Analytical approach

Over the course of the four-year study, the research team systematically gathered descriptive data on recruitment methods, enrollment figures, and frequencies and corresponding percentages. The data were displayed in table format to depict variations in recruitment sites, clinical environmental contexts, strategies, over time related to pre, during, and post COVID-19 timing. Graphical techniques were employed to illustrate the flow of time series data [8]. The research team met weekly to record and audit the recruitment approaches from each clinical site documented in data files and tables of participant enrollment by calendar months. Rigor was maintained in data collection and analysis by holding weekly team meetings using a systematic meeting procedure that was recorded in the meeting minutes, followed by a review of the data and by a discussion of new recruitment approaches [7]. Additionally, to address Aim 3 of the original RCT which focused on evaluating the implementation of by characterizing patient- and provider-level facilitators and barriers, virtual focus groups were conducted with clinical staff who participated in recruitment. These discussions explored facilitators and barriers impacting both the implementation process and outcomes. IRB approval was obtained for these qualitative aspects of the study as well (ID: IRB0007275). Research staff contacted relevant clinical staff by email. Study information was sent to those agreeing to participate in the focus group. Verbal consent was obtained from the clinical staff at the beginning of each focus group after questions were answered and was documented on the transcripts. Due to the COVID-19 restrictions, focus groups were accomplished near the end of the study. Thus, we obtained retrospective views from five focus groups from 10 clinic staff from 4 clinics using thematic analysis [9]. Final case analysis was a convergence of evidence triangulated from team observations, documented in data tables, flow diagrams, archival records, and focus group themes.

Chronological strategies report

From June 2019 to October 2022, the research team enrolled a heterogenous sample of cancer survivors with insomnia symptoms from multiple clinical sites, referral sources, and outreach. Cancer survivors were identified as individuals who were ≥ 1 month after treatment (hormones and targeted therapies were acceptable) for stages I – III of either breast, prostate, colorectal or lung cancer types. Insomnia symptoms were defined as those that met the criteria by scoring above 7 on the Insomnia Severity Index (ISI) [10].

Overall, the participants (n = 136) mean age was 63.82 years [SD = 9.98] and were female (55.9%). Breast cancer comprised 45.6% of the sample, followed by prostate, colorectal and lung cancer (35.3%, 14%, 5.1%, respectively). The majority had stage I cancer (46.3%) followed by stage II (28.7%) and stage III (23.5%). Refer to Table 1. for detailed demographic and clinical characteristics.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Socio-demographic characteristics of the participants (N = 136).

https://doi.org/10.1371/journal.pone.0327806.t001

The initial recruitment target goal was 158 cancer survivors with a predicted attrition of 30%, leaving 110 participants (55 in each intervention arm) to have completed the primary endpoint, ISI at 1-month. Based upon power calculation, after 40 months recruitment, the study enrollment was complete with 136 participants with lower than anticipated attrition rate (12.5%). The CONSORT Diagram (S1 Fig) details the participant flow. Combined recruitment strategies yielded 402 potential participants of which 136 (34%) were enrolled from June 2019 to October 2022 and data collection was completed in 2023. Of the 136 participants screened, 132 were randomized into experimental or control arm, and 122 completed one month follow up (primary endpoint). The study personnel achieved a high retention rate of 87.5% by keeping the participants engaged.

Recruitment strategies summary.

Active in-person recruitment was the most effective strategy overall (n = 75/136, 55.1%). In-person approaches included the collaboration between the CCC and BSN research students yielding 5 enrollments. Due to limitations in in-person access during and after COVID-19, in-person recruitment continued to yield less recruitment overtime and the need for alternative strategies remained. The research team adding a print outreach including study flyers, newsprint and newsletters on social media strategies resulting in successful recruitment numbers n = 50/136, (36.8%). Research registries provided very few participants. Table 2 summarizes strategies, timing, and percentages per recruitment site. S3 Fig demonstrates chronological flow and variations across time.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Recruitment strategies utilized to enroll 136 participants during the study enrollment period.

