In developed regions of the world, men are more likely to be diagnosed with prostate cancer than any other cancer, and those diagnosed are more likely to develop distress or serious psychological problems than healthy men [1,2]. Over 60% of men with prostate cancer report unmet psychological needs and up to a third experience pronounced clinical distress [3-7]. They also have a higher risk of suicide than their healthy male counterparts [8]. A range of factors contribute to men’s psychological comorbidity. Side effects of treatment such as urinary, sexual, bowel, and body-image problems can have a negative effect on cancer-related distress for as much as 2-3 years after diagnosis [9-11], and men’s psychological well-being can be adversely affected by lack of support, the threat of cancer, and the perceived loss of masculine identity [12-14].

The numbers of men with prostate cancer living with and beyond diagnosis are predicted to grow. There are over 1.1 million new prostate cancer cases globally per year, accounting for some 15% of all cancer diagnoses in men [2]. Incidence varies but trends indicate increasing diagnoses and decreasing mortality rates, particularly in developed countries and where screening programs have been implemented [15]. Five-year survival rates now exceed 84% in Western Europe and approach 100% in the United States and Australia [16-18], and in the United Kingdom for instance, incidence rates are expected to rise by 12%, to over 77,000 new cases per year by 2035 [19]. The growing number surviving prostate cancer means there will be more men experiencing reduced psychological well-being and quality of life, resulting in increased care utilization and health service costs. Innovative, accessible, and low-cost care delivery solutions are required to meet this long-term challenge.

Providing early psychological support is vital to ensure men experiencing distress after prostate cancer treatment do not fall into a cycle of negative thinking and avoidance behaviors, which can escalate symptoms and lead to the need for more intensive, prolonged support [5,6]. However, men’s engagement with psychological support is frequently restrained: reticence to communicate and delays in presenting to clinicians are underpinned by fears of stigmatization and the desire to normalize their illness experience by not needing help [20]. To support men’s psychological needs, it is essential to develop interventions that address these barriers.

The effectiveness of cognitive behavior therapy (CBT) is well documented; it can offer an acceptable, brief intervention within mental health services for people experiencing emotional difficulties as a consequence of comorbid problem(s) [21,22]. More recently, Web-based CBT has proved as effective as CBT delivered face-to-face [23]. However, there is mixed evidence for the role of clinician guidance in Web-based interventions. Although clinician support has been considered important for beneficial outcomes, there is evidence to suggest that the level of training for those providing guidance may be of limited importance; in some cases therapist effect may be minimal, and support can equally come from nonclinicians [23,24]. Conversely, a recent review concluded that there is limited evidence to show that self-guided interventions, in any delivery mode, can reduce psychological distress after cancer, but the authors did consider that efficacy may be increased if interventions are targeted at people formally assessed as being distressed [25]. Notably, recent meta-synthesis of qualitative studies in long-term conditions established that building Web-based social ties with peers can support self-management and improve illness experiences in aspects that are hard for individuals to negotiate offline [26].

Web-based CBT for cancer patients, and prostate cancer in particular, is a less-developed area compared with other chronic physical conditions [27-31]. For men with prostate cancer, Web-based delivery of psychological support is promising, it can facilitate access and engagement by providing a faceless, perceptually private environment to ameliorate men’s fears of stigmatization; it can also prove cost-effective for health services. A systematic review has shown that psychological interventions for prostate cancer survivors can improve mental health [32], but although 10 of the 21 effective patient-focused interventions identified were based on or contained components of CBT, only 2 were Web-based interventions. Both showed an improvement in depression [33] or distress [31], but neither of the interventions was carried out among a sample of men who had been formally assessed as being distressed nor were they delivered within a clinical setting. Outcomes from these studies are more relevant to worried prostate cancer patients than to a clinically distressed prostate cancer population requiring a therapeutic service.

In this study, we describe the development and feasibility of delivering a Web-based intervention in clinical practice for men with mild and moderate distress after treatment for prostate cancer. The program offers self-guided CBT augmented with filmed peer support and low-level chat room facilitation to encourage self-management; it aims to offer men the ability to monitor their condition and to affect the cognitive, behavioral, and emotional responses necessary to maintain an acceptable level of psychological well-being [32,34,35]. The program is intended to provide a cost-effective, brief intervention that can be offered through health services with minimal practitioner input. Reflecting recommended foci for feasibility studies [36], we assessed (1) demand through uptake and attrition, (2) acceptability by adherence and participant satisfaction, (3) potential for improvement in distress (and self-efficacy phase II) through limited efficacy testing, and (4) potential challenges for implementation in clinical practice.

We conducted 2 phases of feasibility research. Phase I assessed the program prototype in a low-intensity, primary care psychological service within which it was developed. Data from that phase informed further development, and phase II tested a slightly revised version in a secondary care cancer service.

