記住我
Cancer is often associated with psychological distress in patients and their family members. Emerging evidence shows that psychological distress contributes to cancer mortality [,]. Given that over 2 million new cancer cases are expected to be diagnosed in 2024 in the United States, psychosocial distress is a significant public health problem []. Psychosocial distress can be triggered by many challenges, such as decision-making regarding treatment, self-care challenges due to side effects from cancer treatment, maintaining work-life balance, and financial burden. A large body of research documents the negative influence of a cancer diagnosis and treatment on a patient’s experience, including depression, anxiety, and decreased quality of life [,]. Cancer not only affects the patient but also imposes changes on the family []. Family members, who often assume caregiving roles to complement the roles of the health care team, often experience deteriorating quality of life and significant psychological distress [,]. For many years, researchers have examined psychosocial interventions addressing patients’ and family members’ needs to help maintain psychosocial well-being and quality of life during the cancer experience [-].
Increasingly, studies have used digital technology to deliver psychosocial interventions. In this report, we refer to digital health intervention as the use of digital, mobile, and wireless technologies to deliver an intervention. Digital health interventions have gained popularity due to their geographic accessibility, self-paced nature, user-friendly design, up-to-date information provision, and time-sensitive interaction with health care providers [,]. Further, digital interventions have significant potential for reaching people, mainly in rural areas or people with limited mobility []. There are various delivery modes for digital interventions, such as smartphone apps, websites, the internet, and virtual reality. There are also drawbacks, including concerns related to security and privacy and inaccessibility for people without smart device ownership. Psychosocial interventions may incorporate various components, such as communication skills training, cognitive behavioral therapy, patient education, peer support, and problem-solving training [].
Despite the plethora of individual research studies, a synthesis of digital psychosocial interventions for patients with cancer and their families is needed to provide a summary of existing evidence regarding the effects of interventions and provide directions for future research and clinical practice. A range of systematic reviews have examined digital health psychosocial interventions for patients with cancer [-] and their family members [,]. However, these reviews have limitations. For example, some reviews primarily focused on a specific population, such as individuals with breast [] or prostate cancer [,]; a particular delivery mode, such as internet-based [,,]; or a specific psychosocial outcome, such as quality of life or psychological distress [,]. In addition, Slev et al [] synthesized evidence from systematic reviews of interventions delivered through computers or the internet for patients with cancer and their caregivers; however, the authors failed to quantify the effectiveness of interventions across studies using advanced statistical techniques, such as a meta-analysis. To date, no studies have used meta-analytical strategies to quantify the impact of digital health interventions on psychosocial outcomes in patients with cancer and family members. To fill these gaps, we conducted a systematic review and meta-analysis to comprehensively review the characteristics and effectiveness of digital psychosocial interventions on psychosocial outcomes across different available delivery modes in adult patients with cancer and their family members.
The specific aims were to answer the following questions:
What are the characteristics of digital psychosocial interventions for adult patients and families living with cancer? (ie, intervention component, theoretical or conceptual framework, tailored or standardized, mode of delivery, prescribed dosage, duration of the intervention, and actual dosage)?What is the efficacy of interventions on psychosocial outcomes for adult individuals diagnosed with cancer and their family members and associated factors (ie, delivery mode, control condition, and dosage, including the number of sessions, frequency, and duration)?The review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist [].
Study IdentificationThe medical librarian (DR) and first author (YZ) worked together to identify search terms to build a comprehensive search strategy (). Using controlled vocabulary and keywords when available, the search strategy was executed in the following databases: PubMed, Cochrane Library, Web of Science, Embase, CINAHL, PsycINFO, ProQuest Dissertations and Theses Global, and ClinicalTrials.gov. The results were limited to the English language and those published from each resource’s inception until March 2019, when the search was completed. An initial limited search of PubMed and CINAHL was undertaken, followed by an analysis of the text words in the abstract and the index terms used to describe the article. Relevancy was determined by the first author (YZ) and medical librarian (DR). A second search was undertaken across all included databases using all identified keywords and index terms.
