Systematic review of technology-based interventions to improve anxiety, depression, and health-related quality of life among patients with prostate cancer

Abstract

Objectives:

This review aimed to systematically evaluate the characteristics of technology-based interventions (TBIs) and their effectiveness on anxiety, depression, and health-related quality of life (HRQoL) among patients with prostate cancer.

Methods:

We identified eligible research reports published in English language between January 1, 2000, and September 15, 2018, from CINAHL; Embase; “Library and Information Science Abstracts”; “Library, Information Science and Technology Abstracts”; “Library and Information Science Source”; PsychINFO; and PubMed. We abstracted randomized control trials and quasi-experimental studies that measured the outcomes related to anxiety, depression, or HRQoL. We extracted the data using the Cochrane Handbook for Systematic Reviews of Interventions guideline.

Results:

Among the six randomized control trials and four quasi-experimental studies that met our inclusion criteria, the TBIs aimed to provide informational, psychosocial, self-care management, and communication support. About 60-92% of the participants in six studies completed all required contents and 77-94% of the participants in four studies logged onto the TBI platform. Compared with the patients in usual care, the TBI users reported a significant reduction in anxiety (N=1 study) and depression (N=2 studies) and improvement in HRQoL (N=2 studies). We also identified the limitations of the existing TBI trials.

Conclusions:

We found insufficient evidence to support the effectiveness of TBIs in improving health outcomes (anxiety, depression, and HRQOL) among patients with prostate cancer. Future research needs to (1) use rigorous randomized control trials, (2) be sufficiently powered to examine the effects of TBIs, and (3) examine how the effect of TBIs on health outcomes vary by actual intervention use, intervention components, and duration.

1 |. BACKGROUND

Approximately 18 to 27% of patients with prostate cancer experience depression, anxiety, and reduced health-related quality of life (HRQoL).^1,2 Health care providers often fail to meet the psychosocial care needs for prostate cancer patients³ due to health care costs and insufficient health care resources.^4–6 National guidelines on cancer survivorship from the American Cancer Society has called for programs to better address the unmet care needs for prostate cancer patients and their families. Although in-person interventions such as nurse-led group counseling have been effective in reducing depressive symptoms among prostate cancer patients, they are expensive and inconvenient because of the time and travel required.⁷

Psychosocial and behavioral care delivered using technology-based interventions (TBIs) has the potential to provide cost-effective psychosocial care services to a large number of cancer patients and their families, to facilitate engagement in survivorship care, and to reduce the fear of stigma related to attending mental health services.^8,9 Frequently used TBIs include web-based programs, mobile apps, telehealth communication, and monitoring sensors to enhance individualized health care.¹⁰ Although TBIs can be scalable to improve the psychological well-being and HRQoL among prostate cancer patients and meet their survivorship care needs,^11–13 little is known about the characteristics and effectiveness of TBIs on anxiety, depression, and HRQoL among patients with prostate cancer.⁸ The purpose of this systematic review is to evaluate the characteristics of TBIs and their effectiveness on anxiety, depression, and HRQoL among patients with prostate cancer. The findings from this review will provide evidence integral to designing effective TBIs enhancing supportive care for prostate cancer patients.

2 |. METHODS

2.1 |. Identification of studies

We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to guide the systematic review process (Figure 1). A research librarian developed the search strategy and performed the search to identify articles reporting the effectiveness of TBIs on anxiety, depression, and HRQoL among prostate cancer patients. For the purposes of this review, we defined TBIs as health management or behavioral change interventions implemented via digital platforms, such as mobile apps, web-based programs, multimedia applications, or delivered through electronic devices, including smartphones, tablets, and computers.¹⁴

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart of study identification and selection

We identified eligible research articles from the following electronic databases: CINAHL; Embase; “Library and Information Science Abstracts”; “Library, Information Science and Technology Abstracts”; “Library and Information Science Source”; PsychINFO; and PubMed. We adjusted these keywords with vocabulary and syntax across databases: (eHealth OR “digital health” OR mHealth OR multimedia OR web* OR app OR apps OR “mobile technology” OR internet OR online OR “mobile phone” OR smartphone OR tablet OR computer* OR application* OR “mobile applications”) AND English[lang] AND (“prostate neoplasm*” OR “prostate cancer” OR “prostate cancers” OR “prostate tumour” OR “prostate tumours” OR “prostate tumor*”) AND (depress* OR anx* OR “quality of life” OR “health related quality of life” OR QOL OR HRQOL).

