Timing and Tailoring of Internet-Based Cognitive-Behavioral Treatment for Insomnia for Cancer Survivors: A Qualitative Study

BACKGROUND

Roughly one in three of the 15.5 million cancer survivors in the U.S.¹ experiences clinically significant distress, depression, anxiety, fatigue, and/or insomnia.^2–4 Internet-delivered interventions hold significant promise to increase access to effective behavioral treatments for these symptoms, given that Internet interventions are highly cost-effective, quickly scalable to accommodate demand, and accessible with an Internet connection anywhere, anytime, and privately.⁵ Understanding how to best tailor and time the delivery of efficacious behavioral Internet interventions for cancer survivors is essential to best meet the significant psychosocial needs among this growing population. Appropriately timing the delivery of behavioral Internet interventions to survivors is important to maximize survivors’ uptake and sustain their use of these programs to ultimately ensure greatest impact.⁶ User feedback from survivors about their preferences regarding the level of tailoring will help developers of Internet-delivered behavioral interventions strike the right balance between specificity and generalizability. Therefore, to address questions regarding timing and tailoring of behavioral Internet-based interventions for cancer survivors, we examined qualitative feedback from focus groups of cancer survivors who piloted a standard version (i.e., not tailored for cancer survivorship) of an Internet-delivered CBT for insomnia intervention.

METHODS

Participants

Participants were cancer survivors from a pilot trial of SHUTi (Sleep Healthy Using the Internet) between July 2008 and July 2009. Pilot trial data have been published;⁷ eligibility criteria are listed in Supplemental Material. Of the 28 participants enrolled in the trial, 21 were contacted to participate in focus groups as they remained in remission from cancer and had completed SHUTi. Of those contacted, 11 participated in focus group interviews, while three did not respond and seven were unable to be scheduled for an interview.

Procedure

Study procedures were approved by the University of Virginia Institutional Review Board (IRB #13685). All participants provided informed consent. Two trained interviewers led the semi-structured focus group interviews (see Supplemental Materials for COREQ reporting guide).

Qualitative analysis

Focus group recordings were transcribed verbatim and were then free-coded using inductive thematic textual analysis⁸ by a trained research assistant. Next, KS completed a secondary coding of the discussion generated from two select interview questions pertinent to the research questions (i.e., timing: “At what point during treatment or recovery from cancer would this sleep program have been helpful to you?” and tailoring: “Is there any cancer-specific information that would have been helpful to have included in the program?”). All authors reviewed coding results and generated overarching themes.

RESULTS

Focus group and participant characteristics are listed in Tables 1 and 2, respectively. Thematic codes and supporting quotations are presented in Supplemental Materials. Quotations are denoted with a participant number (letter indicates the focus group).

Table 1.

Focus group characteristics

Group	N	Date (duration)	Modality
A	5	9/28/2009 (90 minutes)	In-person (video recorded)
B	2	11/3/2009 (60 minutes)	Conference call (audio recorded)
C	4	11/23/2009 (90 minutes)	In-person (video recorded)

Table 2.

Participant characteristics (N=11)

	n (unless specified)
Age	M=57 (range:48–68)
Education
Some college	2
College degree	3
Graduate degree	6
Marital status
Married/living with partner	9
Widowed	2
Comfort with the Internet
A little uncomfortable	1
Comfortable	1
Very comfortable	2
Extremely comfortable	7
Time since diagnosis (years)	M=6.5 (SD=6.5)
Stage at diagnosis
I	5
II	3
III	3
Treatments received (Yes)
Chemotherapy	9
Radiation	6
Surgery	11
Total nighttime sleep time (hours)
Pre-intervention	M=5.6 (SD=1.1)
Post-intervention	M=6.0 (SD=1.5)
Insomnia Severity Index
Pre-intervention	M=17 (SD=3.5)
Post-intervention	M=11 (SD=5.4)

Timing the intervention

A consistent message from the participants was that they would not have been interested in, or able to effectively enact, behavior change techniques recommended by SHUTi during chemotherapy (e.g., sleep restriction, stimulus control). Specifically, participants cited the burden of side effects, shock of the diagnosis, effort to learn as much as possible about their cancer and treatment, and “just concentrating on living” (C1) as factors that would have made it “too complicated, too much” (C4) to initiate and sustain behavior changes. Participants did appreciate low-effort strategies during this time that offered symptom relief, such as relaxation exercises. In two focus groups, participants specifically recommended offering the intervention following the transition off active cancer treatment, when participants frequently report an uptick in distress, often due to the reduced support and surveillance from their oncology team. Several participants discussed that the period transitioning off-treatment was when their insomnia symptoms worsened. There was general agreement among participants that the optimal timing is “when you recognize that you have a problem… and that you want help” (C4).

