ABSTRACT
Background
Although depression is prevalent and has significant consequences among individuals living with likely incurable cancer (ILLIC), optimal methods of identifying and treating depression in this population remain unknown.
Aims
To evaluate a paradigm of (1) proactive identification (ID) (i.e., remotely and asynchronously from clinical encounters) of depression among ILLIC and (2) digital mental health intervention (DMHI) for depression treatment.
Methods
In this decentralized randomized clinical trial, ILLIC with elevated depressive symptoms were proactively identified using electronic health record data and randomized 2:1 to a DMHI‐based Behavioral Activation treatment or usual care (UC) depression treatment. Measures of feasibility (accrual, retention) and acceptability (engagement) were described; depression severity (change in PHQ‐9 scores through 4 weeks post‐randomization) was modeled with a generalized estimating equation.
Results
Among 88 ILLIC who completed screening, 30 were eligible and randomized to the trial. No patients were lost to follow‐up or withdrew; 80% of patients randomized to proactive ID + DMHI used the app through 4 weeks. Proactive ID + DMHI improved depression from baseline to 4 weeks relative to proactive ID + UC (mean difference in change from baseline to week 4 = −2.7; 90% CI: −4.9 to −0.4). At 4 weeks, the odds of a clinical response (PHQ‐9 decrease of ≥ 5 points) was 9.0‐fold higher for patients in proactive ID + DMHI relative to proactive ID + UC (OR 9.0; 90% CI: 1.1–74.2).
Conclusions
A proactive ID + DMHI approach to identifying and treating depression among ILLIC is feasible, acceptable, and potentially efficacious. These promising data support conducting a large efficacy trial evaluating this approach.
Trial Registration
ClinicalTrials.gov identifier: NCT05932810
Keywords: cancer, depression, digital mental health, metastatic, oncology, randomized clinical trial, survivor
1. Introduction
With improvements in treatment and supportive care, there are a growing number of individuals living longer with advanced or metastatic cancer that is likely incurable [1, 2]. This understudied survivor subpopulation is highly heterogenous and has unique survivorship care needs [1, 3]. Individuals living with likely incurable cancer (ILLIC) may transition on and off therapy, have considerable prognostic uncertainty, and experience a unique symptom profile due to cumulative effects of multiple cancer‐directed therapies and novel agents. Living longer with likely incurable cancer often comes with burdensome psychosocial symptoms and up to 50% of people living with advanced or metastatic cancer report depressive symptoms [4, 5, 6]. Depression in this patient population is highly significant as it is associated with worse quality of life (QOL), reduced adherence to anti‐cancer therapies, suicidal ideation, and desire for hastened death [7, 8].
Although numerous trials [9, 10] and meta‐analyses [5, 11] have documented that evidence‐based psychosocial treatment improves depression outcomes for people living with advanced or metastatic cancer, multilevel barriers limit access. Patients lack treatment information, experience stigma, do not receive treatment referrals, and are faced with geographic barriers to care [12, 13]. Providers lack treatment knowledge and have limited time and resources [14]. Organizational barriers, such as unclear referral pathways, further limit uptake [14]. Novel care delivery approaches that address these multilevel barriers and extend psychosocial treatment access for ILLIC are clearly needed.
One promising approach to improve access to depression treatment is to proactively link patients with evidence‐based treatment outside the context of (i.e., asynchronous from) a medical encounter [15, 16]. As a pre‐requisite to this care delivery approach, it is necessary to first identify patients with (1) elevated depressive symptoms and (2) likely incurable advanced and metastatic cancer. Currently, information available in the electronic health record (EHR) can be leveraged to accomplish both goals. For example, short depression screening tools (e.g., PHQ‐2 [17]) are routinely administered in oncology care settings to systematically identify people with cancer at risk for experiencing elevated depressive symptoms [18, 19]. In addition, patients with advanced or metastatic cancer that is likely incurable can be identified using both structured and unstructured data from the EHR [20].
Digital mental health interventions (DMHIs), which offer adaptations of evidence‐based mental health interventions, are another component of a promising strategy to improve access to mental health treatment in this population. DMHIs overcome many of the barriers to depression treatment among cancer survivors including stigma, a shortage of trained mental health professionals, and geographic barriers. DMHIs have comparable effects to standard psychosocial interventions [21, 22]. Although trials have demonstrated that DMHIs are feasible, acceptable, and may improve depressive symptoms among cancer survivors [23, 24, 25, 26], they have not been evaluated specifically among ILLIC. Self‐guided DMHIs, which allow on‐demand access with each survivor selecting their optimal frequency and duration, are particularly well suited to the unique needs of ILLIC. Unlike traditional psychotherapies and guided DMHIs which present content in a “sessionized” fashion, the on‐demand scheduling and flexible, self‐directed pace of the self‐guided DMHI can accommodate the fluctuating and unpredictable physical and mental health symptoms in this population related to unpredictable cancer progression, accumulating treatment toxicity, and cycling on and off of cancer‐directed therapy. Among DMHIs that could be deployed to treat depression among ILLIC, those featuring Behavioral Activation (BA) treatment are particularly appealing because BA specifically targets decreased activity level, a key mechanism purported to underly depression in this population [27, 28].
To address the lack of adequate mental health care delivery for people living with likely incurable advanced or metastatic cancer and depressive symptoms, we conducted a pilot randomized clinical trial (RCT) to evaluate the feasibility, acceptability, and preliminary efficacy of proactive identification (ID) + DMHI versus proactive ID + Usual Care (UC) for treating depression in this population.
2. Methods
This decentralized 2‐arm parallel‐group RCT was conducted among patients receiving care from the Medical University of South Carolina (MUSC) Hollings Cancer Center. The study was approved by the MUSC institutional review board (Pro00126828) and registered at ClinicalTrials.gov (NCT05932810). Trial information is presented according to the CONSORT extensions for randomized pilot [29], psychological intervention trials [30], and e‐health trials [31].
