Recruitment problems in psychosocial oncology research

Faculty of Psychology and Educational Science, Open University of the Netherlands, Heerlen, the Netherlands Department of Health Psychology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands Department of Medical and Clinical Psychology, Tilburg University, Tilburg, the Netherlands Department of Psychiatry, Amsterdam Public Health Research Institute, VU University Medical Center, Amsterdam, the Netherlands

patient recruitment problems, psychosocial cancer research, adaptive distress, maladaptive distress, desire for help In our professional experience with conducting psychosocial cancer research, we have all experienced problems with recruiting cancer patients and survivors for participation in outcome research. As researchers, we work in a highly competitive environment in which many compete for scarce resources. We sometimes end up submitting ambitious grant proposals whose feasibility, in patient recruitment, may be less than optimal. For instance, in 2013, both van Lankveld and colleagues and Schroevers and Fleer and colleagues received funding for innovative and methodologically rigorous (ie, following Consolidated Standards of ReportingTrials criteria) intervention studies. In both studies, the inclusion rate was disappointingly low (both lower than 1%), even though power calculations were based on prevalence rates of psychosocial problems reported in the recent literature. Similar problems are probably experienced by many fellow researchers, and have also been reported or hinted at in some publications, eg, Fredman 1 and van Scheppingen. 2 Our goal here is to make a plea for more extensive reporting about this issue, both in the interest of minimizing the risk of publication bias in the field of psychosocial oncology in case study results would remain unpublished, and to allow colleagues to learn from these negative experiences, and thus to avoid wasting additional investigational resources.
We believe that many of the problems experienced in recruiting patients into our studies reflect a lack of awareness among researchers of barriers to recruitment, that result from a mismatch between patient characteristics, the types of intervention that we offer, and the ways we select and approach potential research participants. Patient characteristics include, for example, their felt desire for help, personality characteristics, and their preference for specific treatment type. Types of intervention include, for example, individual vs group treatment and cognitive behavioral vs supportive-expressive approaches. We are convinced that both methodological and patientrelated factors contribute here.
Methodological factors include the use of screening instruments that have low predictive validity and do not succeed in identifying those patients who desire supportive psychosocial care, as well as inefficient procedures for approaching cancer survivors. An example of the first factor was provided by van Scheppingen and colleagues 2 who conducted an RCT investigating the efficacy of problem-solving therapy for cancer survivors. They reported that screening based on symptom checklist scores and 1 additional question on need for services proved unsuccessful for recruiting sufficient cancer patients in need of care.
An example of the impact of procedural characteristics on recruitment rates was published by Fredman and coworkers. 1 They investigated a couple-based intervention aimed at relational enhancement for breast cancer survivors. Couples were more likely to participate when they were contacted at home or at a follow-up appointment at the cancer clinic, compared to when they were asked to participate when first diagnosed. Shorter geographical distance between home and treatment facility location also increased the likelihood of participation. Another example of the latter point was given by Rabin and colleagues. 3 They observed significant differences among young adult cancer survivors in enrollment rates for an exercise intervention study between personal recruitment at a clinical facility and at cancer-related social events, and when using nonpersonal procedures via postal mail, telephone, advertisements on the internet, radio, television, social media, or other means of contact. Recruitment including personal contact in an oncology clinic yielded the highest inclusion rates, although the costeffectiveness of nonpersonal recruitment strategies was higher.
Patient-related factors are also related to diminished study participation in psychosocial oncology research, including patient characteristics, such as both younger and older age, 3  A possible explanation for the discrepancy between distress and service uptake is that the presumed link between cancer-related distress and patients' desire for professional help is not based on emotion theory and the relationship between emotion and psychopathology. 8 Dekker and colleagues argued that it is not the intensity but the nature of emotional responses which determines whether emotions are maladaptive. Using the results of research on the relationship between emotion dynamics and psychopathology, 9 they argued that emotional responses, reflected in elevated distress scores, often serve an adaptive purpose. Emotional reactions to cancer and cancer treatment only become maladaptive when they linger and perpetuate over time, or when they are extreme and unstable. 9 Emotional responses may also become maladaptive when they hinder adaptive coping. Dekker and colleagues 8 further argued that patients' need for psychosocial care depends on whether emotional responses are adaptive or maladaptive.
Patients who experience maladaptive emotional responses seem to be in need of professional mental health care, defined as psychotherapy, pharmacotherapy, or emergency psychiatric care. In contrast, patients who experience adaptive emotional responses do not seem to be in need of professional mental health care. Instead, they may be primarily in need of support from relatives, friends, and primary caregivers (ie, doctors and nurses) and low-intensity psychosocial interventions. Further research on the distinction between adaptive and maladaptive emotional responses and on tailoring of interventions to the needs of patients with cancer is needed. Such research has now begun to emerge. 5 In conclusion, improving the recruitment of patients to psychosocial intervention studies that target emotional needs can be facilitated by carrying out feasibility studies that assess both the prevalence of problems in the target population and the interest in and willingness of patients to receive professional psychosocial care. Estimates of perceived symptom burden and functional impairment alone can result in serious overestimates of uptake in psychosocial intervention studies.

CONFLICT OF INTEREST
None of the authors have received financial support for the preparation of the manuscript or have any other conflict of interest to disclose.

Key Points
• Patient-reported distress does not necessarily reflect the patient's actual desire for help.
• Distress may be either adaptive and helpful, or maladaptive.
• Feasibility studies should assess interest and willingness of patients to receive professional psychosocial care.
• Research on optimal recruitment strategies will improve enrollment in patient-based research in psychosocial oncology.
• Greater reporting of problems with patient recruitment in clinical outcome studies in the field of psychosocial cancer treatment will help minimizing publication bias. VAN