The studies were approved by the UK NHS National Research Ethics Service, phase I reference 13/SC/0065; phase II reference 15/SC/0690.

In accordance with the Medical Research Council framework for developing and evaluating complex interventions [37], we used a pre-post, within-participant comparison, mixed-methods design in both phases. In phase II, sequential qualitative interviews were conducted after the final assessments to provide complementary context to the data [38]. In phase I, the intervention ran in February and again in March 2015, with separate facilitators and participant cohorts. In phase II, the intervention ran once with a single cohort and facilitator in June 2016.

Identification, eligibility, and screening are outlined in Figure 1. Men diagnosed with prostate cancer, not receiving palliative care for metastatic disease, were invited. In phase I, 432 men diagnosed up to 5 years were invited by a letter from their primary care physician; in phase II, 606 men diagnosed up to 3.5 years were invited by a letter from a nurse consultant in a secondary care cancer service. The letter contained full participant information and a link to the study website where all further contact took place. Interested men visited the website and gave informed consent. Consented men were then assessed for eligibility and those eligible were screened for distress. Men experiencing mild distress and men experiencing moderate distress were asked to complete the remainder of baseline assessments and were offered the intervention. (All inclusion and exclusion criteria are summarized in Table 1.) A risk-assessment protocol was administered throughout phases I and II (see Multimedia Appendix 1).

The program concept was developed from our previous research in urinary symptom self-management after prostate cancer treatment [39]. In response to a custom-made motivational peer support film used in the randomized controlled trial of that intervention, service users requested self-guided, easily accessible support to help manage their psychological distress. This need was confirmed in a scoping review of available psychological care, literature review of the status of cancer-related psychological interventions [40], and emerging policy initiatives to provide psychological support within primary care services for people struggling with a chronic condition.

The Web-based program, Getting Down to Coping, was produced in conjunction with a low-intensity psychological service that accepted general physician and self-referrals. The service was part of England’s National Health Service, Improving Access to Psychological Therapies (IAPT) program, which offered low- or high-intensity therapy, or referral to specialist mental health services [41]. Particular features included the use of standardized and manualized evidence-based CBT and routine outcome monitoring at each clinical session. The low-intensity service typically offered brief courses for people with mild and moderate anxiety or depression, mostly by telephone, but face-to-face and computer-based support was available. Published evaluations have reported recovery rates in excess of 50%, supporting service objectives, and IAPT is now developing services to provide tailored support for people with mental health needs associated with a long-term physical condition [42,43].

The research team, in collaboration with a software engineer, senior mental health practitioners, urologist, psychosexual therapist, specialist nurse, and 3 user representatives, codeveloped the initial program prototype based on the manualized short course of CBT delivered by the service. The course was then tailored to reflect prostate cancer-related examples and concerns and supplemented with links to medical, physical, emotional, social, and financial prostate cancer information. The program website was styled with graphics and language to appeal to a male audience and to reduce connotations of mental health.

A fundamental component of the program was peer support, but providing this over the internet is complex. Active engagement with others through posting messages can mediate positive outcomes, but not all men are prepared to do this, and evidence shows that passively viewing messages is not associated with the same beneficial outcomes [31,44]. To support men, we embedded theory-driven peer support films [45], as well as a platform for interactive support via an asynchronous chat forum. A single chat forum thread ran weekly; each week the facilitator started the thread by posting a question relevant to the module topic. The program was beta-tested by service users and the films were evaluated in focus groups. Comments from users and research participants were incorporated to refine the program and films.

The program contained 4 weekly, consecutive CBT modules with an introduction at the beginning of Module 1 (see Multimedia Appendix 2). It was intended that men should spend up to an hour per week on each module, including chat room activity. Modules were available 1 each week and progress was sequential. A male narrator supported the text. Men were invited to create a profile that other men could view. All worksheets and materials were available to download.

The IAPT service used the Patient Health Questionnaire-9 (PHQ-9) [46] and the General Anxiety Disorder-7 scale (GAD-7) for outcome monitoring [47]. To mirror the service’s practice, these were administered weekly within the program at the beginning of each module. Except in a situation of risk, it was intended that no formal feedback from these measures would be given to participants. In phase II, the measures were replaced with a noncompulsory weekly mood diary that invited participants to rate how they were feeling on 5 scales: down/cheerful, irritable/calm, vulnerable/in control, weary/active, and worried/relaxed; participants had the opportunity to review their previous scores. Phase II incorporated email notifications of chat room posts, chat room access from any page, easier navigation, and frequently asked questions for IT support. (screenshots in Figure 2.)

The prototype was developed with hard-coded software, which limited functionality. For phase II, the program was redeveloped within a content management system to enable integration into clinical practice and facilitate further research.