2.1. Participants
Eligible patients were ≥ 18 years of age, had any type of advanced or metastatic cancer that was likely incurable, reported current elevated depressive symptoms (defined per ASCO Guidelines [8] as a PHQ‐9 [32] score ≥ 8), owned an iOS‐ or Android‐compatible smartphone, and had English fluency. Patients were excluded if they had severe cognitive impairment precluding informed consent (self‐report or diagnosis of dementia or neurocognitive disorder) or if they had current suicidal ideation (≥ 2 on PHQ‐9 item 9). If a participant screened positive for suicidal ideation during the eligibility assessment (or during post‐randomization follow‐up), they were contacted for completion of a risk assessment and provided with follow‐up care as appropriate. Potentially eligible patients were sent a study invitation via the EHR patient portal (MyChart), text message, email, and/or phone (based on preferences in the EHR) and underwent remote final eligibility assessment, including completion of the PHQ‐9 [32] to ensure current depression at accrual.
2.2. Procedures
Following remote informed e‐consent and completion of baseline assessments, patients were randomized 2:1 to proactive ID + DMHI or proactive ID + UC using a mixed block design with block sizes of 3, 6, and 9. The 2:1 randomization allocation was selected to maximize information about the acceptability of proactive ID + DMHI. The randomization sequence was generated by the biostatistician using a computer‐generated algorithm and implemented in REDCap. Patients were compensated $40 for completion of baseline assessments, $40 for completion of each post‐baseline assessment, and $100 if all study assessments were completed (total potential compensation = $300).
2.3. Proactive ID + DMHI (Intervention)
Proactive ID refers to the use of structured and unstructured EHR data asynchronous from clinical encounters to proactively identify patients who are (1) living with advanced or metastatic cancer that is likely incurable and (2) who likely have elevated depressive symptoms. This paradigm is contrasted with the standard of care approach of screening patients for depression during a routine clinical encounter and linking them in real time to mental health treatment. To date, the gold standard method of identifying ILLIC is through manual review of the EHR. To proactively identify ILLIC for this trial, we performed an automated EHR search of all patients with a cancer diagnosis who had undergone or were undergoing treatment and had a clinical encounter at the MUSC Hollings Cancer Center on Jan 1, 2023 or later. We then manually reviewed charts to identify those with likely incurable advanced or metastatic cancer. To proactively identify the subset of survivors likely to have elevated depressive symptoms, we performed an automated EHR search to identify patients with a PHQ‐2 score of ≥ 3 (i.e., positive screening [33]), depression on their problem list, or a depression‐related billing code associated with their last visit.
The DMHI deployed within this trial was “Moodivate,” a DMHI‐based approach to deliver Behavioral Activation (BA) Treatment for Depression [16, 34, 35]. BA is an evidence‐based first‐line depression treatment for cancer survivors [8, 27, 28, 36] which aims to help patients identify, schedule and reengage in positive activities. BA is ideal for people living with advanced or metastatic cancer that is likely incurable because it specifically targets decreased activity level, a key mechanism purported to underly depression in this population [27, 28]. As illness progresses among ILLIC, daily tasks may be abandoned, leading to decreased activity levels, an environment that perpetuates depressed mood, and subsequent exacerbation of depressive symptoms. BA works by helping ILLIC re‐engage in meaningful activities, thereby improving mood.
Moodivate has been described previously [16, 34, 35]. For this study, psychoeducational content was tailored within the app specifically for survivors living with advanced or metastatic cancer that is likely incurable. Moodivate is available on both iOS and Android and is free for initial download. Following successful app download, study staff provide a brief, scripted overview regarding app utilization and provide the patient with 10 min to use the app and ask questions. Moodivate users can track daily mood to monitor progress and are prompted to complete a clinical assessment of depressive symptoms once every 2 weeks. Users can view a graph of mood and depressive symptoms overlaid upon a graph of the number of completed activities, illustrating the connection between activity and mood. Gamification elements (e.g., badges) promote continued engagement with Moodivate over time [34]. Patients are encouraged to use the app regularly at least once per day, throughout the study and have access to the app through the entire study period and beyond. As a self‐guided DMHI, there is no prescribed minimum “dose” of Moodivate; the app may be used for shorter or longer duration than the 8–12 weeks suggested in the original BA manual [37] depending on user preference.
2.4. Proactive ID + UC (Control)
Proactive ID of ILLIC and elevated depressive symptoms was performed in an identical fashion for those randomized to UC. UC mimics existing depression treatment for people living with advanced or metastatic cancer that is likely incurable. Participants in either arm could receive treatment (e.g., pharmacotherapy) or a referral for evidence‐based psychotherapy (e.g., cognitive behavioral therapy) either within our integrated behavioral medicine clinic or to an outside mental health provider from their oncology team as part of their usual clinical care (i.e., unrelated to trial participation).
2.5. Study Measures and Outcomes
Demographic data were collected via self‐report. Clinical characteristics were extracted from the EHR. All patient‐reported study assessments were collected via a REDCap link which was sent to the participant via text message and/or email. Concomitant mental health treatment was measured at baseline and all post‐randomization visits by self‐report. Rurality was defined based on rural‐urban commuting area (RUCA) codes with RUCA codes of 1‐3 classified as non‐rural and 4–10 as rural. Digital literacy was assessed at baseline via the Mobile Device Proficiency Questionnaire (MDPQ‐16) [38]. The MDPQ‐16 is a validated, 16‐item self‐report measure of mobile device proficiency. Scores range from 0 to 40 with higher scores indicating greater mobile device proficiency.
For the primary objective of evaluating the feasibility of conducting a fully decentralized RCT comparing proactive ID + DMHI with proactive ID + UC, we selected feasibility endpoints of accrual and trial retention. For the secondary objective of determining the acceptability of proactive ID + DMHI among depressed ILLIC, we selected a battery of passively collected app analytics data (collected continuously post‐randomization), a quantitative program evaluation with open‐ended feedback, and the System Usability Scale (SUS) [39] as measured at week 4. The SUS is a validated 10‐item self‐report measure of acceptability. The SUS score ranges from 0 to 100; higher scores indicate better usability.
Depressive symptoms were measured using the PHQ‐9 [32] at baseline and weekly for 4 weeks post‐randomization for participants in both arms. The PHQ‐9 score ranges from 0 to 27; higher scores represent more severe depressive symptoms and severity of depressive symptoms is categorized as moderate (8–14), moderately severe (15–19), and severe (20–27). An additional depression‐related endpoint was a clinically important reduction in depressive symptoms, defined as a decrease in PHQ‐9 score ≥ 5 points from baseline to week 4. Anxiety and health‐related quality of life (HRQOL) were measured with the Hospital Anxiety and Depression Scale, Anxiety subscale (HADS‐A) [40] and Functional Assessment of Cancer Therapy‐7 item version (FACT‐G7) [41], respectively, at baseline and weekly for 4 weeks post‐randomization for participants in both arms. Lower scores on the HADS‐A indicate lesser anxiety while higher scores on the FACT‐G7 indicate better HRQOL. Per protocol, assessments were to be completed within a 72‐h window of their assignment; all assessments were completed within the protocol‐specified window except one end of study assessment at week 4 for one participant.
