A pilot intervention of using a mobile health app (ONC Roadmap) to enhance health-related quality of life in family caregivers of pediatric patients with cancer

Introduction

Over the past decade, our interdisciplinary team developed a positive psychology-based mHealth app (Roadmap) (1). We created this app, because when a child is diagnosed with cancer, the entire family is impacted (2,3). Invariably, cancer alters the health-related quality of life (HRQOL) of family members and their care recipients (patients) (4). Family caregivers face a myriad of challenges navigating the demands of paid jobs with unpaid caregiving tasks (5). Unsurprisingly, caregivers who assume significant caregiving responsibilities at home face higher physical and emotional stressors, impeding their ability to provide care of loved ones, make decisions, and manage self-care (6-9). These chronic stressors can lead to adverse psychological and physiological effects on their daily lives that can adversely impact the patient (10,11).

The Roadmap mHealth app was developed to provide HRQOL support for family caregivers of patients with cancer (12,13). Applying Carbonneau’s conceptual framework on the positive aspects of caregiving (14), iterative cycles of user-centered design were utilized (15). This framework included three central positive aspects of caregiving: (I) quality of caregiver and patient daily relationship; ii) caregiver feeling of accomplishment; and (III) meaning of the caregiver role in daily life. These domains were considered interdependent and working together to reinforce positive outcomes (e.g., caregiver well-being) and protect caregiver HRQOL. Caregiver self-efficacy and caregiver enrichment events in daily life influenced the enhancement of positive aspects of caregiving (14).

Guided by this framework (14), the Roadmap mHealth app was studied in a pilot intervention to support the HRQOL of family caregivers of pediatric patients with cancer. Herein, this Roadmap app was leveraged to: (I) assess the feasibility and acceptability of the Roadmap mHealth app (henceforth, ONC Roadmap, abbreviated for “oncology”) in caregivers of pediatric patients with cancer; (II) characterize self-reported outcomes of physical, mental, and social HRQOL domains; and (III) evaluate the wearable sensor data outputs in both caregivers and patients. This work is important because it may inform future mHealth design and intervention considerations for families of children with cancer. We present the following article in accordance with the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) reporting checklist (16) (available at https://mhealth.amegroups.com/article/view/10.21037/mhealth-22-24/rc).

Methods

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Institutional Review Board of the University of Michigan Medical School (IRBMED HUM#01176584) and was registered on ClinicalTrials.gov (NCT04480541). IRBMED-approved informed consent/assent was taken from all the study participants.

Study site

The study was conducted at the University of Michigan, Ann Arbor, MI (U-M). All study activities were conducted remotely with no in-person contact, and all study materials were mailed to participants’ residences. The design and development of ONC Roadmap have been previously published, including graphical images of the app (https//:www.roadmap.study) (1).

Recruitment and enrollment

Inclusion Criteria: Eligibility for study participation of caregivers included: age ≥18 years and self-reported as the primary caregiver of their pediatric patient with cancer. Patients were required to be at least age ≥5 years and receiving cancer care at the data coordinating site. The study team has other IRB-approved studies in this patient population (age ≥5 years) where technology and wearable sensors are being examined. Patients in this age group have been participants in similar studies (17). While the intervention specifically targeted the caregiver, both members of the dyad (i.e., caregiver and patient) had to agree to participate. Additional eligibility requirements included both members having access to necessary resources for participating in an mHealth technology-based intervention (i.e., smartphone/tablet and internet access), and being willing to use personal equipment/internet for the study. All participants provided informed consents/assents within the ONC Roadmap app. Of note, children (age 10–14 years) signed the IRB-approved Assent Form document and adolescents (age 15–17 years) signed the IRB-approved Consent Form document; children (age 5–9 years) did not sign any Assent Form documents.

Recruitment occurred between September 2020–September 2021. IRBMED-approved paper flyers and postings were distributed throughout the outpatient Pediatric Hematology/Oncology (PHO) waiting rooms, clinic rooms, and infusion center. Interested participants who contacted the study team by phone or email received additional study information (e.g., overview of study procedures). All recruitment and participant onboarding were conducted remotely due to the COVID-19 pandemic. The target sample was 50 dyads (see Figure 1: TREND (16) Diagram of Participant Flow).

Figure 1 Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) diagram of participant flow.

The intervention: study procedures

The study procedures of the intervention are outlined in Figure S1.