To meet inclusion criteria, the studies (1) targeted patients with prostate cancer at different stages; (2) included TBIs for health management or behavioral change; (3) measured health outcomes of anxiety, depression, or HRQoL; (4) used a randomized control trial (RCT) or quasi-experimental research design; and (5) were published between January 1, 2000, and September 15, 2018 (to maximize the number of studies). Quasi-experimental studies were included in this review because of the lack of published RCTs in this area. Studies were also excluded if they (1) included only telemonitoring (eg, telephone calls, video calls, or other approaches for self-monitoring) of chronic illness and/or symptoms, (2) were published in a language other than English, and (3) only included abstracts but full articles were unavailable. Two investigators independently screened and evaluated the abstracts to determine if the studies met inclusion criteria, and then reviewed and assessed the full reports of those meeting inclusion criteria. Disagreements were resolved among members of the research team.

2.2 |. Data extraction

We used the Cochrane Handbook for Systematic Reviews of Interventions as a guide to extract from the eligible studies.¹⁵ The extracted data included study design, aims, participant demographics, sample size, retention rate, TBI characteristics, TBI use, and health outcomes. If research projects assessed multiple outcomes, our reviewers extracted the outcomes related to anxiety, depression, and HRQoL. Thereafter, we developed a preliminary synthesis using the textual description as recommended in “The Guidance on the Conduct of Narrative Synthesis in Systematic Reviews”¹⁶ (see Table S1 in the Supporting Information). The textual description is a preliminary starting point in the synthesis to describe the studies in more depth than can be given in tables.¹⁶

The Cochrane risk of bias criteria were used to evaluate whether the risk of bias was low, unclear, or high risk.¹⁷ We rated studies as having a low risk of bias if all eight Cochrane criteria were met (see Table S2). Studies not meeting any of the Cochrane criteria were rated as high risk of bias. If a study omitted mention of Cochrane criteria, the study was rated as having an unclear risk of bias (Higgins et al, 2011).

3 |. RESULTS

3.1 |. Study characteristics

The initial search yielded 1,512 publications (Figure 1). After removal of duplicate papers and ineligible studies, 10 studies met our inclusion criteria, and thus, were included in this review. These studies included randomized control trials (N=6),^12,18–22 and quasi-experimental studies (N=4: three “2-group pretest-posttest” studies and one “1-group pretest-posttest” study).^23–26 These studies were conducted in Australia, Canada, Germany, Netherlands, Norway, Sweden, the United Kingdom, and the United States (Table 1).

TABLE 1.

Characteristics of included studies and technology-based interventions (TBIs)