There were 3 chat room facilitators, each responsible for 1 cohort of men and all were trained by a subteam of researchers, clinicians, and user representatives. Training covered Web-based facilitation, prostate cancer issues, and/or CBT theory. In phase I, 2 low-intensity psychological practitioners from the collaborating service carried out facilitation. Each practitioner allotted 2, predesignated 2-h slots per week to facilitate the chat forum for their cohort. They also visited the website intermittently during office hours to assess risk. During these slots, the practitioners did not continue with their usual clinical caseload. In phase II, a specialist cancer nurse delivered facilitation and accessed and interacted with the program on an ad hoc basis during the time it was available to participants.

We assessed participant demographic, disease, treatment and satisfaction profiles, and uptake and adherence descriptively. We used descriptive statistics (mean, median, standard deviation [SD], interquartile range [IQR]) and box plots to examine the distribution of distress measured by the GHQ-28 and self-efficacy. The samples were not powered to detect significance in the outcome measures, nevertheless we present nonparametric data in relation to distress and self-efficacy to aid understanding of the potential effect of the program within these samples and provide data on which to base a power calculation for a larger study of efficacy. Statistical testing was performed on the 2 samples of core users, which for feasibility testing in this design provides a more useful measure of overall outcome [48].

A total of 14.1% (61/432) of invited men consented to the take part in phase I, and 12.2% (74/606) of invited men consented to take part in phase II (Figure 3).

In phase I, 47 eligible men were screened, of whom 32% (15/47) were experiencing distress: 21% (10/47) indicated mild and moderate distress and were offered the intervention and 11% (5/47) indicated suicidal thoughts and were referred for clinical assessment.

In phase II, 49 eligible men were screened, of whom 43% (21/49) were experiencing distress, 41% (20/49) indicated mild and moderate distress and were offered the intervention, and 1 indicated a suicide risk and was referred for clinical assessment.

In phase I, 10 participants were offered and started the intervention, and 2 withdrew in the first module (1 felt the program was not appropriate for his needs and 1 gave no reason). A total of 80% (8/10 across 2 cohorts) remained in the program for the 4 weeks and completed all assessments (Figure 3).

In phase II, 20 participants were offered the intervention, 1 did not start (no reason given), and 3 withdrew during the first module (1 declined because he had been recently bereaved and 2 gave no reason). A total of 80% (16/20 in 1 cohort) remained in the program for the 4 weeks and completed all assessments (Figure 3).

Baseline demographic, disease, and treatment characteristics for core users are reported in Table 3.

Questionnaire responses indicated that the Web-based recruitment and consent process was understood, appropriate in language and style, swift to respond, and easy to access (Table 4). Response to the program and the films was also positive, but there were some issues raised and clarified in open-ended and qualitative responses that will inform future development. Issues were related to the following: (1) length of sessions, the last module Getting There was shorter than the preceding modules, which was disappointing; (2) questions about suicide (at screening and assessment) were alarming for some; and (3) the need to enhance identification with the men in the films by providing details of what treatments they had received. In phase I, the chat room was poorly endorsed: it had been difficult to locate, there had been little activity and opportunity to chat, and the facilitation was not perceived as supportive. These issues were addressed in phase II and satisfaction improved: access and ease of use was enhanced, notifications of chat room activity were provided, and facilitator interaction was increased.

Phase II interviews helped clarify the questionnaire responses. (Participant profiles: Multimedia Appendix 3; verbatims: Multimedia Appendix 4). Acceptability of the program was high, even among those with low IT skills, and there was a readiness to improve skills. Men were comfortable using tablets, mobile phones, laptops, and desktop computers: they mainly found it effortless and flexible in comparison with other forms of support, even those who were finding concentration, or the availability of free time, a challenge. Some expectations were not met: disappointment with the length of modules noted in the assessment was attributed to less interaction time and things to do in the final module.

Men readily identified that the program targeted issues they found difficult to talk about. Learning about the link between the effects of treatment and mood and behavior was a fresh perspective, and they felt that the skills developed to manage now would be useful should things change in the future. Men referred more consistently, however, to other aspects of the program. The weekly provision of information via a range of links related to that week’s learning was used enthusiastically because it provided access to immediate information that meant men could control when and how much they consumed. They identified that having the links was an improvement on their usual Internet use because they were direct and avoided lengthy searching and inappropriate or potentially scary information. The films provided discreet stories, preventing the unpredictable, which helped men feel connected and in control, particularly those with social and information needs.

For those who engaged in the chat room, it was a safe environment where they could be honest with each other without the inhibitions they often experienced with clinicians. It was also a source of quick answers to spontaneous questions, which for some was very appealing; for others, it could be daunting if something was revealed that was incorrect or alarming. Despite the improved chat room satisfaction scores in phase II, some felt that the facilitator could have made more attempts to encourage men to open up and interact.