Self-reported assessments

All self-reported HRQOL data were collected using ONC Roadmap, which utilized Qualtrics (Qualtrics, Provo, UT), an online research tool that enables researchers to create study-specific websites for administering study surveys and storing participant data. Participants were prompted by app alert to complete assessments at baseline (pre-study period (T0) and days 30 and 120 (post-baseline assessments at T1 and T2) using ONC Roadmap. Caregiver socio-demographic data (e.g., age, gender, race, ethnicity, education, occupation), household finances, caregiving experiences, and use of mobile devices/technology were obtained at T0, based on our National Caregiver Health Survey (18-20). A list of the HRQOL PROMIS^® measures assessed at T0, T1, and T2 by self-report or parent proxy are provided in Table S1 (21,22). The reference population for PROMIS^® measures is the U.S. general population (23), whereas the reference population of the affiliate PROMIS^® measures (the Neuro-QoL TBI-CareQoL measures) is other caregivers (24). PROMIS^® Measure-Specific Scoring Guides available online through the PROMIS^® Assessment Center were used to score the measures and calculate T-scores (25). A higher PROMIS^® T-score represents more of the concept being measured. For example, an individual with T-score of 60 for the Global Mental or Physical Health scale is one standard deviation better (i.e., healthier) than the U.S. general population. While patients (age ≥8 years) and family caregivers (age ≥18 years) completed self-assessments, parent proxy assessments were also completed by family caregivers for patients (age 5–17 years only).

Roadmap and Fitbit^® Apps

Participants (caregivers and patients) were instructed to download ONC Roadmap and Fitbit^® apps on their smartphone or other mobile device (both free of charge and publicly available via Apple and Google app stores). As previously described (1), caregivers received the full-version of ONC Roadmap, which included positive activities, chat forums, resources, and graphs (mood, sleep, and steps). Graphs only were visible to patients (i.e., they did not have access to positive activities, chat forums, or resources).

Wearable sensor

Fitbits^® were mailed to the participants’ homes. They were instructed to wear it continuously, except while charging, to measure their physical activity, heart rate, and sleep during the 120-day monitoring period.

Feasibility and acceptability

Feasibility of the study, defined a priori in the Protocol, was calculated as the percentage of caregivers who logged into ONC Roadmap and engaged with it at least twice weekly for at least 50% of the 120-day study duration. These data were based on data use logs (i.e., timestamps) of: (I) positive activity completed; (II) chat/reply to chat posted or viewed in the forum, (III) and mood score reported (1).

Caregivers completed a Feasibility and Acceptability questionnaire, which was informed by existing measures of feasibility and acceptability (26,27), and the Mobile App Rating Scale (MARS) (28) at the end of study (i.e., day 120) to specifically assess app-quality. Our a priori hypothesis was that more than 50% of respondents would Agree or Strongly Agree with the feasibility and acceptability of ONC Roadmap. The MARS is a simple, objective, and reliable tool for assessing the quality of mHealth apps and has demonstrated internal consistency (alpha =0.90) and interrater reliability intraclass correlation coefficient (ICC =0.79) (28). The MARS was scored by calculating the mean scores of engagement, functionality, aesthetics, and information quality objective subscales, and an overall mean app quality total score. Each MARS item used a 5-point scale (1-Inadequate, 2-Poor, 3-Acceptable, 4-Good, 5-Excellent). Higher total and subscale scores indicate better app-quality.

Statistical analyses

For the descriptive statistics, continuous measures were described using means/medians (M) and standard deviation (SD)/interquartile range (IQR), while categorical measures were summarized using frequencies and proportions. These data were analyzed using R (version 4.1.1).

Fitbit^® automatically generated accelerometer-based summary data (per proprietary algorithms) based on “activity counts” collected over the course of the day. We assessed participant compliance in wearing the Fitbit^® by identifying when heart rate data were present through the Roadmap platform using the Fitbit application programming interface (API) (29). As previously reported, we measured daily wear time using heart rate data with a minutes-level resolution. Compliance was expressed both in hours (0–24 h) and in percentages (i.e., by dividing the hours spent wearing the device by 24 h) (30,31). Using this assessment of compliance, we calculated the average daily step count for participants who wore the Fitbit^® more than 6 h between 8 AM and 8 PM. We chose a cut-off of 6 h because the distribution of average daily step count did not change significantly for higher cut-offs. No compliance cut-off was applied for the calculation of asleep hours because the daily average changed by only about 0.05 h between a cut-off of 0 h and a cut-off of 11 h between 8 PM and 8 AM.