https://doi.org/10.1371/journal.pone.0327806.t002

Focus group evaluation of program implementation

A series of five virtual focus groups of (1–3) participants with a duration of 20–45 minutes were held June 2023-October 2023 (IRB STUDY0007275) to obtain feedback from a purposive sample from 10 clinical staff (2 MD and 8 nurses) who participated in recruitment in the study. An expert (SD) in qualitative methods virtually conducted the virtual focus groups. Participants gave IRB verbal consent to the recording of discussion at the beginning of the meeting and were reminded of the nature of focus groups related to keeping confidential others information. Participants were asked (n = 1–3) to describe the barriers and facilitators for recruitment in their settings. Transcripts were de-identified and provided 50 pages of data for thematic analysis. The research team of 4 individuals trained in qualitative analysis coded individually and met as team to discuss emergent themes of the interview transcripts until reaching saturation. Themes emerged from the data. The team identified key themes reflecting recruitment barriers at the patient, provider, and clinic levels, highlighting time-constraints during clinic visits, with staff prioritizing focus on the patients’ cancer disease outcomes versus insomnia symptom burdens. Low staffing exacerbated issues with clinic staff workload. Facilitators for recruitment themes identified the helpfulness of research staff attending the clinic and flagging eligible patients; onsite research team availability during visits to address patients’ questions; and organizing informational lunches to keep clinic staff informed about the study. In the later stages of recruitment, research team delivered flyers to clinic staff with eligible patients named on each flyer as a cue for specialty clinic nurses prior to the visit. Table 3 delineated themes and related quotes.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Focus group themes and subthemes.

https://doi.org/10.1371/journal.pone.0327806.t003

Strategies

Initially, our research team did not meet the goal of 4 participants per month due to the time-intensive prescreening of eligible participants recruitment strategy at the CCC. When in-person recruitment was possible pre- and after COVID-19, recruitment required the research team’s time totaling one FTE per week. Screening medical records for insomnia required a more nuanced review and interview between the research team and the potential participants, thus nurse practitioners were the most effective research team member for screening. To address the issues of identifying potential participants who have insomnia, the research team members and clinical respondents suggested screening for sleep problems at the CCC by including a pre-appointment screening questionnaire (including problems with sleep) for patients as part of the appointment intake process [11–14]. However, the CCC would not facilitate the pre-visit assessment and clinic staff were overloaded often forgetting to do sleep assessments. Thus, recruitment approaches were needed that provided direct contact with potential participants to ask the question “are you having trouble with your sleep?” A number of studies [11,15,16] suggest a variety of recruitment strategies to mitigate low enrollment; however, COVID-19 slowdown created barriers to recruitment due to 8-month shutdowns and subsequent hesitancies. The research team found using print outreach including flyers and social media although health news sources to be particularly effective during the down time to broaden outreach to the community in general to those suffering from sleep issues. We also suggest employing community partnerships and assigning representative recruiters to assist in recruiting a broader representation of participants of the population. The teams’ continuous monitoring of recruitment and enrollment goals through ongoing data collection for this case study, facilitated a critical appraisal of strategies which facilitated reaching our enrollment goal.

Technological adaptations

Before COVID-19, participants preferred in-person interactions followed by email survey queries, but after COVID-19 there as a necessity to shift totally to virtual interactions. When in-person protocols were reinstated after vaccines became available, participants were given the choice of in-person or virtual with 60% opting for virtual interactions. These virtual in-person approaches allowed participants to still establish trust with the project manager, mirroring the importance of initial in-person communications promoting year-long study retention. Additionally, virtual outreach through health newspapers extended our geographic reach to potential participants who recognized their need for sleep interventions.