Study SelectionThe inclusion criteria were studies that (1) included adult patients (≥18 years of age with any cancer diagnosis) or their adult family members (eg, partner, caregiver, adult children, parent, or relative); (2) tested a digital health psychosocial intervention, which was defined as any nonpharmacological therapeutic intervention that addressed the psychological, social, personal, or relational adjustment needs associated with cancer through a digital health mechanism (eg, application and website); (3) measured at least 1 psychosocial outcome; and (4) used an experimental (randomized controlled trial [RCT]) or a quasi-experimental design. Studies were excluded if they enrolled pediatric patients with cancer; were review articles, letters to the editor, editorial reports, case reports, or commentaries; were published as abstracts only; and were not published in English. For meta-analysis, we excluded articles that did not provide data or when only a single study included the outcome measure.
After removing duplicates, the first author (YZ) read all titles and abstracts to identify articles based on inclusion and exclusion criteria. The full texts of all included articles were then screened independently by 2 reviewers (master’s-level or above), and final decisions were made based on consensus. Finally, articles identified in the search were imported to Endnote X8 (Clarivate Analytics).
Data Extraction and ManagementA Microsoft Excel (Microsoft Corporation) spreadsheet was used to record information [], including the description of the interventions (eg, theory basis, mode of delivery, content, actual dosage, planned dosage, standardized, or tailored), study sample (eg, age, sex, education, race, ethnicity, and cancer diagnosis), study characteristics (eg, design, randomization method, and control condition), intervention outcome variables and measurements, follow-ups, and quantitative data (ie, mean, SD, and sample size). Dosage was described as the number of intervention sessions, frequency, and duration of access to intervention. A standardized intervention was defined as all participants receiving the same intervention, while a tailored intervention involved customization of the intervention based on individual characteristics or needs []. We defined the prescribed dosage as the intended treatment dose, including the number of intervention sessions, frequency, and total length according to the study protocol. A codebook was created for data extraction, and the team’s decisions were tracked and recorded. All authors extracted data from 3 articles to pilot-test the spreadsheet. The research team discussed any ambiguity, resolved differences in interpretation, and modified the data extraction spreadsheet. Subsequently, each article underwent independent data extraction by YZ and another author (6 trained reviewers). The research team met throughout the study period every other week to resolve any discrepancies. A total of 15 original study authors were contacted to request missing information (eg, mean, SD, and sample size), and no additional data were received.
Assessment of Methodological QualityThe reviewers assessed the included studies for methodological rigor using standardized critical appraisal instruments from the 13-item Joanna Briggs Institute (JBI) Critical Appraisal Checklist for RCT and the 9-item JBI Critical Appraisal Checklist for quasi-experimental studies []. Reviewers answered each risk of bias item as “yes” (score=1), “no” (score=0), “unclear” (score=0), or “not applicable.” Possible composite scores ranged from 0 to 9 for quasi-experimental studies and 0-13 for RCTs, with higher scores indicating less risk of bias and better study quality. The applicable score (range 0-1) was calculated by dividing the composite score by the maximum score possible after subtracting any “not applicable” responses []. All studies were double-coded, and any disagreements were resolved through discussion with the research team [].
Data Synthesis and Meta-AnalysisData SynthesisData synthesis was completed on all articles that met the inclusion criteria. Only primary study results were included if multiple articles were published from the same intervention study. Simple descriptive statistics (ie, mean, SD, frequency, and percentage) were used to summarize study characteristics (eg, study design and participant characteristics) and key features of interventions (ie, theory, mode of delivery, number of sessions, frequency, and total length). Intervention content was grouped and narratively summarized according to the description of the intervention components.