Although this pilot study was not powered to examine efficacy, exploratory analyses were conducted to assess for changes in HRQOL scores across time (i.e., T0, T1, T2). Baseline vs. day 30 and baseline vs. day 120 HRQOL mean T-scores with SD were compared using two-tailed T tests with probability level of 0.05.

Next, we used a longitudinal regression model with random effects to determine whether caregiver PROMIS^® global health outcome changed over time. The model included age, gender, self-report of any mental health condition, and caregiving hours/week.

To examine the relationship between caregiver and patient, we treated data from care partners as a paired or dyadic longitudinal series where the pairing was modeled at each timepoint in the series. Because longitudinal, dyadic data present a special case of nested data whereby interdependence exists at two hierarchies in the data, we employed the Actor-Partner Interdependence Model (APIM) (32-35). An important feature of the APIM analyses was to create within- and between-member versions of the outcome variable (e.g., PROMIS^® global mental health), separately for caregivers and patients. The APIM also included age, gender, baseline self-report of any mental health condition, and caregiving hours/week. We specifically examined whether baseline (T0) caregiver and patient HRQOL (anxiety, depression) domains influenced caregiver and patient global (mental) health at 120-day (T2).

Results

Socio-demographic characteristics

One hundred participants consented/assented and enrolled in this study (Figure 1) with 50 family caregivers and 50 pediatric patients with cancer. There was low study attrition (<5%; Table 1). Nearly half of the caregivers were unable to work (N=21; 42%) due to caregiving responsibilities or unemployed (N=2; 4%). Twenty-nine caregivers (58%) reported annual family income ≤$99,999 and three (6%) <$10,000 for a mean number of 4.5 persons in the household (range, 2–8). Only one caregiver did not own a smartphone and opted to use a mobile tablet for study participation.

Feasibility and acceptability

The majority of caregivers (N=32; 65%) logged into ONC Roadmap and engaged with it at least twice weekly for ~50% of the study duration. Eighty percent of caregivers (N=39) logged in at least once weekly. The four most common activities used were gratitude journal, pleasant activity scheduling, savoring, and engaging with beauty (Table 2). Not surprisingly, caregiver app use declined over time from 18-days during the first 30-days compared with 12-days during the last 30-days (Figure 2). Nonetheless, the Feasibility and Acceptability questionnaire responses indicated that the Fitbit^®, ONC Roadmap app, and longitudinal self-reported assessments were feasible and acceptable with the majority reporting Agree or Strongly Agree with positive Net Favorability in all the categories (Table S2).

Figure 2 ONC Roadmap App use over time in family caregivers. Each boxplot represents the daily compliance averaged chronologically for each 30-day of the 120-day study period (N=49). M1 vs. M2, P=0.002; M1 vs. M3, P<0.001; M1 vs. M4, P<0.001. M, month; IQR, interquartile range.

The MARS was also utilized to provide a multidimensional measure of ONC Roadmap app quality indicators of engagement, functionality, aesthetics, and information quality (28). The ONC Roadmap app quality total mean score was 3.59 (SD =0.78) and overall star rating was 3.38 (SD =0.86). The sub-scales, functionality and aesthetic, had the highest reported mean subscale scores of 4.01 (SD =0.66) and 3.88 (SD =0.74), respectively (Table 3), followed by information and engagement [mean subscale scores of 3.76 (SD =0.75) and 2.99 (SD =0.87), respectively].

Using ONC Roadmap to obtain HRQOL data

Completion rates for the HRQOL assessments by caregivers at T0, T1, and T2 were 100% (N=50/50), 88% (N=43/49), and 88% (N=38/49), respectively. Completion rates by patients (age 8 years and older) were also 100% (N=34/34) at T0 but were lower than caregivers at T1 and T2 [61% (N=20/33), and 45% (N=15/33), respectively]. The Parent Proxy (age 5–17 years) assessments were completed at similar rates to the caregiver self-assessments [100% (N=41/41), 85% (N=35/41), and 76% (N=31/41)], respectively.