2.6. Statistical Analysis
Feasibility and acceptability endpoints were summarized by mean (SD) or median (IQR) descriptive statistics. A generalized estimating equation (GEE) with identity link was used to model expected PHQ‐9 scores at weeks 1–4 as a function of treatment group, time (continuous), and treatment group‐by‐time interaction, with adjustment for baseline PHQ‐9 score and gender. A compound symmetric correlation structure was used, and the variance of the fitted model parameters was estimated using the robust sandwich estimator. The functional form of time was evaluated and no significant departure from linearity was detected. The fitted model was used to estimate average change from baseline for the proactive ID + DMHI and proactive ID + UC arms. To estimate average group scores based on the fitted linear predictor, coefficients for baseline PHQ‐9 and gender were weighted using the average overall baseline PHQ‐9 score and the proportion of male study participants, respectively. Linear contrasts were constructed to estimate the difference between arms in change from baseline at week 4. A similar GEE modeling approach was utilized to analyze the HADS‐A and FACT‐G7 scores. The probability of a ≥ 5‐point change in PHQ‐9 score from baseline to week 4 was modeled using a binary regression generalized linear model with logit link and treatment arm, baseline PHQ‐9 score and gender as independent variables. An odds ratio (OR) and corresponding 90% confidence interval (CI) were reported to evaluate the association between treatment arm and the incidence of a > 5‐point change in PHQ‐9 score from baseline to week 4.
All participants randomized to proactive ID + DMHI (n = 20) or proactive ID + UC (n = 10) were included in the regression analyses of PHQ‐9, HADS‐A, and FACT‐G7 scores. However, one participant did not have week 4 measurements due to mortality. A change from baseline to week 4 could not be estimated for this participant and they were not included in the analysis of a ≥ 5 point change in PHQ‐9 score.
Aligned with the pilot stage of the research and the intention to identify promising findings for follow‐up evaluation in a fully powered RCT, model‐based differences between treatment arms were used to assess the direction and strength of the efficacy signal for changes in score from baseline to week 4 on the PHQ‐9, FACT‐G7, and HADS‐A. We constructed 90% confidence intervals to convey uncertainty around these point estimates.
3. Results
The CONSORT diagram is shown in Figure 1. Patients who did not complete the screening eligibility assessment (n = 261) were similar to those who completed the assessment (n = 88) in terms of demographic characteristics except that Black patients were less likely to complete the assessment than White patients (Supporting Information S1: Table 1). Baseline characteristics for patients randomized to proactive ID + DMHI (n = 20) and proactive ID + UC (n = 10) are summarized in Table 1. The median (IQR) age was 57.5 (47.0, 65.5) years; 40% of patients identified as male; 23% self‐reported being from a racial or ethnic minority population, and 13% lived in rural areas. Overall, 73% of patients had metastatic cancer that was likely incurable and 27% had locoregionally advanced cancer that was likely incurable. At randomization, 30% of participants were receiving psychotherapy for their depression and 80% were on pharmacotherapy for their depression. Post‐randomization usage of concomitant mental health therapies through the end of the study was similar to baseline (Proactive ID + DMHI: 35% psychotherapy, 70% pharmacotherapy; Proactive ID + UC: 20% psychotherapy; 90% pharmacotherapy).
FIGURE 1.
CONSORT diagram. Total of ineligible patients by reason sums to > n = 36 because patients may be ineligible for > 1 concurrent reason.
TABLE 1.
Baseline demographic and clinical characteristics.
| Variable | Proactive ID + DMHI (N = 20) | No. (%) | Overall (N = 30) |
|---|---|---|---|
| Proactive ID + UC (N = 10) | |||
| Age, median (IQR), years | 60.0 (50.0, 66.2) | 56.5 (43.2, 62.8) | 57.5 (47.0, 65.5) |
| Sex | |||
| Male | 10 (50%) | 2 (20%) | 12 (40%) |
| Female | 10 (50%) | 8 (80%) | 18 (60%) |
| Gender | |||
| Male | 10 (50%) | 2 (20%) | 12 (40%) |
| Female | 10 (50%) | 8 (80%) | 18 (60%) |
| Race and ethnicity | |||
| Non‐Hispanic White | 15 (75%) | 8 (80%) | 23 (77%) |
| Hispanic White | 0 (0%) | 0 (0%) | 0 (0%) |
| Black | 4 (20%) | 1 (10%) | 5 (17%) |
| American Indian or Alaskan Native | 1 (5%) | 0 (0%) | 1 (3%) |
| Multiracial | 0 (0%) | 1 (10%) | 1 (3%) |
| Health insurance | |||
| Private, employee provided | 3 (15%) | 1 (10%) | 4 (13%) |
| Private, other | 6 (30%) | 2 (20%) | 8 (27%) |
| Medicare | 6 (30%) | 4 (40%) | 10 (33%) |
| Medicaid | 2 (10%) | 2 (20%) | 4 (13%) |
| Other | 3 (15%) | 0 (0%) | 3 (10%) |