Environmental influences on recruitment

Identifiable issues emerging from the focus groups analysis focused on the struggle to keep CCC clinical staff committed to the recruitment process while recognizing their challenges in addressing ongoing patient concerns. Clinical staff prioritized concerns with disease symptoms, thus patient’s questions related to sleep were neglected. Although clinical providers voiced interest in helping recruit participants, they seldom mentioned the study to their patients due to other competing priorities. Time constraints of a busy clinic setting made their commitment to study recruitment less of a priority. For these reasons, it was imperative for research team personnel to be physically present in the clinic setting during the recruitment process. Potential participants from the survivorship clinic who had completed treatment had the advantage of having a prescreening quality of life survey that could identify sleep issues and were more willing to participate in clinical trials to mitigate troubling quality of life issues. After the COVID-19 lockdown ended and in-person visits resumed, many patients transitioned to telehealth clinic visits where the research team did not have screening access. Furthermore, themes from the focus groups clinicians reported that patients were worried about recurrence and waiting on scan results and did not prioritize insomnia. Some clinicians reported tending to refer patient with sleep problems to their primary care provider or to other support services such as acupuncture. Furthermore, suggesting sleep referrals may assist patients in need.

Limitations and future directions

Despite the strengths and valuable insights that the current study provides, there are several limitations that should be considered. First, despite efforts to obtain a diverse sample of cancer survivors, the participants in our study were primarily non-Hispanic White individuals from a targeted geographic area (Western New York), which may limit the generalizability of the findings. However, we have provided detailed descriptions of the real-life context in which this study was conducted – specifically, the conditions imposed by the COVID-19 pandemic and related shutdown measures. Including such contextual information may enhance the transferability of our findings, as it allows readers and researchers to understand and adapt the intervention approach and results to similar real-world scenarios. Additionally, the participants were largely skewed towards either breast or prostate cancer types and may not accurately reflect the views and experiences of individuals from a heterogeneous sample of cancer survivors. During recruitment at the survivorship clinic, there was a high interest by survivors with gynecology and head-and-neck cancer types. Interest by these groups should be further explored for future trials. Second, the reasons that eligible participants declined to take part by recruitment strategy were not formally collected, which would have further aided analysis of the approach. Third, we found that some participants hesitated in joining in the study upon learning about the two arms in the study. Future research should also consider providing an option for cross over design for those randomized to the control arm after the study is complete.

To address validity threats in this case study analysis, we employed rigor by using a systematic procedure in a time series data collection approach that explicated recruitment outcomes with documentary evidence (i.e., using the consent dates). Further, we operationalized recruitment measures with flow charts, timelines and graphics in chronological order [7]. In evaluating data, the team searched for patterns in the data that predicted outcomes, identified problems, and tested solutions. These patterns provided answers to the propositions that explained the alternative recruitment processes with documentary evidence of when,where, and who was enrolled in the context of the COVID-19 pause.

During this RCT, the consent dates provided documentary evidence of the recruitment process challenges in a real-life context that was highly pertinent to the phenomenon. In team meetings we reviewed recruitment patterns that offered potential explanations of the effect of various approaches, answered initial propositions, and brainstormed new ideas. Data gathered systematically regarding different approaches provided a chain of evidence [7] to explain the usefulness of each approach facilitating analyzing the initial propositions. Furthermore, having multiple sources of evidence, such as, including the voices of clinic staff in focus groups, assisted in data triangulation to understand the chronological influence of the COVID-19 shut down on recruitment.

Lastly, rebuilding trust with the partnering clinics and providing them with continual feedback and updates during the height of COVID-19 were definite challenges. Numerous turnovers in providers, clinical staff members, and changes in rules/regulations at the partnering clinics coupled with limited in-person interaction made our study difficult to promote. For instance, most stage I cancer survivors were pivoted to telehealth visits or were referred back to their primary providers. Internally there were some changes in the research team resulting in a reduction in the number of personnel which slowed recruitment. The research team in our study consisted of former health care providers (i.e., nurse practitioners) translating to higher personnel costs. Focus group interviews with study collaborators clearly identified barriers of workload and cancer focus during clinic visits. They also suggested adding sleep assessment questions to preclinic visit assessments to sensitize patients to their sleep health behaviors and the role of good sleep in improving their health in the future.