In caregivers, the median wear time of sensors across the 120-day study period was: 17.8 h (of the 24 h day); 9.2 h during daytime (between 8 AM–8 PM), and 8.4 h during nighttime (between 8 PM–8 AM). In patients, the median wear time of sensors was: 6.3 h (of the 24 h day); 4.1 h during daytime (between 8 AM–8 PM), and 3.0 h during nighttime (between 8 PM–8 AM). Figure S2 shows the distribution of compliance (24 h) of Fitbit^® wear across the study period. A decline in caregiver compliance was observed from a median of 19.3 h (first 30 days) to 15.8 h (last 30 days) of the study period (Figure S2A), while patient compliance declined from a median of 11.0 h (first 30 days) to 1.3 h (last 30 days, Figure S2B).

When we explored potential differences of steps, sleep, and self-reported mood, caregiver and patient mood were higher in the 5–11 years compared with 12–17 or 18+ years age-groups, P=0.003 or P=0.022 (caregiver) and P<0.001 or P<0.001 (patient), respectively (Table S3). Patient step count was also higher in those aged 5–11 compared with those aged 12–17 (P=0.024). There were no differences in sleep among caregiver and patient age-groups.

Exploratory analyses: Roadmap’s preliminary efficacy on HRQOL outcomes

The change in mean pre- (T0) and post- (T1 and T2) HRQOL scores for participants are shown in Table S4. In general, although most analyses failed to meet conventional levels of statistical significance, there were improvements in almost all the different mental HRQOL domains across all of the groups over time. Specifically, at 30-day (T1), caregivers had higher levels of global mental health and lower levels of caregiver-specific anxiety. At 120-day (T2), caregiver-specific anxiety and strain and general anxiety were lower compared with baseline. Interestingly, caregiver-specific anxiety at T0 was negatively correlated with app use at T1 (i.e., higher baseline anxiety scores were associated with less app use over the next 30 days; Table S5).

In patients (8–17 years), there were lower levels of depression at T1, without reaching statistical significance at T2; patients 18+ years reported better global mental health at T1, without reaching statistical significance at T2. However, in parent proxy reports (8–17 years), depression and fatigue were both rated lower at T2. No significant changes were observed by parent proxy in patients 5–7 years; however, the means were in the anticipated direction.

To adjust for participant age, gender, number of caregiving hours per week, and baseline self-report of any caregiver mental health condition, we then performed linear mixed models. While caregiver age >40 years and self-report of any mental health condition were negatively associated with global mental health of the caregiver, this outcome was improved at T1 compared with T0 (Table S6). Additionally, patient (8–17 years) depression was significantly lower over at T1 and T2 compared with T0. In our generalized linear APIM models, we found that “actor” (i.e., caregiver or patient) anxiety or depression at baseline influenced day 120 global mental health outcomes of the caregiver or patient, respectively. Interestingly, when we also assessed for “partner” effects (Table S7), caregiver depression at T0 was negatively associated with patient global mental health at T2.

Discussion

In this study, family caregivers of pediatric patients with cancer met our a priori defined measure for Feasibility for the study duration (120-day). Most family caregivers agreed that the ONC Roadmap app, Fitbit^®, and study design were feasible and acceptable. They also indicated they were likely to engage in a similar future study lasting up to 6-month.

Our study also incorporated the MARS, which provided a multidimensional measure of engagement, functionality, aesthetics, and information quality (28). ONC Roadmap caregivers reported the sub-scale mean scores as well as overall mean app quality total score at least consistent with or higher than 50 apps that were previously selected for rating by the MARS (28). Caregivers generally Agreed/Strongly Agreed with the perceived impact of ONC Roadmap on users’ knowledge and attitudes. Specific strengths of ONC Roadmap were functionality, aesthetic, and information quality. While overall a specific weakness was engagement, specific elements of entertainment and interest had mean scores higher than previously published studies on independent ratings of 50 mental health and well-being apps (20). This likely contributed to the decline in app use over the 120-day study period. Thus, we are considering strategies in future app refinements to enhance engagement by presenting content in interesting ways (e.g., alerts, messages, reminders, feedback) (36,37).

Nonetheless, caregivers were compliant with completing assessments, reporting mood scores, and wearing Fitbit^®, consistent with what we observed in our recently completed college student study (30). However, we found patients to be more variable in completing the study-related procedures. With the near ubiquitous use of technology in children, adolescents, and young adults (38), this was somewhat surprising. It is possible that caregivers experienced the benefit of having access to positive activities, chat forums, mood/steps/sleep graphs, and caregiving resources, whereas for patients, their only access to ONC Roadmap was limited to graphs. Alternatively, patients may have established their technology-related support and did not leverage ONC Roadmap to the extent that caregivers did. Semi-structured interviews with our dyads are ongoing to better understand these factors.