First Author, Year, Design, and Country	Study Aims	Sample Size (n); Setting and Severity of illness	Method of Recruitment	Control	TBI Name	TBI Modality and Platform	TBI Content	TBI Duration and Sessions Time	Interventionist
Cockle-Hearne et al 2018; One group pretest-posttest; UK	Assess feasibility of online CBT among men experiencing mild to moderate distress after prostate cancer treatment.	(n=16) Outpatient diagnosed with localized prostate cancer within the last 3.5 years	Physician and nurse consultant invitation	N/A	Getting Down to Coping	Online CBT support website	CBT supplemented by peers’ support and chat room to encourage self-management. Self-monitor, improve cognitive, behavioral, and emotional responses necessary to maintain psychological wellbeing	4 weekly sessions lasting one hour per session	Psychological practitioner, and oncology nurse
Hawkins et al 2017; RCT; USA	Examine computer-based support to improve HRQoL	(n=310) Outpatient newly diagnosed with prostate cancer (Stage I or II)	Hospital and cancer centers	Cancer mentor	CHESS	Interactive online support website	Address patient needs for information, communication and support, build skills, decision map tool, and a module on managing sexual problems	6 months (as needed)	N/A (self-guided)
Lange et al 2017; Two group pretest-posttest; Germany	Examine the effectiveness of a guided chat groups in psychosocial aftercare	(n=143) Outpatient postprostatectomy; prostate cancer (Stage I to IV)	Prostate cancer clinic	Usual care	Web-based chat	Online chat group support website	Exchange concerns, problems and support with peers about incontinence, fear of progression, sexuality, communication, occupational reintegration, resource orientation, and coping	1 to 2 months (Weekly group session lasting 60 to 90 min)	Psycho-therapist
Loiselle et al 2010; Two group pretest-posttest; Canada	Examine the impact of a multimedia informational system on health-related outcomes	(n=45) Outpatient newly diagnosed with breast or prostate cancer; (Stage I or II)	Oncology clinics	Usual care	Oncology Interactive Educational Series	Cancer related multimedia information (websites and CD-ROM)	Information about cancer prevention, detection, symptoms, diagnosis, treatment, nutrition, pain, psychosocial care, community support, and answers to frequently asked questions	8 weeks (as needed)	N/A (self-guided)
Osei et al 2013; RCT; USA	Examine the effect of an online support system on HRQoL	(n=40) Outpatient diagnosed with prostate cancer within the last 5 years	Mailing	Prostate Cancer Resource kit	Us TOO	International online support group website	Offers information about prostate cancer, symptoms and side effects management, informed decisions, active surveillance, and treatment options	6 weeks (at least three times per week)	N/A (self-guided)
Ruland et al 2013; RCT; Norway	Examine the effects of online support system on symptom distress, depression, self-efficacy, HRQoL, and social support	(n=136) Outpatient undergoing treatment of breast or prostate cancer (Stage I to IV)	Advertising and mailing	Public cancer website	WebChoice	Online interactive health communication website	Enhanced symptoms monitoring, information and self-management, e-communication with health professionals, and e-forum group with peers.	1 year (as needed)	N/A (self-guided)
Sundberg et al 2017; Two group pretest-posttest; Sweden	Examine the effect of a smartphone application on symptom burden and HRQoL	(n=130) Outpatient undergoing radiotherapy for localized prostate cancer	Hospitals	Usual Care	Interaktor	Interactive smartphone application for detecting, reporting and managing of symptoms	Selfcare advice, symptom history graphs, symptom assessment for bladder, bowel function, fatigue, pain, anxiety, distress, sleep, and flushing, and risk assessment model based on symptoms with alert notification.	5 to 11 weeks (as needed)	Clinical nurses
Van De Wal et al 2017; RCT; Netherland	Examine the efficacy of blended CBT for high fear of cancer recurrence	(n=88) Outpatient survivors 6 months to 5 years posttreatment of breast, prostate, or colorectal cancer	Hospitals and mailing	Usual Care	bCBT	Blended CBT (website, face to face, and telephone call)	Psychoeducation, cognitive restructuring, and behavioral modification modules delivered as face-to-face and e-consultations or may be replaced with telephone consultations along with a workbook.	3 months (8 sessions total: five 1-h face-to-face sessions and three 15 minutes online consultation)	Oncology psychologist
Wootten et al 2015; RCT; Australia	Examine the efficacy of an online intervention to reduce participants’ psychological distress	(n=142) Outpatient undergoing treatment or posttreatment of localized prostate cancer within 5 years	Advertising, mailing, and urologist invitation	Moderated forum	My Road Ahead	Self-guided online psychological CBT website	Psycho-education, interactive exercises, automated feedback. The topics about the emotional impact of illness, cognitive strategies, communication, coping, and planning.	10 weeks (as needed; recommended weekly)	N/A (self-guided)
Yanez et al 2015; RCT; USA	Examine the feasibility, acceptability, and efficacy of a technology assisted psychosocial program	(n=74) Outpatient with advanced prostate cancer (Stage III or IV)	Hospitals	Health promotion	CBSM	Online cognitive behavioral stress management website	Stress management, education about (impotence, incontinence, intimacy, acceptance, existential concerns, and life narratives), stress reduction/relaxation technique, coping skills, interpersonal skills, and social networks.	10 Weekly group sessions: 30 min of relaxation and 60 min of stress management.	Masters level therapist

