients with advanced cancer experience the same symptoms or similar symptoms at the same severity, which makes it difficult to establish treatment plans to reduce symptom burden for these patients.

To develop targeted interventions aimed at improving patient QOL and reducing symptom burden, longitudinal research examining interactions among cognitive, biological, and behavioral variables over the course of treatment is needed with the goal of describing mechanisms underlying symptom generation and exacerbation. In addition to development and implementation of tailored symptom management interventions for patients with cancer, a better understanding of factors that predict subgroups of patients at risk for different symptoms based on their clinical phenotypes is critical. To achieve these two goals, it is necessary to have comprehensive baseline data of key variables (e.g., biological samples, symptom reports, measure of perceived stress) from prior to the start of cancer treatment rather than from after treatment has ensued, so that symptoms can be prevented or mitigated before they occur or become severe.

For all of these reasons, recruiting research participants at the time of a new diagnosis, prior to the start of treatment, and keeping them engaged in the study during treatment is critical in the conduct of biobehavioral research in patients with advanced cancer; however, research in this patient population is fraught with challenges, particularly with regard to recruitment and enrollment of participants. Previous reviews have focused on challenges in conducting research with individuals who are at the end of life (Mackin et al., 2009; Steinhauser et al., 2006) or are very old (Townsley, Selby, & Siu, 2005) or with individuals participating in clinical trials for treatment of cancer (Lara et al., 2001; Manne et al., 2015; Wright et al., 2004). Ammari, Hendriksen, and Rydahl-Hansen (2015) described recruitment challenges and reasons for nonparticipation in a family-coping-oriented palliative home care trial with a population recruited from medical, surgical, and oncology hospital wards in Denmark. Of the families who declined to participate, 65% cited one of the two main reasons: “burden of illness too great” and “too soon” to receive palliative support. Many of the same challenges are present in studying advanced cancer patients but with two critical differences: (1) the window of time for researchers to identify and enroll individuals newly diagnosed with advanced cancer is extremely short and (2) patients may already be participating in a clinical trial for treatment purposes, adding burden to their lives. Given the unique circumstances of patients newly diagnosed with advanced cancer and the critical importance of conducting prospective biobehavioral symptom research in this population, our purpose in this article is to describe challenges and solutions to recruitment and enrollment of patients with advanced cancer by illustrating two biobehavioral research studies in which we are examining cancer-related symptoms as exemplars.

Exemplar Studies

Exemplar 1

The purpose of the first study is to evaluate interactions of symptom severity, inflammatory response, cognitive appraisals, and coping approach over 24 weeks in a sample of older adults with advanced lung, colorectal, or pancreatic cancer. The combining of data on cognitive appraisals and coping with those related to inflammatory mechanisms of cancer-related symptoms provides a unique perspective on both the psychological and physiological experiences of symptoms in advanced cancer. The specific aims are to (1) describe the associations among inflammatory response (i.e., levels of peripheral interleukin 1, interleukin 6, and tumor necrosis factor α), severity of six common cancer-related symptoms (i.e., pain, fatigue, sleep disturbance, loss of appetite, depressed mood, cognitive dysfunction), cognitive appraisals, and coping (i.e., emotion-focused, problem-focused, and mixed focus) at baseline and over 24 weeks after diagnosis and (2) to evaluate possible predictors (e.g., demographics, type of cancer, cognitive appraisal, coping approach, and inflammatory response) of worse symptom severity scores at baseline and over 24 weeks after diagnosis. Patients are being recruited at the University of Iowa Holden Comprehensive Cancer Center (HCCC). HCCC is the only National Cancer Institute (NCI)-designated cancer center in the state of Iowa and serves a three-state area comprising both rural and urban residents. We hypothesized that cognitive appraisals and coping behaviors feed back into the immune response and contribute to changes in cancer-related symptoms over time. The institutional review board (IRB) of the University of Iowa approved this study.