Not surprisingly, over one-third of the caregivers reported significant burden (i.e., providing more than 40 hours of unpaid caregiving) and up to one-half were unable to work due to caregiving responsibilities or unemployment. We also found that between 32–50% of our caregivers reported household incomes as less than or equal to 200% federal poverty level ($55,500 for a family of 4 in 2022) (39). While these data were in line with reported rates of financial hardship in Michigan and U.S. families with children below 200% (~45%) (40), they highlight the potential implications of poverty, such as material hardship, when designing interventions to address the needs of families undergoing intensive medical management, such as cancer care. Indeed, psychosocial support that is low-burden and dyadic-focused are needed to integrate seamlessly with cancer care delivery (4,41).

Research across multiple disciplines is emerging on the importance of designing interventions that are truly dyadic in nature, integrating both caregiver and patient (42). Dyadic-level processes have been shown to influence the health and well-being of members within a dyad (43). For example, pediatric and adolescent and young adult patients’ reports of subjective illness severity may indicate their own as well as their caregiver’s mental health (44). Accordingly, while exploratory, herein we examined the influence of baseline HRQOL domains of the caregiver and patient on their mental health and found that baseline depression in caregiver influenced patient’s global health 120-day later. Thus, with growing emphasis on interpersonal, dyadic-level processes likely contributing to the health, illness, recovery, treatment, and/or overall well-being of both members of the dyad (41,43), future studies should integrate both members of the dyad. Indeed, mHealth technologies offer scalable and flexible solutions for delivering family-based or dyadic-level interventions (45,46).

While this study was not powered to assess the efficacy of the ONC Roadmap in HRQOL outcomes, exploratory analyses suggest preliminary efficacy on HRQOL outcomes. In general, we saw improvements in mental HRQOL over the 120-day study period for all groups, although this difference only met conventional levels of significance in a few instances, primarily for the caregiver group (where the intervention was intended for). Specifically, there were significant improvements in caregiver-specific strain, caregiver-specific anxiety, and general anxiety at day 120. Interestingly, higher baseline care-specific anxiety scores were associated with less app use; and higher baseline patient (8–17 years) depression scores were associated with less caregiver app use (data not shown). There was also evidence to suggest that patient fatigue improved with caregivers receiving the intervention. Accordingly, while the intervention’s primary target was caregivers, it was unexpected that patients experienced reduced depression. Indeed, the “partner effects” observed in our APIM analyses suggest a potential interaction between both members of the dyad. Interestingly, while caregiver global mental health improved at day 30, this was not sustained at day 120, and physical function and positive affect declined at day 120.

It is possible that caregivers experienced increased physical burden while caring for other family members at home during the pandemic. We also speculate that with the PROMIS^® positive affect’s measure of “In the past 7 days: I felt cheerful,” it may better reflect pleasurable engagement (e.g., ecstatic happiness) (47), separate from general well-being, as assessed by the PROMIS^® global mental health (21). Lyubomirsky and Layous’ positive activity model (48) suggests that the dosage and variety of positive activities coupled with motivation and effort of the individual (i.e., so called, person-activity fit) may influence the degree to which well-being is enhanced. Thus, this intersection between characteristics of the individual and the positive activities may be important considerations in how well those activities are able to enhance individual well-being. In the present research, we did not obtain information regarding psychological conditions/disorders or personality traits. These variables could have impacted the findings and will be considered in the design of our future studies. In our ongoing qualitative interviews, we are assessing these considerations in more detail.

Strengths of this study included the broad inclusion criteria of pediatric cancer diagnoses and phases of care, low-burden on participants, and a pre-registered study design plan. Our study also had limitations. Our findings are likely more generalizable to caregivers who were similar, mostly White, non-Hispanic, with at least some college education, and who own mobile devices and routinely use apps. In the present study, patient compliance waned over time. The patient was not emphasized as the primary member of the study. It is possible their involvement was no longer considered novel or important over time. In future studies, we hope to incorporate the dyad as its primary target rather than one member of the dyad alone (e.g., caregiver). Additionally, we recognize the inherent biases afforded by single-arm, single-center study designs. Nonetheless, we are encouraged with the high proportion of caregivers who reported the intervention to be feasible and acceptable with no adverse events reported. While caregivers were compliant with completing assessments, reporting mood scores, and wearing Fitbit^®, patients were more variable in completing the study-related procedures.