Meta-Analytical ProcedureAn a priori decision was made to only include studies in the meta-analysis if at least 2 studies used the same instrument to assess the same psychosocial outcome []. Standardized mean differences (ie, Hedges g) were calculated to compare intervention effectiveness across studies that used different scales or measurements. Mean differences between the scores before the intervention and the follow-up assessment after the intervention were calculated for pre-post interventions. Similarly, for the RCT studies, the results from follow-up in each study were selected and analyzed using difference scores from before and after the intervention for both intervention and control groups, with the pooled SDs. We computed the overall effect size across different time points for studies with multiple follow-ups. By doing so, we captured the time-varying effect on intervention effectiveness []. The overall effect (including all information across all time points) and time-varying effects, including the interim effect (during the intervention period), immediate effect (after the intervention), short-term effect (follow-up ≤8 weeks after completion of the intervention), and long-term effect (follow-up >8 weeks after completion of the intervention), were calculated. A cutoff of 8 weeks was chosen because it was the median length of the follow-up period across the included studies.
To assess study heterogeneity, the I² statistic was examined. The I² statistic quantifies the proportion of total variance across studies caused by a fundamental difference between trials rather than chance. An I² statistic of <25% indicates low heterogeneity, between 25% and 75% indicates moderate heterogeneity, and >75% indicates high heterogeneity []. Lower heterogeneity is better. Funnel plots (ie, to visually assess the asymmetry) and Egger test (ie, to test the asymmetry statistically) assessed publication bias []. In funnel plots, if points are distributed equally between positive and negative effects, bias is lacking; variability is expected to be greater near the bottom of the chart among smaller sample size studies. For the analysis of data from studies with more than 1 digital psychosocial intervention group, we compared each digital psychosocial intervention group to the control group separately. Additionally, subgroup analysis was planned based on the review’s focus on examining the effect of delivery mode, type of control condition, and dosage on outcomes. Furthermore, we performed sensitivity analyses by including and excluding studies with extreme weights in the analyses. We used the DerSimonial-Laird random-effects model to weight and pool the individual estimates to capture variance across different studies, as all included studies were conducted in heterogeneous populations across various settings []. We performed all statistical and meta-analyses using STATA (version 17; StataCorp LLC).
After removing duplicates, a total of 2108 studies were identified. shows a flow diagram of studies identified, screened, included, and excluded from this systematic review and meta-analysis. After screening titles and abstracts and applying inclusion and exclusion criteria, a total of 70 records with 65 unique studies (for multiple manuscripts published from the same intervention study, only primary manuscripts were included) were included in the systematic review [-] and 32 studies [,,,,,,,,,-,,,,-,,,,,,,,,,,,] with available data were included in the meta-analysis. A total of 33 studies were excluded from the meta-analysis because either data were unavailable to calculate the effect size (n=14) [,,,,,,,,,,,,,] or no other study used the same measure (n=19) [,,,,,,,-,,,,,,,,].
Figure 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram. Study CharacteristicsOverviewOf the 65 studies, 48 (74%) were RCTs [-,-,-,-,,,-,,-,-], and 17 (26%) were quasi-experimental [,,,-,,,,,-,]. More than half (n=37, 58%) of the studies were conducted in the United States [-,,-,,-,,,-,,,,,,,-,-,,,], and the rest were from the Netherlands (n=9, 14%) [,,,,,,,,], Australia (n=5, 8%) [,,,,], and other countries (eg, Denmark and Ireland). In total, 10,361 patients (mean 159, SD 166; range 9-803 patients per study) were included: 7098 female patients and 3263 male patients; 1045 caregivers or partners were enrolled (mean 16, SD 54; range 9-244 caregivers or partners per study), including 781 female individuals and 264 male individuals. The average age of patients ranged from 39.9 to 72 years, and the average age of caregivers or partners ranged from 51.5 to 58.8 years. In the 33 studies that provided information about race and ethnicity, most patients (n=3495, 90%) and family members (n=259, 97%) were described as “White” or “Caucasian.” The cancer diagnoses varied across studies, with the most prevalent being breast cancer (n=24, 37%) [,,,,,,,,,,-,-,,,,,,,,], mixed cancer diagnosis (n=19, 29%) [,,,,,,,,,,,,,,,,,,], and prostate cancer (n=7, 11%) [,,,,,,]. The attrition rate ranged from 0% to 76%, with a median of 16.8% (mean 20.4%, SD 13.7%). The recruitment rate ranged from 4.4% to 94.2%, with a median of 59.5% (mean 56%, SD 24.6%). Detailed information about the study and sample characteristics from the included studies is provided in Table S1 in [-].