Abbreviations: bCBT: blended cognitive behavioral therapy; CBSM: cognitive-behavioral stress management; CBT: cognitive behavioral therapy; CHESS: Comprehensive Health Enhancement Support System; HRQoL: health-related quality of life; N/A: not applicable; RCT: randomized controlled trial.

3.2 |. Characteristics of research participants

The study sample sizes ranged from 16 to 310 participants (Table 2) recruited from outpatient settings in hospitals, cancer centers, and prostate cancer clinics through advertisements, mailings, or clinical provider’s invitation. Seven studies targeted only prostate cancer patients,^{12,18,19,22,23,25,26} whereas three studies targeted breast and colorectal cancer patients in addition to prostate cancer patients.^20,21,24 The cancer patients were in remission after being diagnosed with localized cancer or at varying stages of cancer (Table 1). The average age of participants across studies ranged from 57 to 69 years. Study participants were mostly White (57-100%), married (65-87%), and retired (30-77%). Participants reported educational levels ranging from elementary to graduate school, and most had post-secondary school degrees (73-89%). Two studies reported participants’ locations of residence,^12,23 and 73% and 89% of these participants were urban residents, respectively.

TABLE 2.

Effect of technology-based interventions on anxiety, depression, and health-related quality of life

Study	Assessment Time Point	Outcomes	Study Measures	Between Group Difference*	Within Group Difference (Intervention)*	TBI Use	Retention Rate (By End of Follow-up)
Cockle-Hearne et al 2018	Baseline, and end of intervention	Anxiety Depression	General Health Questionnaire-28	N/A	Positive change in anxiety between baseline and post-intervention (A Wilcoxon signed-rank test: (z= −3.466, r= −.613, p=.001) NS	92%a	80%
Hawkins et al	Baseline, 2, 6, 12 and 24 weeks	HRQoL	WHOQOL	NS	NS	78%	94%
Lange et al	Baseline and follow up	Anxiety Depression HRQoL	ET-5 and MAX-PC ET-5 SF-8	NS NS NS	N/A N/A N/A	N/A	31%
Loiselle et al	Baseline, end of intervention, and 3-month postintervention	Anxiety Depression HRQoL	STAI CESD SF-36	NS NS NS	NS NS NS	94%	93%
Osei et al	Baseline, 6 and 8 weeks	HRQoL	EPIC-26	NS	Improved HRQoL (p =.036) from baseline to 6 weeks	N/A	N/A
Ruland et al	Baseline, 3, 6, 9, and 12 months	Depression HRQoL	Epidemiological Studies Depression Scale 15-D HRQoL	NS NS	Improved depression over study period (mixed effect model: slope estimate= −0.41, CI= −0.71 to −0.11, t= −2.71; p= .007) NS	77%	75%
Sundberg et al	Baseline (T1), end of treatment (T2), and 3-month posttreatment (T3)	HRQoL	EORTC QLQ-C30	Interaktor users reported improved HRQoL at T2 and T3(p <.05)	Improved HRQoL (p<.05) over study period (baseline to T3)	N/A	88%
Van De Wal et al	Baseline and 3 months	Anxiety Depression HRQoL	HADS EORTC QLQC30	bCBT users reported decreased anxiety (ANCOVA: CI= −4.131 to −2.763, ES “Cohen’s d”= .67, p≤.001) bCBT users reported decreased depression (ANCOVA: CI= −2.950 to −0.781, ES Cohen’s d=.56, P≤.001) bCBT users reported improved HRQoL (ANCOVA: CI= 5.441 to 19.325, ES Cohen’s d=.64, p=.001)	N/A N/A	66%a	93%
Wootten et al	Baseline, 5 weeks, 10 weeks, 3 months and 6 months	Anxiety Depression	DASS-21	NS NS	NS NS	60%	73%
Yanez et al	Baseline and 6 months	Depression HRQoL	PROMIS depression item bank CAT FACT-G	CBSM users reported improved PROMIS depression (ANCOVA: F= 5.11, ES Cohen’s d= 0.69, p=.03) NS	N/A N/A	70 %a	82%