Conclusions

Our findings suggest that mHealth technology can be used to support the HRQOL of caregivers and their patients. We are currently exploring mechanisms of ONC Roadmap on HRQOL outcomes. We are also examining strategies to enhance engagement, such as just-in-time adaptive interventions (36) or digital coaching (49). Considering the growing populations of survivors and aging caregivers (50), developing and rigorously testing mHealth platforms that provide cancer supportive care for both members of the dyad are needed.

Acknowledgments

Funding: This work was supported by an American Society of Hematology Bridge Grant and National Institute of Health/National Heart, Lung, and Blood Institute grant (No. 1R01HL146354) and the Edith S. Briskin and Shirley K Schlafer Foundation (SWC).

Footnote

Reporting Checklist: The authors have completed the TREND reporting checklist. Available at https://mhealth.amegroups.com/article/view/10.21037/mhealth-22-24/rc

Data Sharing Statement: Available at https://mhealth.amegroups.com/article/view/10.21037/mhealth-22-24/dss

Peer Review File: Available at https://mhealth.amegroups.com/article/view/10.21037/mhealth-22-24/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://mhealth.amegroups.com/article/view/10.21037/mhealth-22-24/coif). CNN is supported by an NIH/NCI T32 Training Grant (T32CA236621). JPT reports that the analytic work for this study was supported by the National Center for Advancing Translational Sciences (NCATS Grant Number: UL1TR002240) for the Michigan Institute for Clinical and Health Research. JAM is supported as a research staff member on HL146354 (Roadmap mHealth Study). AH, NEC, DLB, DAH report that being Co-I of National Institute of Health/National Heart, Lung, and Blood Institute grant (1R01HL146354). SWC reports that this work was supported by an American Society of Hematology Bridge Grant and National Institute of Health/National Heart, Lung, and Blood Institute grant (1R01HL146354) and the Edith S. Briskin and Shirley K Schlafer Foundation (SWC). SWC is supported by grants 1R01HL146354 and K24HL156896. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Institutional Review Board of the University of Michigan Medical School (IRBMED HUM# 01176584) and IRBMED-approved informed consent/assent was taken from all the study participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Rozwadowski M, Dittakavi M, Mazzoli A, et al. Promoting Health and Well-Being Through Mobile Health Technology (Roadmap 2.0) in Family Caregivers and Patients Undergoing Hematopoietic Stem Cell Transplantation: Protocol for the Development of a Mobile Randomized Controlled Trial. JMIR Res Protoc 2020;9:e19288. [Crossref] [PubMed]
2. Jones BL. The challenge of quality care for family caregivers in pediatric cancer care. Semin Oncol Nurs 2012;28:213-20. [Crossref] [PubMed]
3. Kazak AE, DiDonato S, Schneider SJ, et al. Assessing family psychosocial risks in pediatric cancer. In: Abrams AN, Muriel AC, Wiener L. editors. Pediatric Psychosocial Oncology: Textbook for Multidisciplinary Care. Cham: Springer International Publishing; 2016:51-65.
4. Kent EE, Rowland JH, Northouse L, et al. Caring for caregivers and patients: Research and clinical priorities for informal cancer caregiving. Cancer 2016;122:1987-95. [Crossref] [PubMed]
5. Northouse L, Williams AL, Given B, et al. Psychosocial care for family caregivers of patients with cancer. J Clin Oncol 2012;30:1227-34. [Crossref] [PubMed]
6. Schulz R, Newsom J, Mittelmark M, et al. Health effects of caregiving: the caregiver health effects study: an ancillary study of the Cardiovascular Health Study. Ann Behav Med 1997;19:110-6. [Crossref] [PubMed]
7. Vitaliano PP, Zhang J, Scanlan JM. Is caregiving hazardous to one's physical health? A meta-analysis. Psychol Bull 2003;129:946-72. [Crossref] [PubMed]
8. Lee S, Colditz GA, Berkman LF, et al. Caregiving and risk of coronary heart disease in U.S. women: a prospective study. Am J Prev Med 2003;24:113-9. [Crossref] [PubMed]