| None | 0 (0%) | 1 (10%) | 1 (3%) |
| Marital status | |||
| Married | 14 (70%) | 6 (60%) | 20 (67%) |
| Separated/divorced | 3 (15%) | 3 (30%) | 6 (20%) |
| Widowed | 2 (10%) | 0 (0%) | 2 (7%) |
| Single/never married | 1 (5%) | 1 (10%) | 2 (7%) |
| Highest level of education | |||
| Some high school | 0 (0%) | 2 (20%) | 2 (7%) |
| High school graduate or GED | 4 (20%) | 1 (10%) | 5 (17%) |
| Some college | 11 (55%) | 3 (30%) | 14 (47%) |
| College graduate (4 years or more) | 4 (20%) | 2 (20%) | 6 (20%) |
| Graduate school/Advanced degree | 1 (5%) | 2 (20%) | 3 (10%) |
| Household income | |||
| < $25,000 | 6 (30%) | 0 (0%) | 6 (20%) |
| $25,000–$50,000 | 5 (25%) | 4 (40%) | 9 (30%) |
| $50,000–$75,000 | 4 (20%) | 2 (20%) | 6 (20%) |
| $75,000 = $100,000 | 1 (5%) | 0 (0%) | 1 (3%) |
| > $100,000 | 3 (15%) | 4 (40%) | 7 (23%) |
| Don't know/not sure | 1 (5%) | 0 (0%) | 1 (3%) |
| Employment status | |||
| Full‐time | 2 (10%) | 1 (10%) | 3 (10%) |
| Unemployed | 6 (30%) | 1 (10%) | 7 (23%) |
| Disabled | 4 (20%) | 4 (40%) | 8 (27%) |
| Employed part time | 1 (5%) | 1 (10%) | 2 (7%) |
| Retired | 6 (30%) | 3 (30%) | 9 (30%) |
| Other | 1 (5%) | 0 (0%) | 1 (3%) |
| Rurality | |||
| Non‐rural | 16 (80%) | 10 (100%) | 26 (87%) |
| Rural | 4 (20%) | 0 (0%) | 4 (13%) |
| Baseline depression severity (PHQ‐9), mean (SD) | 12.8 (3.3) | 14.0 (2.7) | 13.2 (3.1) |
| Baseline depression severity (PHQ‐9 score) | |||
| Moderate (8–14) | 14 (70%) | 4 (40%) | 18 (60%) |
| Moderately severe to severe (15–27) | 6 (30%) | 6 (60%) | 12 (40%) |
| Concomitant mental health treatment a | |||
| Psychotherapy | 7 (35%) | 2 (20%) | 9 (30%) |
| Pharmacotherapy | 15 (75%) | 9 (90%) | 24 (80%) |
| Cancer type b | |||
| Brain | 2 (10%) | 1 (10%) | 3 (10%) |
| Breast | 3 (15%) | 0 (0%) | 3 (10%) |
| Colon | 1 (5%) | 1 (10%) | 2 (7%) |
| Gynecologic | 1 (5%) | 1 (10%) | 2 (7%) |
| Lung | 1 (5%) | 1 (10%) | 2 (7%) |
| Multiple myeloma | 2 (10%) | 2 (10%) | 4 (13%) |
| Neuroendocrine | 2 (10%) | 1 (10%) | 3 (10%) |
| Pancreas | 1 (5%) | 1 (10%) | 2 (7%) |
| Prostate | 4 (20%) | 0 (0%) | 4 (13%) |
| Renal | 1 (5%) | 1 (10%) | 2 (7%) |
| Other b | 2 (10%) | 1 (10%) | 3 (10%) |
| Type of likely incurable cancer | |||
| Advanced | 5 (25%) | 3 (30%) | 8 (27%) |
| Metastatic | 15 (75%) | 7 (70%) | 22 (73%) |
| Time since diagnosis of likely incurable cancer, median (IQR), years | 2.5 (1.5, 3.6) | 2.5 (2.2, 5.4) | 2.5 (1.8, 4.1) |
| Current cancer therapy a | |||
| Chemotherapy | 7 (35%) | 3 (30%) | 10 (33%) |
| Hormone therapy | 4 (20%) | 0 (0%) | 4 (13%) |
| Immunotherapy | 2 (10%) | 2 (20%) | 4 (13%) |
| Targeted therapy | 8 (40%) | 3 (30%) | 11 (37%) |
| None (treatment break) | 5 (25%) | 4 (40%) | 9 (30%) |
| Duration of cancer‐directed therapy, median (IQR), years | 4.0 (2.0, 8.7) | 2.2 (2.0, 3.7) | 2.5 (2.0, 5.5) |
| ECOG performance status | |||
| 0 | 7 (35%) | 4 (40%) | 11 (37%) |
| 1 | 4 (20%) | 3 (30%) | 7 (23%) |
| 2 | 4 (20%) | 0 (0%) | 4 (13%) |
| 3 | 1 (5%) | 0 (0%) | 1 (3%) |
| Unknown | 4 (20%) | 3 (30%) | 7 (23%) |
| Charlson comorbidity index | |||
| 0 | 15 (75%) | 6 (60%) | 21 (70%) |
| 1 | 3 (15%) | 2 (20%) | 5 (17%) |
| 2 | 0 (0%) | 1 (10%) | 1 (3%) |
| ≥ 3 | 2 (10%) | 1 (10%) | 3 (10%) |
| Cell phone type | |||
| iPhone | 11 (55%) | 7 (70%) | 18 (60%) |
| Android | 9 (45%) | 3 (30%) | 12 (40%) |
| Mobile device health proficiency Questionnaire‐16 score, median (IQR) | 37.0 (32.4, 40.0) | 39.0 (33.8, 40.0) | 37.8 (33.0, 40.0) |
Percentages can sum > 100% as a participant can have more than one current therapy.
Other = 1 (DMHI) Hematologic, 1 (DMHI) Melanoma, 1 (UC) Sarcoma.
3.1. Feasibility and Acceptability
For the primary feasibility endpoint of accrual to the RCT, among 349 ILLIC with likely depression identified in the EHR, 88 (25%) completed screening, and 30 met trial eligibility criteria. Among the 30 eligible potential participants, 100% were enrolled and randomized to the trial. For the feasibility endpoint of retention, one participant in proactive ID + UC died between weeks 3 and 4; no patients were lost to follow‐up or withdrew from the study.
In terms of acceptability among patients randomized to the Proactive ID + DMHI arm, the Moodivate app was rated as having good usability (mean [SD] System Usability Scale score = 73.0 [18.9] at week 4). Table 2 shows engagement metrics for participants randomized to proactive ID + DMHI. Engagement was high, as 80% of patients randomized to proactive ID + DMHI continued to use the app through 4 weeks. Participants completed a mean (SD) of 35.9 (26.6) Moodivate app sessions and scheduled a mean (SD) of 66.8 (114.7) activities within the app. The quantitative program evaluation and open‐ended questions regarding the acceptability of the proactive ID + DMHI approach are shown in Supporting Information S1: Table 2.
TABLE 2.