Abbreviations: 15-D HRQoL, 15 dimensions HRQoL instrument; ANCOVA: analysis of covariance CESD, Center for Epidemiological Studies Depression; DASS-21, 21 items depression, anxiety, and stress scale; EPIC-26, 26 items Expanded Prostate Cancer Index Composite; EORTC QLQ-C30, European Organisation for Research and Treatment of Cancer Quality of life Questionnaire Core 30; ET-5, The Emotion Thermometers; FACT-G, The Functional Assessment of Cancer Therapy–General; HADS, Hospital Anxiety and Depression Scale; HRQoL, health-related quality of life; MAX-PC, Memorial Anxiety Scale for Prostate Cancer; N/A: no available data; NS: no significant difference; PROMIS, Patient-Reported Outcomes Measurement Information System; SF-36, 36-Item Short Form Health Survey; SF-8, Short Form survey 8 items of health outcomes; STAI, State-Trait Anxiety Inventory; TBI: technology-based intervention; WHOQOL, World health Organization Quality of Life; depression item bank CAT, Adult Depression computerized adaptive testing; T1, 2, 3= time 1, 2, 3.

^aCompleted TBI contents.

^bLogged onto TBI.

^*Statistical significance at p ≤.5.

3.3 |. Characteristics of the TBIs

The intervention characteristics are displayed in Table 1. The TBIs aimed to provide informational, psychosocial, self-care management, or communication support. These studies used self-guided websites (N=5) ^{12,18–20,24}; web-based programs supported by medical professionals (N=4)^21–23,26; and a mobile health app with clinical nurse supervision (N=1).²⁵

The intervention duration of eight studies ranged from 4 weeks to 3 months,^{12,19,21–26} and two studies continued the interventions for 6-12 months.^18,20 Six studies specified that participants needed to complete 4 to 18 TBI sessions,^{12,19,21–23,26} whereas participants in four studies used the TBI programs as needed.^18,20,24,25 Four studies reported that the length of each session ranged from 15 to 90 minutes.^21–23,26 Seven studies provided technical assistance, IT training, or TBI user manuals to the participants.^{18,20,22–26}

We defined platform use as how often the study participants logged onto the TBIs or completed all TBI content.²⁷ Seven studies (Table 2) reported varied platform use, ranging from 60-92% of the participants completing all required TBI contents (N=4 studies),^12,21,22,26 to between 77 and 94% of the participants logging onto TBI sessions (N=3 studies).^18,20,24

3.4 |. Contents of the TBIs

The contents of the TBIs in the studies under review are outlined in Table 1. However, these studies have not determined which components of the TBIs are more or less effective and whether a combination of different TBI components or a particular component are the most effective. The TBI components can be categorized as (a) informational, (b) psychosocial, (c) self-care management, and (d) communication support. Informational components included patient needs in cancer,¹⁸ information about cancer illness,^{12,19,20,22,24,25} cancer prevention,²⁴ cancer management,^19,23–25 and decision making skills.^18,19 Psychosocial components involved psychological wellbeing and social integration contents including psychoeducation,^12,21 Cognitive Behavioral Therapy support,^12,21,22,26 stress management skills,²² behavioral modification strategies,^21,26 coping skills,^12,22,23 peer support,^20,23,26 community support,²⁴ social networks enhancement,²² occupational reintegration,²³ and interpersonal skills.²² Self-care management components included self-monitoring skills,^20,24–26 symptoms history graph,²⁵ symptoms management skills,^20,24,26 and self-care advice.²⁵ Communication support components included e-forum group discussion with other patients,^20,23,26 e-communication with health professional,^18,20,21,23 risk assessment alert notification,²⁵ and interactive exercises with automated feedback.¹²