Participant recruitment starts with patient identification via electronic health record (EHR) monitoring from four referring provider clinics. Once we identify patients as potentially appropriate for participation, we approach the attending oncologist to verify eligibility. A member of the clinic staff (e.g., oncologist, fellow, and nurse) then asks the patient for permission to have a member of the research team contact him or her to discuss the study. Once the patient provides permission, we contact the patient. If the patient agrees to participate, we obtain informed consent and enroll her or him into the study.

In the first 3 months of recruitment, we asked clinic staff members to assist with recruitment by discussing the study with eligible participants. Weekly meetings with the research assistant (RA) occurred to help remind the staff about the study. We recruited only two participants with this approach. In Month 4, we submitted a protocol modification to the IRB requesting access to the clinic schedule to prescreen participants for eligibility. Following this modification, enrollment improved significantly from one enrolled participant per month to one per week.

We see each participant for data collection at each chemotherapy visit until treatment concludes or the sixth study visit is completed, whichever comes first. On the day of the first chemotherapy infusion, we collect baseline data including a blood draw for inflammatory cytokines and questionnaires covering demographics, symptom severity, cognitive appraisal, and coping. A RA meets participants at the clinical phlebotomy laboratory for the blood draw. Research laboratory tests are drawn at the same time as clinical labs to eliminate multiple venipunctures. The RA then gives participants the questionnaire packet to complete between appointments (e.g., lab, oncologist, and infusion clinic) and returns to the infusion suite later in the day to pick up the questionnaire from the participant and review it for completeness or, if needed, stay with the participant to assist with the questionnaire packets.

During the first 3 months of data collection, we gave questionnaire packets to participants when they checked in for their infusion appointment rather than at the laboratory. However, participants frequently returned the packets with incomplete data. When we asked them why, participants explained that they were confused by certain items and/or were getting tired during infusion (possibly from the antihistamine premedication prior to chemotherapy). In response, we changed the procedure in three ways: (1) We order the packet so that the most important questionnaires were at the front of the packet and put the questionnaires on demographic and covariates into separate packets so that participants could start them at baseline and complete them at a separate visit, if needed; (2) we provide participants with the packets at the laboratory rather than the infusion appointment, so that participants could start them prior to the administration of chemotherapy premedication that causes drowsiness and work on them between clinic appointments, potentially reducing the monotony and fatigue associated with working on them continuously; and (3) an RA checks the packet for completeness when the participant finishes and is available to assist the participant with packet as needed (e.g., reading the questions out loud and explaining confusing items). These strategies have dramatically improved the quality and completeness of the data collected.

Exemplar 2

The purpose of the second study is to examine genetic, clinical, and biomarker correlates of EGFR-inhibitor-related rash to better understand why some, but not all, patients with advanced cancer who receive EGFR-inhibitor treatment develop a painful rash. The rash affects the face, neck, scalp, and upper trunk (Lacouture et al., 2011; Solomon & Jatoi, 2011; Wickersham et al., 2014) and is physically and emotionally painful (Lacouture et al., 2011; Perez-Soler et al., 2005; Wickersham et al., 2014; Wong et al., 2010), making it difficult for individuals to adhere to therapy (Boone et al., 2007; Wagner & Lacouture, 2007). Therefore, understanding who is at risk of developing the rash and how severely they are likely to develop it may help to focus intervention development in this area. The specific aims are to test whether (1) circulating soluble EGFR levels are associated with development and severity of EGFR-inhibitor-related rash and (2) a unique transcriptional signature in fibroblasts distinguishes, at a molecular level, patients who develop the rash from those who do not.