Moodivate acceptability and engagement (n = 20).
| Acceptability measure | Mean (SD) |
|---|---|
| Number of Moodivate app sessions | 35.9 (26.6) |
| Time per Moodivate session, minutes | 3.6 (2.3) |
| Total time using the Moodivate app, minutes | 107.1 (89.7) |
| Number of goals created within Moodivate | 4.0 (6.0) |
| Number of activities created within Moodivate | 3.2 (4.4) |
| Number of activities scheduled within Moodivate a | 66.8 (114.7) |
| Number of activities completed a | 28.3 (41.0) |
| Use of app within week following download (engagement) | No. (%) |
|---|---|
| Week 1 | 20 (100%) |
| Week 2 | 17 (85%) |
| Week 3 | 17 (85%) |
| Week 4 | 16 (80%) |
Number of activities created refers to the number of unique activities created (e.g., “go for a 10‐min walk”). However, each activity created could be scheduled and completed > 1 time (e.g., the person could schedule the activity of “go for a 10‐min walk” daily for 7 days). As a result, it is possible for the number of activities scheduled and completed to exceed the number of activities created.
3.2. Depression
The estimated average change in PHQ‐9 score from baseline over time is shown in Figure 2. For participants in the proactive ID + DMHI arm, PHQ‐9 scores decreased by a mean of 3.2 points (90% CI: −4.5 to −1.9) from baseline to week 4. Proactive ID + DMHI improved depression severity as measured by PHQ‐9 scores relative to proactive ID + UC (mean difference in change in PHQ‐9 score from baseline to week 4 = −2.7; 90% CI: −4.9 to −0.4). At 4‐week post‐randomization, 45% (9/20; exact 90% CI: 26%–65%) of patients in proactive ID + DMHI experienced a clinically important decrease in depression severity (i.e., decrease in PHQ‐9 of ≥ 5 points). The odds of experiencing a clinically important decrease in depression was 9 times higher for patients randomized to proactive ID + DMHI relative to those randomized to proactive ID + UC (OR = 9.0; 90% CI 1.1–74.2).
FIGURE 2.
Mean change from baseline of PHQ‐9 scores for patients in proactive ID + DMHI versus proactive ID + UC. Line graph demonstrating the mean change from baseline in PHQ‐9 scores over time by intervention allocation with 90% CIs. Proactive ID + DHMI arm: n = 20 patients at all timepoints; Proactive ID + UC arm: n = 10 at all timepoints except n = 9 at week 4.
3.3. Anxiety and HRQOL
Proactive ID + DMHI decreased anxiety over time (change in HADS‐A score from baseline to week 4 = −1.7; 90% CI: −2.8 to −0.6), although this result did not differ relative to proactive ID + UC (mean difference in change in HADS‐A score from baseline to week 4 = −1.0; 90% CI: −2.8 to 0.8). Proactive ID + DMHI improved HRQOL over time (change in FACT‐G7 score from baseline to week 4 = 2.0; 90% CI: 0.8–3.2). Proactive ID + DMHI also improved HRQOL relative to proactive ID + UC (mean difference in change in FACT‐G7 score from baseline to week 4 = 2.9; 90% CI: 1.1–4.7).
4. Discussion
In this pilot RCT, we demonstrated that a proactive ID + DMHI approach to treating depression is feasible and acceptable among ILLIC and results in high levels of app engagement. These preliminary data demonstrate that this treatment paradigm may improve depression severity relative to proactive ID + UC in this patient population. Although we are cautious to over‐interpret these results, particularly the wide confidence intervals, the finding that 45% of patients in proactive ID + DMHI experienced a clinically meaningful improvement in depression severity, which is 9‐times the odds of patients in proactive ID + UC, suggests that the benefits of this intervention have the potential to be realized by a large proportion of patients with advanced and metastatic cancer suffering from depression.
4.1. Implications (Clinical and Research)
The proactive ID + DHMI approach to treating depression among ILLIC is appealing at a population level because it addresses many of the barriers to screening for depression among cancer survivors seen with current standard of care approaches [42, 43, 44, 45]. Although the strong recruitment, retention, and engagement data confirm the feasibility and acceptability of the proactive ID + DMHI approach for this patient population, our data also highlight opportunities to improve intervention reach in two key areas. First, the overall response rate to the proactive outreach was low, as only 25% of ILLIC with likely elevated depressive symptoms completed the screening eligibility assessment. Second, even among those who completed the assessment who likely had elevated rates of depression per EHR documentation, an additional 37.5% did not have elevated depressive symptoms upon confirmatory eligibility assessment. These data suggest that additional research is necessary to tailor message content and/or method of delivery to promote reach while concurrently refining methods to enhance the accuracy of identifying individuals with likely elevated depressive symptoms using structured and unstructured data within the EHR.
Our study also extends normative data about mobile device proficiency among cancer survivors, a key feasibility concern for expanding DMHI usage. Our sample was highly proficient in terms of their mobile device use, with proficiency scores similar to a community‐based sample of people < 65 years from a study published in 2016 [38]. However, as overall population‐level norms or cancer survivor‐specific norms for the MDPQ‐16 have not been established, it is unclear whether our study sample with a median age of 57.5 years is exceptionally proficient or rather merely representative in terms of mobile device proficiency in the current age. It is also unknown whether mobile device proficiency systematically differed between those who enrolled in the study and those who did not. Future studies evaluating DMHI in this patient population should characterize mobile device proficiency among participants and those who decline to understand how to maximize intervention reach.
Our study adds to the growing evidence base supporting the efficacy of DMHIs for treating depression among cancer survivors [23, 24, 25, 26]. Prior studies evaluating DMHI for treating depression among cancer survivors have been small and single arm, establishing feasibility and acceptability. Building on these single arm data, Zion et al. conducted a decentralized RCT of 449 patients with nonmetastatic stage I‐III cancer survivors, demonstrating that a cognitive behavioral stress management (CSBM) app reduced anxiety and depression severity relative to sham educational control [25]. This important study provided a solid evidence base for DMHIs among cancer survivors but was limited through its inclusion of individuals without clinically significant depressive symptoms, exclusion of those with severe depression, and reliance on social media and internet platforms for recruitment (i.e., untethered to clinical oncology care).
To our knowledge, our trial is the first to show the potential effectiveness of DMHIs for treating depression specifically in the growing survivor population of ILLIC. In their large RCT, Zion et al. excluded individuals with stage IV and metastatic cancer because of the need for additional tailoring of their CSBM‐based DMHI to meet the unique needs of this population [25]. In particular, relevant considerations for ILLIC include how longstanding treatment‐related toxicity (e.g., chemotherapy induced cognitive dysfunction, cancer‐related fatigue), need for ongoing cancer‐directed therapy (e.g., busy daily schedules with clinical appointments), and prognostic uncertainty (e.g., fear of cancer progression) may affect the delivery and efficacy of DMHIs for depression treatment. Herein, we provide novel data demonstrating that a BA‐based DMHI tailored to the unique needs of survivors living with advanced and metastatic cancer has potential as a novel treatment paradigm.