3.5 |. The risk of bias

The studies under review demonstrated various levels of risk of bias in different domains as measured using the criteria established in the Cochrane risk of bias assessment tool (Table 3). All RCTs maintained a random sequence generation of group assignments using computer-generated lists; however, none included information about the process of concealment of allocation of group assignments and blindness. No study reported strategies to maintain intervention fidelity (eg, consistent intervention use among participants). Three studies used an intention-to-treat approach to examine the intervention effects.^12,20,21 Two RCTs reported unbalanced baseline measures among the study populations^19,23; these studies also lacked population diversity in terms of demographical characteristics among the participants, which limited the generalizability of the study results.

TABLE 3.

Quality of study assessment tool

Study	Random Sequence Generation	Allocation Concealment	Blinding of Participants and Personnel	Blinding of Outcome Assessment	Incomplete Outcome Data	Selective Reporting	Deferential Intervention use	Baseline Imbalance	Level of Risk
Cockle-Hearne et al	H	H	H	H	L	L	L	H	H
Hawkins et al	L	U	U	U	L	L	H	L	H
Lange et al	H	H	H	U	L	L	U	H	H
Loiselle et al	H	H	H	U	L	L	H	L	H
Osei et al	L	U	U	U	L	L	U	H	H
Ruland et al	L	H	H	U	L	L	H	U	H
Sundberg et al	H	H	H	H	L	U	U	U	H
Van De Wal et al	L	H	U	U	L	L	H	L	H
Wootten et al	L	U	U	U	L	H	H	L	H
Yanez et al	L	U	U	U	L	L	H	L	H

Abbreviations: L: low risk; H: high risk; U: unclear.

Two RCT studies^18,21 and one quasi-experimental study²⁵ reported having sufficient power (>0.80) to detect medium intervention effect size, whereas the remaining studies reported that the estimated sample sizes were underpowered. The participant retention rates between baseline and at the end of follow-up assessments ranged from 73 to 94% in eight studies.^{12,18,20–22,24–26} One quasi-experimental study reported a retention rate of 31%,²³ and one RCT included all of the randomized participants in the analysis and did not report information about participants dropout between the baseline and follow up points.¹⁹

3.6 |. Effect of TBIs on anxiety, depression, and HRQoL

Five studies examined the effects of TBIs on anxiety.^{12,21,23,24,26} Among these studies, one study reported a significant reduction in anxiety among TBI users (N = 45) compared with patients receiving usual care (N = 43),²¹ and one study reported a within-group reduction in anxiety in the TBI group only²⁶ (see Table 2).

Seven studies reported the TBI effects on depression.^{12,20–24,26} Among these studies, two reported a significant improvement in depression in the TBI users compared with the participants in control conditions.^21,22 Ruland and colleagues also showed improvement in depression in the TBI group during the study period.²⁰

Eight studies reported on the effects of TBIs on HRQoL,^18–25 with three studies indicating significant improvement in HRQoL among TBI users.^19,21,25 Sundberg and colleagues showed within and between groups improvement in HRQoL among participants using smartphone applications for symptom management.²⁵ Van De Wal and colleagues reported significant improvement in HRQoL among online cognitive behavioral therapy users, but not those in face-to-face psychological support groups.²¹ In their quasi-experimental study, Osei and colleagues reported improvement in HRQoL within online informational and educational support group subjects over time.¹⁹

4 |. DISCUSSION

We systematically reviewed TBI characteristics and effectiveness of TBIs on anxiety, depression, and HRQoL among patients with prostate cancer. We found limited evidence of TBI effectiveness in reducing anxiety and depression and improving HRQoL, which may be explained by the modest number of studies available for the review, the varied and underpowered sample sizes, and the lack of randomization, allocation concealment, and blinding in a number of studies. The significant heterogeneity in the study design, intervention use, outcome measurements, timing of follow-up, and the TBI modalities, components, duration and frequency of sessions, and availability of professional facilitation and technical support, has made the TBI effectiveness comparison across studies challenging. Furthermore, the homogeneity in study participants’ demographic variables (eg, older age, White, and married), although reduced the baseline imbalance in some studies, has limited the generalizability of the study findings among cancer patients with different backgrounds.