The samples were not powered to detect a significant change. Notwithstanding, we carried out nonparametric testing to determine potential for change in distress between baseline and post intervention; the samples performed similarly (Figure 4). A Wilcoxon signed-rank test indicated improvement in distress at the end of the intervention in both phases (phase I z=−2.213, P=.03, r=−.55; phase II z=−3.342, P=.001, r=−.59). In phase II, we also calculated change in domain scores. From baseline to post intervention, there was a positive change in somatic domain symptoms (z=−2.588, P=.01, r=−.458) and anxiety domain symptoms (z=−3.466, P=.001, r=−.613). Scores for social dysfunction domain symptoms were z=−1.531, P=.13, r=−.27 and for severe depression domain symptoms scores were z=−1.283, P=.20, r=−.23 (see Multimedia Appendix 5).

At baseline, men were most confident in performing daily activities (mean 69, median 80.0, SD 22.4, range 18-72), less-confident coping/tolerating symptoms (mean 52, median 53.5, SD 17.2, range 26-80), and least confident managing their symptoms (mean 31.6, median 31.50, SD 9.3, range 12-49; see Multimedia Appendix 7). A Wilcoxon signed-rank test performed on baseline and post scores showed an improvement in coping (z=−2.329, P=.02, r=−.412). Change was not indicated in managing symptoms (P=.11) or in performing daily activities (P=.08).

This is the first study, to our knowledge, to assess the feasibility of delivering Web-based CBT support in clinical settings among men screened with distress after treatment for prostate cancer. The Getting Down to Coping Program was embraced and acceptable to its target users. It can be delivered in a clinical service and has the potential to provide a therapeutic psychological service remotely.

There are limitations to these studies. The sample sizes were small and were not powered to detect change, and participants were not randomized. Generalization of our findings must therefore be cautious. Where we found change we do not know what variables are responsible; natural recovery could play a role and so could extraneous events. Nevertheless, with feasibility testing in 2 clinical settings, we have developed our knowledge of both the intervention and research required to move to the next trial stage. The consistency across our 2 samples on a number of measures, and the effects found, indicate that larger scale, evaluative research is justifiable. Our further research will include a longer follow-up period to provide an indication of maintenance of change, as well as full cost-effectiveness analysis. Furthermore, we will analyze covariance in respect of facilitator and group effects.

A strength of our studies was that we incorporated the theoretical model of PSD [52], which can provide an objective understanding of adherence. We were able to measure social and task design elements, which we posited were the most important features in our intervention for effecting behavior change. Measuring intended use is reflective of assessing compliance in face-to-face therapy and has been proposed as the most realistic reference standard for adherence in Web-based interventions [50,60]. Analysis on this basis offers more robust comparability within and across studies than static measures of exposure, such as log-in data, which are inherently subject to system and participant ambiguity. For instance, log-in time in phase I gave little indication of what interaction took place. In phase II, we measured time logged-in-and-active, which gave a clearer indication for adherence purposes, but we still had to make assumptions, that is, how long a log-in was deemed to contain active time. Such assumptions are often not reported in studies. More extensive application of the theoretical underpinnings of how design and system components can influence behavior change is called for [79].

To integrate psychological services for cancer patients in the existing care pathways, interventions that fit with health provider parameters of care provision are required [75]. Without research sampling based on defined need, or analysis of implementation processes, many intervention studies do not provide sufficient evidence of viability or efficacy [75]. We have taken the first steps to address this with a Web-based intervention, by assessing feasibility among a population that requires clinical support and by providing that support within a clinical practice context.

The program has the potential to fit within a stepped model of care by providing psychological support for men who are mild or moderately distressed and who fall within the clinical parameters for low-intensity support. Addressing these men’s needs will prevent escalation of symptoms and the need for higher-intensity therapy. Potentially, this would lead to service cost savings in terms of reduced physical and mental health service use.

In a stepped model of therapeutic care, the program requires low-level facilitation for monitoring risk, which raises cost implications versus a completely automated system. However, within the clinical services we researched, risk surveillance was a mandatory requirement, and the true cost comparison would be versus a therapist-led, face-to-face, or telephone approach. The numbers of men who can be supported with the Web-based program at any one time is subject to economies of scale, and cost advantage can increase exponentially with the volume of patients taking part. Furthermore, facilitation need not be carried out by advanced practitioners, which contributes to delivery cost advantage. Assigning staff to the facilitator role who possess appropriate competencies, as well as ensuring their involvement with and commitment to the innovation, will be crucial to its success [80].

Clinical effectiveness, and cost-effectiveness in terms of clinical delivery and health service utilization, need to be tested in an evaluative research design. Future research should be underpinned by exploratory enquiry to establish the most relevant and engaging ways to communicate study and intervention characteristics for this prostate cancer population. In future, we hope to develop the Web-based program for men with other cancers.