Control ConditionOf the 48 RCTs, 29 (60%) studies included a usual care control group [,,,-,,,,,,,,,,-,,,,-,], and 19 (40%) included an active control [,,,-,,,,-,,,,,,,]. Among the 17 quasi-experimental studies, 11 (65%) did not have a control group [,,-,,-], and 6 (35%) studies included a usual care control group [,,,,,].
Outcome AssessmentA total of 21 studies had 1 follow-up assessment [,,,,,,,,-,,,-,-,,], 23 had 2 follow-up assessments [,,,,-,-,,,,,,,,,,,,], 11 had 3 follow-up assessments [,-,-,,,], and 7 had 4 or more follow-up assessments [,,,,,,]. The timing of follow-up assessments varied, ranging from immediately to 6 months after the intervention. The commonly reported outcomes and relevant measures are reported below in Aim 2: Effects on Patients’ and Family Members’ Psychosocial Outcomes.
Quality Assessment (Risk of Bias)The quality assessment scores of the included studies are summarized in Table S1 in . Overall, the RCT studies’ general quality was mixed, with applicable scores ranging from 0.38 to 0.91 (mean 0.61, SD 0.12). Quasi-experimental studies were generally of moderate to high quality, with applicable scores ranging from 0.63 to 0.89 (mean 0.75, SD 0.08) on the JBI Critical Appraisal Checklist for quasi-experimental studies. The publication year and applicable appraisal score were not significantly correlated in RCTs (r=0.12; P=.40) and quasi-experimental studies (r=–0.04; P=.88).
Aim 1: Intervention CharacteristicsOverviewThere was large heterogeneity in intervention components, theoretical or conceptual framework, type of intervention (ie, tailored or standardized), mode of delivery, prescribed dosage (ie, number of sessions, frequency, and length), and received dosage (Table S2 in ).
Intervention ComponentsA total of 37 (57%) out of 65 studies included a single intervention component [,-,,,,,,-,-,,,-,,,-,,,,,,,], 13 (20%) studies included 2 intervention components [,,,,,,,,,,,,], and 15 (23%) studies included 3-5 intervention components [,,,,,-,,,,,,]. The most common intervention components were information and resources, or psychoeducation (n=29, 45%) [,,-,,,,-,-,,,,-,,,], and cognitive-behavioral strategies (n=20, 31%) [,,,,,,,,,,,,,,-,,,].
Theoretical or Conceptual FrameworkMore than half (n=38, 59%) of the included studies did not identify a conceptual or theoretical framework [,,,-,,,,,-,,,,,,,-,-,,,,,].
Standardized or Tailored InterventionOf the 65 studies, 26 (40%) included both standardized and tailored interventions [,,,,,,,,,-,,,,-,-,,,,,], 28 (43%) studies included only standardized interventions [,,,,,,,-,,-,,,,,,-,,,,,,], and 11 (17%) studies had only tailored interventions [,,,,,,,,,,].
Modes of DeliveryThe majority of studies conducted interventions through an internet website (n=40, 62%) [-,-,,,-,,,,,,,,,,,-,,-] or smart device app (n=8) [,,,,,,,]. A total of 7 (11%) studies conducted interventions through virtual reality [,,,,,,], 3 (5%) studies through telehealth [,,], and 2 (3%) studies through a computer program [,]. Electronic health information systems [], interaction portals [], and videoconferences [] were each used in 1 study. Overall, 2 studies used multimodal interventions delivered through the combination of either telephone and videoconference [] or internet and telephone [].