Our review, different from previous reviews of TBIs in other cancer populations, has shown that TBIs have the potential to reduce anxiety and depressive symptoms and improve HRQoL among prostate cancer patients.^{19–22,25,26} Web-based survivorship interventions have been shown to reduce anxiety and improving HRQoL among breast cancer patients.²⁸ Fridriksdottir et al reported in their review that web-based interventions positively affected anxiety, depression, and HRQoL for cancer patients who mostly were breast or female cancer survivors.⁹ Previous studies have indicated that website attributes, Internet usage patterns, and habits are different across genders.^29–31 The differences in the findings of our review and those from the other reviews suggest that the effects of TBIs may also vary by gender. Future research needs to explore the potential gender difference in TBI effects and the causes of these gender differences.

The studies in this review varied significantly in the TBI characteristics. For example, the TBI components included information and psychosocial support for symptom management; the TBIs also varied in whether they were self-guided by participants or involved professionals’ guidance and control. There was significant variation in TBI dosages in terms of total program duration, length of each session, and frequency of the TBI implementation. Nonetheless, we found limited evidence about how these TBI characteristics (eg, components, formats, and TBI dosage) influenced TBI effects on anxiety, depression, and HRQoL because of the limited number of studies available for this review, and insufficient data about how the TBI components and format contributed to the health outcomes. However, in their review, Talboom-Kamp and colleagues (2018) suggested that a combination of TBI components would show more promising effects than single component TBIs. Furthermore, although no studies in our review have reported whether and how the TBI duration and frequency changed the outcomes, previous reviews did not find evidence that TBIs lasting longer than 3 months, compared with those lasting less than 3 months, achieved a greater reduction in anxiety and depression or improvement in HRQoL among breast cancer patients ^28,32. Likewise, Lee and colleagues found in another systematic review that positive health outcomes were not directly related to the length of the TBIs.³³ Future research needs to examine rigorously whether and how the effects of TBIs on health outcomes vary by the TBI characteristics.

Neither did we find evidence about participants’ health outcomes being related to their use of TBIs (eg, completed all TBI sessions, discontinued use, or completed sessions over time). Our findings are consistent with the results from a previous systematic review of TBIs that showed that little studies monitored intervention dose (number of log on and website exploration) and its influence on HRQoL among breast and prostate cancer patients.⁸ Maintaining intervention use over time during study is a major moderator of TBI effectiveness³⁴ and is likely to influence participants’ long-term health outcomes.^28,35,36 In this review, participant retention rates between baseline and the end of follow-up were considered high, ranging from 73 to 94%.³⁷ The retention rate, however, reflects the percentage of participants retained for postintervention assessments, but not necessarily the actual intervention uses among participants. For example, Van De Wal and colleagues reported that 93% of the participants completed the follow-up assessments, indicating a high retention rate; yet only 66% of the participants completed all the required TBI sessions. In the meanwhile, although seven studies reported the actual intervention use over time, none reported how TBI use affected the study results.

Amid growing concerns regarding barriers to continuous TBI use among patients with chronic illness patients and how to overcome these barriers,³⁶ investigators have commented on the importance of developing assessment tools to explore facilitators and barriers to maintaining TBIs engagement.^38,39 Furthermore, the number of times a participant logs into the website may not reflect the time of TBI use (eg, completing sessions).⁴⁰ Future research, thus, should not only consider the benefits of TBI on participants’ health but also closely monitor, analyze, and report TBI use by participants.