Participants in Exemplar 2 are being recruited from two cancer centers in the Baltimore, MD, area: (1) The University of Maryland Marlene and Stewart Greenebaum Comprehensive Cancer Center (GCCC), an NCI-designated comprehensive cancer center that defines its primary catchment area as central Maryland, which includes Baltimore city and the six surrounding counties and services the central Maryland region (primarily urban and suburban) and (2) the University of Maryland St. Joseph’s Medical Center Cancer Institute, a regional medical center in the Baltimore area and a recognized affiliate of the GCCC. The IRB of the University of Maryland approved this study. Receipt of a Health Insurance Portability and Accountability Act (HIPAA) waiver from the University of Maryland IRB has allowed initial screening to be performed by a research team member via electronic medical record review. We review the schedules of seven oncologists and two nurse practitioners weekly for potential study participants according to IRB-approved inclusion and exclusion criteria. Once we identify a potential participant, we contact the oncologist or nurse practitioner who then introduces a research study staff member to the potential participant to further discuss the study. We also post IRB-approved study advertisements in waiting areas of the cancer centers and in Baltimore-area pharmacies, such as CVS^® and Rite Aid, public libraries, and community centers/senior centers.

This study has been open and recruiting participants since February 2014. As of May 2016, of the patients who were approached for potential participation in the study, 23 had agreed and consented to participate and 23 had declined. Of the 23 who consented, 4 withdrew before completing study procedures and 19 completed study procedures.

Recruitment Challenges for Both Exemplars

The recruitment rates for Exemplar 1 and Exemplar 2 (i.e., number who agree to participate over the number of eligible patients screened) have been fairly similar. Exemplar 1 had a recruitment rate of 11.7% over a period of 18 months, and Exemplar 2 had an overall recruitment rate of 9.9% over a 2-year period. Exemplar 2 had both a longitudinal component and a cross-sectional component; the recruitment rate was 1.2% for the first cohort and 20.5% for the second. Recruitment challenges fell into two categories: clinician referral and timing of consent and/or baseline data collection. Table 1 provides a summary of recruitment challenges and potential solutions for both exemplars.

Table 1.

A Summary of Challenges for Recruiting Patients With Advanced Cancer for Studies on Cancer- or Treatment-Related Symptoms and Potential Solutions.

Challenge	Potential Solution
Clinician referral
Missed enrollment due to starting treatment prior to baseline data collection	Develop “flagging” system in electronic health record to alert clinicians that this person is eligible for clinical trial and to notify research personnel
Recruitment rates lower than projected	Reevaluate inclusion criteria to determine how they can be broadened. For example, add additional cancer types with similar disease trajectories and treatments Add an additional site for recruitment of participants Work with cooperative groups (e.g., Alliance and GOG), especially when studying a relatively small patient population
Burden on clinical staff	Obtain HIPAA waiver to allow initial screening via electronic health record Tailor recruitment strategies to fit the unique needs of each clinic
Clinicians forget details of the protocol	Conduct a study-initiation visit to review scientific background, study purpose, and participant entry requirements Provide the oncologist with a brief, bulleted summary of the study including inclusion/exclusion criteria and participation requirements immediately before he or she enters the patient’s clinic room (i.e., research personnel present in clinic to answer questions) Identify a clinical champion who has a broad understanding of the study including short-term and long-term goals of the research Consider using frontline staff to offer study materials to all eligible patients upon check-in to clinic appointments, followed by a phone call by the research staff
Clinician gatekeeping (e.g., limiting access to potential study participants)	Provide in-service to staff about the study purpose, design, and methods and address their concerns about the impact of study on patients, staff, clinic flow, etc.
Timing of consent and data collection
Timing of a new diagnosis, treatment options and decision, and potential clinical trial enrollment for treatment with your additional study evaluating symptoms related to treatment	“Piggyback” onto an existing clinical trial that has similar data collection time points Collaborate/coordinate efforts with the clinical trials research office
Timing of informed consent and collection of baseline data with treatment initiation	Flexibility in location and timing of completion of informed consent and data collection (e.g., clinic, participant’s home, or other private location)
Patients are unable to decide about participation because they overwhelmed with amount of information presented about diagnosis and treatment at initial clinical appointments	Utilize telephone reminders after initial contact to review questions or concerns about the study Develop opt-in rather than opt-procedures Develop recruitment materials for audiovisual learners along with written materials
Family concerns about patient participation	Schedule time to speak with patient and family together about the study when they are not in a rush or focused on other appointments Emphasize potential benefits (if there are any) and benefits to future patients and families like them Consider using other sources of data whenever possible (e.g., validated proxy measures, medical recorder data, clinical trial measures) to minimize perceived survey burden

Note. Alliance = Alliance for Clinical Trials in Oncology; GOG = Gynecology Oncology Group; HIPAA = Health Insurance Portability and Accountability Act.