DosageThe dosage prescribed and received were highly variable. The number of intervention sessions ranged from 1 to 56, with a median of 6. A total of 27 (42%) studies did not specify the prescribed dose; 19 (29%) only stated the number of days participants had access to the intervention [,,,,,-,,,,,,,,,,] and 8 (12%) did not provide information on the prescribed dosage [,,,,,,,]. Frequency was highly variable, with self-paced (n=26, 40%) as the most common [,,,,,,,-,,,,,,-,,-,], meaning no specific intervention frequency was defined and the intervention content was available throughout the study period. The other common frequencies of intervention sessions were weekly (n=17, 26%) [,,,,,,,,,,,,,,,,] and 1-time intervention sessions (n=8, 12%) [,,,,,-]. The median length of the intervention was 8 weeks, with the length ranging from 1 hour (ie, use of the intervention on an iPad for an hour) to 24 months. Received dosage was defined as the uptake of the intervention by the participants. A total of 18 (28%) studies did not report the received dosage [,,,,,,,,,,,,,,,,,]. Various information was reported, including attendance rate, number of times participants used the app, frequency with which participants logged into the website, number of website pages reviewed, skill practice time, and intervention session completion rate. Most of the interventions (n=43, 66%) were self-delivered without an interventionist, with self-paced being most common [-,,,,,,-,,-,,,,-,,,,-,,,-,,,].
Aim 2: Effects on Patients’ and Family Members’ Psychosocial OutcomesPatients’ OutcomesOverviewA meta-analysis was conducted on 32 studies. Overall, 5 outcomes were examined. A summary of the interventions’ overall effect sizes; time-varying effect sizes for quality of life, anxiety, depression, distress, and self-efficacy; and heterogeneity statistics for each outcome is displayed in . The forest plots for overall effect sizes and time-varying effects are displayed in . The funnel plots for overall effect sizes and time-varying effects are displayed in .
Table 1. Summary of the meta-analysis.Population, outcome, measure, and valueEffect at different time pointsaThe overall effect accounts for time-varying effect across different time points.
bQOL: quality of life.
cFACT-B: Functional Assessment of Cancer Therapy–Breast.
dES: effect size.
eNot applicable.
fFACT-G: Functional Assessment of Cancer Therapy–General.
gQLQ-30: Quality of Life Questionnaire, 30 items.
hSF36: Short Form Survey 36-item.
iHADS: Hospital Anxiety and Depression Scale.
jCESD: Center for Epidemiologic Studies Depression Scale.
kPHQ9: Patient Health Questionnaire-9.
lSATI: State-Trait Anxiety Inventory.
mDT: Distress Thermometer.
nCBI: Coping Behaviors Inventory.
Quality of LifeQuality of life was measured by the Functional Assessment of Cancer Therapy–Breast [,,,,], Functional Assessment of Cancer Therapy–General [,,,,], European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, 30-items [,,,,,,], and 36-item Short Form Survey [,,,,]. Overall, a total of 21 studies with 1847 participants in the intervention groups showed an increase in quality of life, with a mean difference between groups of Hedges g=0.05 (95% CI –0.01 to 0.10). The impact of heterogeneity within the studies was significant (I2=42.7%; P=.01). With respect to publication bias, the funnel plot displayed a greater number of studies toward the top of the mean (Egger test, P<.001). The time-varying effects were as follows: Hedges g=–0.16 (95% CI –0.39 to 0.06) for the interim effect; Hedges g=0.95 (95% CI –1.99 to 3.89) for the immediate effect; Hedges g=2.25 (95% CI 0.36-4.14) for the short-term effect; and Hedges g=0.18 (95% CI 0-0.35) for the long-term effect. The statistical heterogeneity among studies was I2=70% (P=.07) for the interim effect; I2=100% (P<.001) for the immediate effect; I2=98% (P<.001) for the short-term effect; and I2=18.3% (P=.29) for the long-term effect.