Next, our findings have shown that half of the TBIs offered professional support to users, yet they did not achieve better outcomes than the self-guided interventions. Ugalde and colleagues (2018) recommended that further rigorous evidence is needed to understand the effect of self-guided interventions compared with the role of professional support or interventionists. Professional support in TBIs could provide information, guidance, and assistance for users regarding their symptoms management.⁴¹ The lack of guidance and assistance by health professionals is a substantial barrier to the use of TBIs among elderly people (eg, prostate cancer patients),³⁵ while health care providers’ guidance and assistance have resulted in more enhanced self-management in patients with chronic conditions than TBI access alone.³³ Similarly, Seiler and colleagues found in a systematic review that therapist-guided TBIs were more efficacious than self-guided interventions.⁴² In addition to health professionals’ support, we found that more than half of the studies provided participants with technical support such as user manuals, technical assistance, and technology training, which can reduce technical problems related to TBI use and enhance intervention fidelity in studies that require regular monitoring and attention.³⁶

Finally, all TBIs in this review were web-based, except the smartphone app in a quasi-experimental study (Sundberg et al, 2017), a finding similar to a previous report that TBIs were exclusively delivered via online web-based programs for health promotion among elderly people.³⁵ Smartphone apps, on the other hand, have been the most common type of TBI for health behavioral change among people with different health problems.^43,44 More user-friendly and straightforward web-based platforms are needed to deliver complex and advanced content consistent with illness severity and needs for patients with chronic illness (eg, prostate cancer patients) who are elderly, have cognitive deficits and comorbid conditions, and lack of technology skills.^43,44 These physical, individual, and social challenges coupled with the technical challenges in using mobile apps (small screen and font size) may mean that web-based interactive and comprehensive health interventions are more preferable.^42,45–47 However, new smartphone technologies have advanced features such as interactive mode, augmented reality, flexible screen, and voice recognition that might improve the usability of smartphone apps for complex health issues for older patients.^47,48

4.1 |. Limitations

Our selection criteria intentionally included TBIs that focused specifically on prostate cancer patients because of the unique challenges that prostate cancer survivor face. However, our review included three studies that also targeted breast and colorectal cancer patients, in addition to prostate cancer patients.^20,21,24 Among these studies, Loiselle and colleagues conducted subgroup analyses for breast and prostate cancer participants, whereas the other two studies combined all cancer conditions in the same analysis. In addition to HRQoL, we only assessed the effectiveness of TBIs on anxiety and depression since they are the major psychological symptoms among cancer patients, and they are with a clear definition, measurements, and diagnostic criteria.⁴⁹ Future studies need to assess the effects of TBIs on other psychological outcomes, such as psychological distress and cancer-related stress.

4.2 |. Implications in research and practice

The high risk of bias and underpowered sample sizes suggest that future research will require robust designs to examine the effectiveness of TBIs on health-related outcomes among prostate cancer patients. Randomization, blinding, consistent intervention use, and sufficient sample size should be considered in future research to minimize the risk of bias and enhance the study quality. Future studies must also include analysis of the impact of TBI characteristics such as components, format, duration and frequency of TBI sessions and the role of interventionists on the health outcomes. The variability of outcome measures and the timing of follow-up assessments have made it difficult and sometimes impossible to compare across studies and draw definitive conclusions. Therefore, researchers need to adopt standardized outcome measures such as the common data elements including PROMIS measures in future research and also to set up standardized follow-up assessment periods.

There have been a growing number of studies that develop and evaluate TBIs to improve the health outcomes of prostate cancer patients. Different research studies tend to lack the theoretical frameworks that lead the TBI development and evaluation process. Future research should integrate evidence-based and theory-guided TBIs for health management and behavioral change to determine what components, format, and dose are appropriate for prostate cancer conditions and needs, as well as how to consider biopsychosocial barriers in the use of TBIs.

5 |. CONCLUSION

Our review has found insufficient evidence to support the effectiveness of TBIs in improving health outcomes such as anxiety, depression, and HRQoL among patients with prostate cancer. However, to improve the methodological quality of this line of research, future research needs rigorous designs such as randomized control trials and sufficient sample size and power to examine the effects of TBIs on the health outcomes among patients with prostate cancer. Moreover, future research needs to examine the impact of the actual TBI use, components, sessions duration, and frequencies, as well as involvement of interventionists on patient outcomes.