Although there is a growing interest in the ILLIC population, their long‐term toxicity profile and comorbidity burden remain poorly understood [3, 6]. Despite being on cancer‐directed therapy for a median of 2.5 years, our study sample had relatively good performance status (75% ECOG 0–1) and low severity of comorbidity burden (40% Charlson Comorbidity Index score = 0). However, these measures of performance status and comorbidity burden are likely not the optimal measures to characterize the relevant treatment toxicities experienced by this patient population in terms of their mental health care as they fail to capture relevant toxicities such as chemotherapy induced cognitive dysfunction and cancer‐related fatigue (Charlson Comorbidity Index) or do so in a manner that is too gross to notice small differences that may be relevant to seeking mental health care (ECOG performance status). Future studies that explore the psychosocial needs for the growing ILLIC population should continue to refine measures that precisely characterize the relevant toxicity profile.
4.2. Limitations
Consistent with its pilot nature, the study had a single‐site design, small sample size and short follow‐up. We are therefore cautious to over‐interpret these encouraging but preliminary findings. Further evaluation of proactive ID + DMHI in a fully powered efficacy trial featuring a sample of survivors with diverse demographic characteristics and longer‐term follow‐up is necessary to confirm the efficacy signal and enhance the external validity of these preliminary findings. Although we showed a potential benefit of proactive ID + DMHI relative to proactive ID + UC, we cannot elucidate the marginal benefit of this approach relative to UC without proactive identification (as patients in both arms were identified as having elevated depressive symptoms via proactive ID) nor to a control that would account for nonspecific treatment effects of receiving a mobile app (i.e., an attention control). Although self‐guided DMHIs allow on‐demand access to treatment, prior studies in non‐cancer samples have demonstrated highly variable patient engagement [46, 47]. The relationship between levels of engagement with self‐guided DMHIs and improvement in depression severity is not known and should be studied in future work. Finally, we were unable to passively and systematically collect information about the types of activities that patients create, schedule, and complete, limiting our ability to understand the relationship between activity scheduling and completion with improvement in depression severity.
5. Conclusion
Data from this pilot RCT suggest that proactive ID + DMHI may result in a clinically meaningful improvement in depressive symptoms among ILLIC. These promising preliminary data support conducting a large effectiveness trial to establish proactive ID + DMHI as a novel, evidence‐based paradigm for treating depression among ILLIC.
Conflicts of Interest
E.M.G. reports research support and honorarium from Castle Biosciences; research support and travel from Stryker, and advisory and speaker bureau fees from Merck, all outside of the submitted work. J.D. is a co‐owner of Behavioral Activation Tech LLC, a small business that develops digital interventions for behavioral health treatment. Underlying intellectual property for the Moodivate intervention is owned by the University of Maryland, College Park. Behavioral Activation Tech LLC, has agreed to an exclusive license with the University of Maryland, College Park for the commercialization of Moodivate.
Supporting information
Supporting Information S1
Acknowledgments
This was supported by the National Cancer Institute at the National Institutes of Health (R01 CA281740 to E.M.G and J.D.), R42 MH108219 to J.D., the Biostatistics Shared Resource, Hollings Cancer Center, Medical University of South Carolina (P30 CA138313), the Hollings Cancer Center Clinical Concept Award (P30 CA138313), and the National Center for Advancing Translational Sciences (UL1 TR000062). The funding organizations had no influence on the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Graboyes, Evan M. , Levins Olivia, DeMass Reid, et al. 2025. “Proactive Identification and Digital Mental Health Intervention for the Treatment of Depression Among Individuals With Likely Incurable Cancer: A Pilot Randomized Clinical Trial,” Psycho‐Oncology: e70309. 10.1002/pon.70309.
Funding: This work was supported by the National Cancer Institute at the National Institutes of Health (R01 CA281740 to E.M.G and J.D.), R42 MH108219 to J.D., the Biostatistics Shared Resource, Hollings Cancer Center, Medical University of South Carolina (P30 CA138313), the Hollings Cancer Center Clinical Concept Award (P30 CA138313), and the National Center for Advancing Translational Sciences (UL1 TR000062).
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Stegmann M. E., Geerse O. P., van Zuylen L., Nekhlyudov L., and Brandenbarg D., “Improving Care for Patients Living With Prolonged Incurable Cancer,” Cancers 13, no. 11 (May 2021): 2555, 10.3390/cancers13112555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Gallicchio L., Devasia T. P., Tonorezos E., Mollica M. A., and Mariotto A., “Estimation of the Numbers of Individuals Living With Metastatic Cancer in the United States,” Journal of the National Cancer Institute 114, no. 11 (August 2022): 1476–1483, 10.1093/jnci/djac158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Mollica M. A., Smith A. W., Tonorezos E., et al., “Survivorship for Individuals Living With Advanced and Metastatic Cancers: National Cancer Institute Meeting Report,” Journal of the National Cancer Institute 114, no. 4 (April 2022): 489–495, 10.1093/jnci/djab223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Brandenbarg D., Maass S., Geerse O. P., et al., “A Systematic Review on the Prevalence of Symptoms of Depression, Anxiety and Distress in Long‐Term Cancer Survivors: Implications for Primary Care,” European Journal of Cancer Care 28, no. 3 (May 2019): e13086, 10.1111/ecc.13086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Akechi T., Okuyama T., Onishi J., Morita T., and Furukawa T. A., “Psychotherapy for Depression Among Incurable Cancer Patients,” Cochrane Database of Systematic Reviews 2008, no. 2 (April 2008): CD005537, 10.1002/14651858.CD005537.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Brick R. S., Gallicchio L., Mollica M. A., et al., “Survivorship Concerns Among Individuals Diagnosed With Metastatic Cancer: Findings From the Cancer Experience Registry,” Journal of Cancer Survivorship 19, no. 5 (April 2024): 1568–1579, 10.1007/s11764-024-01573-8. [DOI] [PubMed] [Google Scholar]