Anxiety and DepressionHospital Anxiety and Depression Scale (HADS) total scores (without subscale scores reported) were reported in 5 studies with 338 participants in the intervention groups [,,,,]. Overall, participants receiving interventions reported decreased anxiety and depression with a standardized mean difference of Hedges g=–0.72 (95% CI –1.89 to 0.46). The heterogeneity within the studies was significant (I2=97.6%; P<.001). The funnel plot was found to be asymmetric, and Egger test was found to be not statistically significant (P=.77). The time-varying effects were as follows: Hedges g=–0.04 (95% CI –0.23 to 0.16) for the immediate effect; Hedges g=–0.22 (95% CI –0.54 to 0.10) for the short-term effect; and Hedges g=0.14 (95% CI –0.09 to 0.38) for the medium-term effect. The statistical heterogeneity among studies was I2=0% across all time-varying effects.
DepressionDepression was assessed by the HADS-depression subscale [,,,,,,], Center for Epidemiologic Studies Depression Scale [,,,], Patient Health Questionnaire-9 (PHQ-9) [,], and a combination of the PHQ-9 and HADS-anxiety [,] in 1509 participants in the intervention groups. Overall, interventions were not more effective than control conditions for reducing depression (Hedges g=0.03, 95% CI –0.10 to 0.16), with a high heterogeneity of 60.9% (P<.001). With respect to publication bias, the funnel plot displayed a greater number of studies toward the top of the mean (Egger test, P=.25). The time-varying effects were as follows: Hedges g=0.06 (95% CI –0.10 to 0.22) for the immediate effect and Hedges g=–0.04 (95% CI –0.22 to 0.14) for the interim effect. The statistical heterogeneity among studies was I2=69.4% for the immediate effect and I2=29.8% for the interim effect.
AnxietyAnxiety was assessed by the HADS-anxiety subscale [,,,,,,,], State-Trait Anxiety Inventory (STAI) [,,,,], and a combination of the STAI and HADS-anxiety [] in 1075 participants in the intervention groups. Overall, interventions were not more effective than control conditions for reducing anxiety (Hedges g=0.12, 95% CI –0.19 to 0.43), with high heterogeneity of 90.2% (P<.001). The funnel plot displayed a greater number of studies toward the top of the mean (Egger test, P=.46). The interim effect was Hedges g=–0.04 (95% CI –0.19 to 0.12), and the immediate effect was Hedges g=–0.10 (95% CI –0.19 to 0), and the short-term effect was Hedges g=–0.13 (95% CI –0.43 to 0.17). The statistical heterogeneity among studies was I2=35.1% for the interim effect, I2=6.7% for the immediate effect, and I2=10.5% for the short-term effect.
DistressPsychological distress was assessed in 182 participants in the intervention groups using the distress thermometer [,,]. Overall, participants in the intervention groups showed no reduction in distress, with a mean difference between groups of Hedges g=0.98 (95% CI –0.18 to 2.14). The impact of heterogeneity within the studies was significant (I2=98.5%; P<.001). Regarding publication bias, the funnel plot displayed a symmetric distribution around the mean effect (Egger test, P=.46). The immediate effect was Hedges g=0.51 (95% CI 0.10-0.92), with statistical heterogeneity I2=54.2%.
Self-EfficacySelf-efficacy was measured by the Coping Behaviors Inventory in 174 participants in the intervention groups [,]. Overall, participants in the intervention groups did not report improvement in self-efficacy, with a standardized mean difference of Hedges g=–1.41 (95% CI –4.02 to 1.20). However, the impact of heterogeneity within studies was significant (I2=99%; P<.001). Regarding the publication bias, the funnel plot displayed a symmetric distribution around the mean effect (Egger test, P=.22). The immediate effect was Hedges g=2.56 (95% CI –1.22 to 6.35) with high heterogeneity (I2=98.2%; P<.001).
Subgroup AnalysesGiven the heterogeneity of reporting on dosage information and limited data, the subgroup analysis of dosage on intervention effect was not conducted. includes the results of the subgroup analysis on the effect on quality of life, depression and anxiety, and distress. Overall, the associations between delivery mode and control condition with patient outcomes were not statistically significant (P>.05).
Table 2. Subgroup analyses on the effect of delivery mode (internet vs noninternet) and control condition (usual care vs active control) on patient outcomes.Outcome and moderatorsEffect size, Hedges g (95% CI)SEP valueQuality of life (27 studies)
留言 (0)