Clinician Referral

The primary challenges for recruitment of participants into both studies included relying on clinicians to determine patient interest in the study and obtain permission for a researcher to contact the patient to provide more information about the study. While clinician recommendation does increase the chances that a patient will agree to participate in the study (Eggly et al., 2008), realistically, the health-care professionals in oncology clinics are extremely busy with little time to accomplish all necessary tasks during a short patient visit much less explain and recruit for a study. One strategy we utilized in both studies was obtaining IRB approval to prescreen medical records for eligibility. With an HIPAA waiver, we are able to determine which patients appeared to be eligible based on electronic chart review. Then, we contact the oncologist and clinic nurse indicating that an individual might be eligible for the study and providing the date and time of that individual’s next clinic appointment. Finally, a research team member is available in the clinic on the specified date and time of the patient’s clinic visit. Providing a specific name of a potentially eligible participant and the date and time of their next appointment and being present (but out of the way) on the day of the appointment significantly increases the likelihood that providers will ask eligible patients about their interest in our studies. In addition, Exemplar 1 researchers provide the oncologists with an informational flyer with bullet points describing key aspects of the study just before they enter the patient room. This strategy is effective because it removes responsibility from clinic staff to remember study details such as the inclusion criteria, study participation requirements, and why the patient is being asked to participate. The principal investigator (PI) of Exemplar 2 held a study-initiation meeting at the cancer centers to review the scientific background, study purpose, and participant entry requirements for all oncologists and nurse practitioners prior to opening the study and holds additional meetings periodically during the study to remind clinicians of the study and to update them on enrollment.

Timing of Consent and Data Collection

Second, the timing of informed consent has presented challenges for both Exemplars. Baseline blood samples drawn prior to initiation of chemotherapy or EGFR-inhibitor treatment are required for both, so the informed consent process has to occur before therapy is administered. The combination of a new diagnosis and learning about treatment options and potential clinical trial enrollment for treatment is overwhelming for patients. Approaching these patients for interest in an additional study evaluating symptoms related to treatment tends to compound the feeling of being overwhelmed and dealing with “too much.” This issue may arise especially in patients newly diagnosed with advanced cancers, as they may rapidly become too ill for full study participation and are often too anxious to fully comprehend consent to a research study (Steinhauser et al., 2006).

In Exemplar 1, the change in clinical status common in patients with advanced cancer presents a persistent enrollment challenge. Patients identified on the clinic schedule as likely to be eligible could be suddenly hospitalized with chemotherapy started over the weekend, resulting in a missed enrollment. Another common problem has been changing clinics. Eligible patients might be seen at University of Iowa Hospital and Clinics (UIHC) for a second opinion and/or to start chemotherapy but will then return to their home communities many hours away from UIHC. Figure 2 depicts the most common reasons for missed enrollment opportunities. The research team developed a detailed tracking sheet and monitoring system that involved daily checks of the EHR to help overcome this challenge.

Figure 2.

Most common reasons for missed enrollment opportunities in Exemplar 1.