- 7. Mathes T., Pieper D., Antoine S. L., and Eikermann M., “Adherence Influencing Factors in Patients Taking Oral Anticancer Agents: A Systematic Review,” Cancer Epidemiology 38, no. 3 (June 2014): 214–226, 10.1016/j.canep.2014.03.012. [DOI] [PubMed] [Google Scholar]
- 8. Andersen B. L., Lacchetti C., Ashing K., et al., “Management of Anxiety and Depression in Adult Survivors of Cancer: ASCO Guideline Update,” Journal of Clinical Oncology 41, no. 18 (June 2023): 3426–3453, 10.1200/JCO.23.00293. [DOI] [PubMed] [Google Scholar]
- 9. Savard J., Simard S., Giguere I., et al., “Randomized Clinical Trial on Cognitive Therapy for Depression in Women With Metastatic Breast Cancer: Psychological and Immunological Effects,” Palliative & Supportive Care 4, no. 3 (September 2006): 219–237, 10.1017/s1478951506060305. [DOI] [PubMed] [Google Scholar]
- 10. Kissane D. W., Grabsch B., Clarke D. M., et al., “Supportive‐Expressive Group Therapy for Women With Metastatic Breast Cancer: Survival and Psychosocial Outcome From a Randomized Controlled Trial,” Psycho‐Oncology 16, no. 4 (April 2007): 277–286, 10.1002/pon.1185. [DOI] [PubMed] [Google Scholar]
- 11. Okuyama T., Akechi T., Mackenzie L., and Furukawa T. A., “Psychotherapy for Depression Among Advanced, Incurable Cancer Patients: A Systematic Review and Meta‐Analysis,” Cancer Treatment Reviews 56 (May 2017): 16–27, 10.1016/j.ctrv.2017.03.012. [DOI] [PubMed] [Google Scholar]
- 12. Eakin E. G. and Strycker L. A., “Awareness and Barriers to Use of Cancer Support and Information Resources by HMO Patients With Breast, Prostate, or Colon Cancer: Patient and Provider Perspectives,” Psycho‐Oncology 10, no. 2 (March‐April 2001): 103–113, 10.1002/pon.500. [DOI] [PubMed] [Google Scholar]
- 13. Kumar P., Casarett D., Corcoran A., et al., “Utilization of Supportive and Palliative Care Services Among Oncology Outpatients at One Academic Cancer Center: Determinants of Use and Barriers to Access,” Journal of Palliative Medicine 15, no. 8 (August 2012): 923–930, 10.1089/jpm.2011.0217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Dilworth S., Higgins I., Parker V., Kelly B., and Turner J., “Patient and Health Professional's Perceived Barriers to the Delivery of Psychosocial Care to Adults wIth Cancer: A Systematic Review,” Psycho‐Oncology 23, no. 6 (June 2014): 601–612, 10.1002/pon.3474. [DOI] [PubMed] [Google Scholar]
- 15. Fu S. S., van Ryn M., Sherman S. E., et al., “Proactive Tobacco Treatment and Population‐Level Cessation: A Pragmatic Randomized Clinical Trial,” JAMA Internal Medicine 174, no. 5 (May 2014): 671–677, 10.1001/jamainternmed.2014.177. [DOI] [PubMed] [Google Scholar]
- 16. Dahne J., Wahlquist A. E., Carpenter M. J., et al., “A Digital Depression Treatment Program for Adults Treated in Primary Care: A Randomized Clinical Trial,” JAMA Internal Medicine 185, no. 6 (April 2025): 692, 10.1001/jamainternmed.2025.0494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Maurer D. M., “Screening for Depression,” American Family Physician 85, no. 2 (January 2012): 139–144. https://pubmed.ncbi.nlm.nih.gov/30277728/. [PubMed] [Google Scholar]
- 18. Grassi L., Caruso R., Sabato S., Massarenti S., and Nanni M. G., and The UniFe Psychiatry Working Group C , “Psychosocial Screening and Assessment in Oncology and Palliative Care Settings,” Frontiers in Psychology 5 (2014): 1485, 10.3389/fpsyg.2014.01485. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Grassi L., Spiegel D., and Riba M., “Advancing Psychosocial Care in Cancer Patients,” F1000Research 6 (2017): 2083, 10.12688/f1000research.11902.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kirshner J., Cohn K., Dunder S., et al., “Automated Electronic Health Record‐based Tool for Identification of Patients With Metastatic Disease to Facilitate Clinical Trial Patient Ascertainment,” JCO Clinical Cancer Informatics 5 (June 2021): 719–727, 10.1200/CCI.20.00180. [DOI] [PubMed] [Google Scholar]
- 21. Furukawa T. A., Suganuma A., Ostinelli E. G., et al., “Dismantling, Optimising, and Personalising Internet Cognitive Behavioural Therapy for Depression: A Systematic Review and Component Network Meta‐Analysis Using Individual Participant Data,” Lancet Psychiatry 8, no. 6 (June 2021): 500–511, 10.1016/S2215-0366(21)00077-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Bae H., Shin H., Ji H. G., Kwon J. S., Kim H., and Hur J. W., “App‐Based Interventions for Moderate to Severe Depression: A Systematic Review and Meta‐Analysis,” JAMA Network Open 6, no. 11 (November 2023): e2344120, 10.1001/jamanetworkopen.2023.44120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Skrabal Ross X., Gunn K. M., Olver I., et al., “Online Psychosocial Interventions for Posttreatment Cancer Survivors: An International Evidence Review and Update,” Current Opinion in Supportive and Palliative Care 14, no. 1 (March 2020): 40–50, 10.1097/SPC.0000000000000478. [DOI] [PubMed] [Google Scholar]