Tailoring the intervention

Opinions regarding the inclusion of cancer-specific tailoring were more varied. Several participants indicated that cancer-specific information about how cancer and its treatments relate to their insomnia symptoms would have been “really helpful” (B2), and the acknowledgment of her cancer experience in the intervention would have made at least one survivor feel “very welcomed” (C2). Other participants, however, indicated that cancer-specific information would have been distressing, unwanted, or redundant. Several participants described how they had already received information about the effects of their cancer and related treatments on sleep from their doctors or had accessed that information on their own through trusted educational resources. Participants did generally agree that, if cancer-specific information were offered, that it would be best as optional content, making it “possible to bypass if it doesn’t apply” (B2) to the particular user.

DISCUSSION

This study presents qualitative feedback provided by cancer survivors who completed SHUTi, an Internet-delivered behavioral intervention for insomnia. Intervention components should be timed appropriately in survivors’ treatment trajectories to ensure greatest survivor engagement and ultimately improved outcomes, while the utility and impact of tailoring content to be cancer-specific remains uncertain.

Clinical implications

Focus group participants indicated that educational and low-burden symptom relief measures would be the most acceptable intervention format during active cancer treatment, while treatment components that require more concentrated behavior change and patient commitment may be better targeted to the transitional period off-treatment. Participants did note, however, that awareness of a problem and desire for help are essential to motivating treatment seeking and engagement, so future research is needed to explore just-in-time patient-centered models of care delivery that increase awareness about symptom control issues and treatments and link individual patients to the right care at the right time for them. Participants expressed more diverse opinions on the utility and desirability of cancer-specific tailoring. Contrary to the tacit assumption that interventions should be tailored to different user attributes (e.g., race, gender, chronic disease diagnosis), these findings suggest that behavioral interventions with extensive cancer-related information may not resonate with all survivors. Given that there is a trade-off between tailoring and generalizability, further research is recommended to test the effects of tailoring on program uptake, engagement, and efficacy – and whether these effects hold across interventions targeting different clinical issues.

Study limitations

The primary study limitation is that data were collected in 2009. Both cancer treatment and Internet interventions have changed dramatically in the past decade; however, significant implementation issues remain for delivering psychosocial care to survivors: 60 to 70 percent of distressed cancer survivors still decline psychosocial services⁹ and up to 50 percent of Internet intervention users still drop out of interventions prematurely.¹⁰ The survivors’ perspectives described herein remain pertinent to addressing implementation issues of Internet interventions in psychooncology. In addition, findings are based on a convenience sample of individuals participating in a pilot trial – sample recruitment was therefore not guided by saturation. Interviewed participants were also exclusively breast cancer survivors and tended to be highly educated. As such, perspectives offered by our participants may not fully represent the considerations of all survivors, or even all survivors who used the intervention. Users were also only exposed to one specific Internet intervention, specifically, a CBT for insomnia intervention. It is possible that interventions that require fewer or easier lifestyle changes may be more willingly received by survivors; yet given homework and behavioral changes are key components of CBT interventions broadly, survivors’ reticence to take on additional responsibilities during chemotherapy likely still apply.

Conclusions

Internet-based interventions hold significant promise to help address supportive care needs among the growing population of cancer survivors in a cost-efficient, scalable, and accessible way – if these programs are developed in a way that is appropriately fitting and feasible for survivors. Actively incorporating survivors’ perspectives into the development of these programs will be essential to realize their potential impact.

KEY POINTS.

• Internet interventions can help increase access to supportive care among the growing population of cancer survivors.

• Survivors who used an Internet-delivered CBT for insomnia intervention reported that timing the intervention to begin during transition off-treatment may be optimal.

• Feedback was mixed as to whether cancer-relate tailoring would be helpful or necessary.

• The utility and impact of tailoring content to be cancer-specific remains uncertain.

• Actively incorporating survivors’ perspectives into the development of behavioral Internet interventions is essential to realize the potential impact of these programs.