In addition to the challenges we experienced in the exemplars, challenges to recruitment identified in studies of elders at the end of life (Mackin et al., 2009) are pertinent to studies of patients with advanced cancer. Clinicians acting as “gatekeepers” in an effort to either protect or advocate for their patients or, conversely, approaching only those they feel would be amenable to participation may be a significant barrier to recruitment (Mackin et al., 2009; McMillan & Weitzner, 2003; Steinhauser et al., 2006). Similarly, well-meaning family members may prevent patients with advanced cancer from joining research studies or otherwise interfere with the recruitment process (Mackin et al., 2009). The common belief that family members expressed in the literature (Mackin et al., 2009) that “[patients] have too much on their plate right now” was consistent with our experience in the exemplars. Finally, patients may be unable to decide about participation due to the short window of time between diagnosis and beginning anticancer treatments.

Solutions to Recruitment Challenges

Based on our experiences in the two exemplars and those that other researchers describe in the literature among similar populations (Boland et al., 2015; Bruner & O’Mara, 2014; Mackin et al., 2009; McMillan & Weitzner, 2003; Steinhauser et al., 2006; Treweek et al., 2010), investigators can implement a number of strategies to address the challenges described earlier.

Make It Easy for the Referring Clinician

For both studies, we worked with the appropriate IRBs to secure permission to screen clinic schedules rather than rely exclusively on clinic staff to refer potential participants to the research teams. This strategy is especially important for small pilot projects that may not have funds available to pay clinical research coordinators in the cancer center to assist with participant recruitment and enrollment. In addition, it is important to tailor recruitment strategies to fit the personalities of the clinicians and the unique needs of each clinic. For example, offering flyers and handouts with bulleted key points of the study was helpful for Exemplar 1. Alternatively, Steinhauser et al. (2006) first determined which participants were eligible via chart review then mailed a recruitment letter signed by the attending physician. Whenever possible, predetermination of participant eligibility is helpful, so that the clinician only has to ask the patient if she or he is interested in learning more rather than worrying about potential eligibility. Some clinicians prefer that research staff be on hand ready to talk to the patient during the visit. In these cases, maintaining study staff visibility in the clinic while staying out of the way can help clinicians remember the study. Periodic meetings with the referring clinicians to evaluate and troubleshoot participant recruitment and enrollment strategies is also key for meeting the challenges of particular study sites.

If clinician “gatekeeping” is present, research teams can provide in-service education to the staff about the purpose of the study. Mackin et al. (2009) noted that describing the study purpose, design, and methods and offering adequate time to listen to and address the staff’s concerns about the impact of the research on patients, families, staff, and clinic flow can significantly improve staff attitudes about the study. In addition, investigators can consider identifying a “clinician champion” who has a broad understanding of the study, including short-term and long-term goals of the research and will promote a positive attitude about the study.

Make It Easy for Patients and Families

Telephone reminders can be an effective tactic for addressing questions or concerns that have arisen about the study since the initial contact with the participant (Mackin et al., 2009; Steinhauser et al., 2006). Opt-in rather than opt-procedures that require less effort from the patient and family may also aid recruitment of participants into biobehavioral studies (Boland et al., 2015; Mackin et al., 2009; Steinhauser et al., 2006). Recruitment materials geared toward audiovisual learners can help overwhelmed patients and families who may be experiencing cognitive fatigue and are not able to focus on written materials (Treweek et al., 2010). Finding a time to speak with patients and families together about the study when they are not in a rush or focused on other appointments can counteract a default position of declining “extra things” because their minds are preoccupied (Mackin et al., 2009). Emphasizing the potential benefits (if there are any) of study participation and benefits to future patients and families like them may also be helpful. Finally, using sources of data other than those collected directly from participants whenever possible (e.g., validated proxy measures, medical record data, clinical trial measures) may help to minimize perceived study burden (Boland et al., 2015; Treweek et al., 2010).

Be Ready to Act Quickly

Research staff should be readily available during clinic hours to mirror times that patients may be in the clinic. This goal can be a significant challenge for a pilot study with a small budget, however. If this schedule is not feasible, following up on prospective patients several times per week is an alternative option because the patient’s schedule leading up to a treatment visit can change suddenly.