- 24. Chow P. I., Showalter S. L., Gerber M., et al., “Use of Mental Health Apps by Patients With Breast Cancer in the United States: Pilot Pre‐Post Study,” JMIR Cancer 6, no. 1 (April 2020): e16476, 10.2196/16476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zion S. R., Taub C. J., Heathcote L. C., et al., “Effects of a Cognitive Behavioral Digital Therapeutic on Anxiety and Depression Symptoms in Patients With Cancer: A Randomized Controlled Trial,” JCO Oncology Practice 19, no. 12 (December 2023): 1179–1189, 10.1200/OP.23.00210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lee K., Kim S., Kim S. H., et al., “Digital Health Interventions for Adult Patients With Cancer Evaluated in Randomized Controlled Trials: Scoping Review,” Journal of Medical Internet Research 25 (January 2023): e38333, 10.2196/38333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Fernandez‐Rodriguez C., Gonzalez‐Fernandez S., Coto‐Lesmes R., and Pedrosa I., “Behavioral Activation and Acceptance and Commitment Therapy in the Treatment of Anxiety and Depression in Cancer Survivors: A Randomized Clinical Trial,” Behavior Modification 45, no. 5 (September 2021): 822–859, 10.1177/0145445520916441. [DOI] [PubMed] [Google Scholar]
- 28. Hopko D. R., Armento M. E., Robertson S. M., et al., “Brief Behavioral Activation and Problem‐Solving Therapy for Depressed Breast Cancer Patients: Randomized Trial,” Journal of Consulting and Clinical Psychology 79, no. 6 (December 2011): 834–849, 10.1037/a0025450. [DOI] [PubMed] [Google Scholar]
- 29. Eldridge S. M., Chan C. L., Campbell M. J., et al., “CONSORT 2010 Statement: Extension to Randomised Pilot and Feasibility Trials,” BMJ 355 (October 2016): i5239, 10.1136/bmj.i5239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Montgomery P., Grant S., Mayo‐Wilson E., et al., “Reporting Randomised Trials of Social and Psychological Interventions: The CONSORT‐SPI 2018 Extension,” Trials 19, no. 1 (July 2018): 407, 10.1186/s13063-018-2733-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Eysenbach G. and Group C.‐E., “CONSORT‐EHEALTH: Improving and Standardizing Evaluation Reports of Web‐Based and Mobile Health Interventions,” Journal of Medical Internet Research 13, no. 4 (December 2011): e126, 10.2196/jmir.1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Kroenke K. and Spitzer R. L., “The PHQ‐9: A New Depression Diagnostic and Severity Measure,” Psychiatric Annals 32, no. 9 (2002): 509–515, 10.3928/0048-5713-20020901-06. [DOI] [Google Scholar]
- 33. Kroenke K., Spitzer R. L., and Williams J. B., “The Patient Health Questionnaire‐2: Validity of a Two‐Item Depression Screener,” Medical Care 41, no. 11 (November 2003): 1284–1292, 10.1097/01.MLR.0000093487.78664.3C. [DOI] [PubMed] [Google Scholar]
- 34. Dahne J., Lejuez C. W., Kustanowitz J., et al., “Moodivate: A Self‐Help Behavioral Activation Mobile App for Utilization in Primary Care‐Development and Clinical Considerations,” International Journal of Psychiatry in Medicine 52, no. 2 (March 2017): 160–175, 10.1177/0091217417720899. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Dahne J., Lejuez C., Diaz V. A., et al., “Pilot Randomized Trial of a Self‐Help Behavioral Activation Mobile Application for Utilization in Primary Care,” Behavior Therapy 50, no. 4 (2019): 817–827, 10.1016/j.beth.2018.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Mazzucchelli T., Kane R., and Rees C., “Behavioral Activation Treatments for Depression in Adults: A Meta‐Analysis and Review,” Clinical Psychology: Science and Practice 16, no. 4 (2009): 383–411, 10.1111/j.1468-2850.2009.01178.x. [DOI] [Google Scholar]
- 37. Lejuez C. W., Hopko D. R., and Hopko S. D., “A Brief Behavioral Activation Treatment for Depression. Treatment Manual,” Behavior Modification 25, no. 2 (April 2001): 255–286, 10.1177/0145445501252005. [DOI] [PubMed] [Google Scholar]
- 38. Roque N. A. and Boot W. R., “A New Tool for Assessing Mobile Device Proficiency in Older Adults: The Mobile Device Proficiency Questionnaire,” Journal of Applied Gerontology 37, no. 2 (February 2018): 131–156, 10.1177/0733464816642582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Brooke J., “SUS‐A Quick and Dirty Usability Scale,” in Usability evaluation in industry 189, no. 194 (1996): 4–7. [Google Scholar]
- 40. Mystakidou K., Tsilika E., Parpa E., Katsouda E., Galanos A., and Vlahos L., “Assessment of Anxiety and Depression in Advanced Cancer Patients and Their Relationship With Quality of Life,” Quality of Life Research 14, no. 8 (2005): 1825–1833, 10.1007/s11136-005-4324-3. [DOI] [PubMed] [Google Scholar]
- 41. Mah K., Swami N., Le L. W., et al., “Validation of the 7‐Item Functional Assessment of Cancer Therapy‐General (FACT‐G7) as a Short Measure of Quality of Life in Patients With Advanced Cancer,” Cancer 126, no. 16 (August 2020): 3750–3757, 10.1002/cncr.32981. [DOI] [PubMed] [Google Scholar]
- 42. Gerard B., Robbins M., Putra J., et al., “Applying Lean Six Sigma to Improve Depression Screening and Follow‐Up in Oncology Clinics,” Journal for Healthcare Quality 43, no. 3 (May‐June 2021): 153–162, 10.1097/JHQ.0000000000000294. [DOI] [PubMed] [Google Scholar]
- 43. Lazaro‐Escudero M. I., Burgos‐Cardona C. A., Acevedo‐Fernandez K., and Castro‐Figueroa E. M., “Technology‐Assisted Depression Screening Tools for Patients With Cancer: A Systematic Review Protocol,” BMJ Open 11, no. 3 (March 2021): e041878, 10.1136/bmjopen-2020-041878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Dauchy S., Dolbeault S., and Reich M., “Depression in Cancer Patients,” EJC Supplements 11, no. 2 (September 2013): 205–215, 10.1016/j.ejcsup.2013.07.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Hahn E. E., Munoz‐Plaza C. E., Pounds D., et al., “Effect of a Community‐Based Medical Oncology Depression Screening Program on Behavioral Health Referrals Among Patients With Breast Cancer: A Randomized Clinical Trial,” JAMA 327, no. 1 (January 2022): 41–49, 10.1001/jama.2021.22596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Pham Q., Graham G., Carrion C., et al., “A Library of Analytic Indicators to Evaluate Effective Engagement With Consumer mHealth Apps for Chronic Conditions: Scoping Review,” JMIR mHealth and uHealth 7, no. 1 (January 2019): e11941, 10.2196/11941. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Lipschitz J. M., Van Boxtel R., Torous J., et al., “Digital Mental Health Interventions for Depression: Scoping Review of User Engagement,” Journal of Medical Internet Research 24, no. 10 (October 2022): e39204, 10.2196/39204. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information S1
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.