Strategize With Others Conducting Research

Investigators should work with clinical trials coordinators in the medical center oncology program to coordinate research efforts. For example, coordinators might be able to help by suggesting a better way to approach patients or how to phrase the description of the research study to potential participants? They will also know whether there is another clinical trial being performed at the cancer center that has a similar patient population or visit schedule as the study under discussion. In Exemplar 2, we have used this strategy with some success. The PI of that study coordinates with a treatment trial at one clinic that is enrolling patients with similar inclusion and exclusion criteria and a similar study visit schedule. With the permission of the oncologist and the hospital research coordinator, we approach patients about participation in Exemplar 2 after they have been deemed eligible for the treatment trial.

In addition, the field of oncology nursing research has recently increased its use of cooperative groups (such as Alliance and Gynecological Oncology Group) to conduct multisite studies. This strategy is particularly useful when studies involve a relatively small patient population. While start-up procedures may seem to take longer, an added benefit of cooperative groups is the study support and infrastructure they provided (Bruner & O’Mara, 2014).

Develop a Flexible Protocol That Is Responsive to Patient Needs

Researchers should allow for flexibility in consenting and data collection procedures whenever possible. For example, it may be more feasible for some participants to provide consent over the phone to facilitate data collection procedures upon arrival at the clinic. In Exemplar 2, we offer to complete as many study procedures as possible (including blood draws or collection of biological specimens) at the patient’s home or a location that is convenient and private for the patient to facilitate data collection. In addition, consideration of other health-care venues from which to recruit patients with advanced cancer can be advantageous. For example, other researchers have recruited patients with advanced cancer from a specialty pharmacy clinic to improve recruitment/enrollment, particularly for patients taking oral chemotherapy or oral targeted therapies (Spoelstra et al., 2015).

Limitations

Readers should consider our presentation of challenges and potential solutions to recruitment difficulties in biobehavioral research among patients with advanced cancer with several limitations in mind. First, for our exemplars, we recruited participants from two specific NCI-designated cancer centers and one affiliated regional cancer institute. NCI-designated cancer centers are recognized for their scientific excellence and leadership and focus on the integration of research with patient care. As part of their accreditation, NCI-designated cancer centers conduct centralized scientific review of research protocols, have an IRB to review and monitor the protection of human research participants, and benefit from other resources such as a computerized database for collection and storage of research participant data. Community cancer centers and clinics may not have access to such resources; for example, obtaining an HIPAA waiver for prescreening potential participants via the electronic record may not apply for some private practice clinics in the community. Second, both of the exemplar studies have small sample sizes and are being conducted to plan for future, large-scale studies. Finally, the participants in these studies were involved in active treatment rather than being cancer survivors in the traditional sense (i.e., completed curative treatment), which limits the generalizability of these challenges and solutions. The reasons patients on active treatment have for joining a study about cancer symptoms may be very different from the reasons of those who have completed treatment but are being followed over time in the clinic for surveillance.

Conclusion

Researchers and clinicians are increasingly recognizing the importance of exploring patient-reported outcomes in combination with biomarkers for understanding the complex experience of cancer- and cancer-treatment-related symptoms. Biobehavioral research is needed to better understand these experiences and to identify patients at greatest risk for developing distressing symptoms, but recruitment, enrollment, and retention of patients with advanced cancer in longitudinal biobehavioral research are difficult. Such studies call for the collection of comprehensive data on key variables (e.g., biological samples, symptoms, and perceived stress) before, rather than after, the initiation of therapy. Early identification and negotiation of barriers to recruitment of participants with advanced cancer may help improve collection of robust baseline data for studies aimed at identifying factors contributing to variability in cancer-related symptoms. Future research concerning the experiences of patients who refuse to participate in clinical trials because they are feeling too ill, of patients living in underserved areas, and of patients who do not have the financial or instrumental resources to participate in research at academic medical centers is needed to comprehensively characterize symptom experiences and maximize participation of patients with advanced cancer in biobehavioral research.