9. Pinquart M, Sörensen S. Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis. Psychol Aging 2003;18:250-67. [Crossref] [PubMed]
10. Northouse LL, Katapodi MC, Schafenacker AM, et al. The impact of caregiving on the psychological well-being of family caregivers and cancer patients. Semin Oncol Nurs 2012;28:236-45. [Crossref] [PubMed]
11. Litzelman K, Kent EE, Mollica M, et al. How Does Caregiver Well-Being Relate to Perceived Quality of Care in Patients With Cancer? Exploring Associations and Pathways. J Clin Oncol 2016;34:3554-61. [Crossref] [PubMed]
12. Kaziunas E, Hanauer DA, Ackerman MS, et al. Identifying unmet informational needs in the inpatient setting to increase patient and caregiver engagement in the context of pediatric hematopoietic stem cell transplantation. J Am Med Inform Assoc 2016;23:94-104. [Crossref] [PubMed]
13. Maher M, Kaziunas E, Ackerman M, et al. User-Centered Design Groups to Engage Patients and Caregivers with a Personalized Health Information Technology Tool. Biol Blood Marrow Transplant 2016;22:349-58. [Crossref] [PubMed]
14. Carbonneau H, Caron C, Desrosiers J. Development of a conceptual framework of positive aspects of caregiving in dementia. Dementia 2010;9:327-53. [Crossref]
15. Johnson CM, Johnson TR, Zhang J. A user-centered framework for redesigning health care interfaces. J Biomed Inform 2005;38:75-87. [Crossref] [PubMed]
16. Des Jarlais DC, Lyles C, Crepaz N, et al. Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: the TREND statement. Am J Public Health 2004;94:361-6. [Crossref] [PubMed]
17. Flora C, Tyler J, Mayer C, et al. High-frequency temperature monitoring for early detection of febrile adverse events in patients with cancer. Cancer Cell 2021;39:1167-8. [Crossref] [PubMed]
18. Kedroske J, Koblick S, Chaar D, et al. Development of a National Caregiver Health Survey for Hematopoietic Stem Cell Transplant: Qualitative Study of Cognitive Interviews and Verbal Probing. JMIR Form Res 2020;4:e17077. [Crossref] [PubMed]
19. Gupta V, Raj M, Hoodin F, et al. Electronic Health Record Portal Use by Family Caregivers of Patients Undergoing Hematopoietic Cell Transplantation: United States National Survey Study. JMIR Cancer 2021;7:e26509. [Crossref] [PubMed]
20. Raj M, Gupta V, Hoodin F, et al. Evaluating mobile Health technology use among cancer caregivers in the digital era. Digit Health 2022;8:20552076221109071. [Crossref] [PubMed]
21. Hays RD, Bjorner JB, Revicki DA, et al. Development of physical and mental health summary scores from the patient-reported outcomes measurement information system (PROMIS) global items. Qual Life Res 2009;18:873-80. [Crossref] [PubMed]
22. Forrest CB, Bevans KB, Pratiwadi R, et al. Development of the PROMIS ® pediatric global health (PGH-7) measure. Qual Life Res 2014;23:1221-31. [Crossref] [PubMed]
23. Jensen RE, Potosky AL, Moinpour CM, et al. United States Population-Based Estimates of Patient-Reported Outcomes Measurement Information System Symptom and Functional Status Reference Values for Individuals With Cancer. J Clin Oncol 2017;35:1913-20. [Crossref] [PubMed]
24. Carlozzi NE, Kallen MA, Hanks R, et al. The TBI-CareQOL Measurement System: Development and Preliminary Validation of Health-Related Quality of Life Measures for Caregivers of Civilians and Service Members/Veterans With Traumatic Brain Injury. Arch Phys Med Rehabil 2019;100:S1-S12. [Crossref] [PubMed]
25. Assessment Center [Internet]. [cited 2022 Jun 9]. Available online: https://www.assessmentcenter.net/
26. Vandelanotte C, De Bourdeaudhuij I. Acceptability and feasibility of a computer-tailored physical activity intervention using stages of change: project FAITH. Health Educ Res 2003;18:304-17. [Crossref] [PubMed]
27. Haerens L, Deforche B, Vandelanotte C, et al. Acceptability, feasibility and effectiveness of a computer-tailored physical activity intervention in adolescents. Patient Educ Couns 2007;66:303-10. [Crossref] [PubMed]
28. Stoyanov SR, Hides L, Kavanagh DJ, et al. Mobile app rating scale: a new tool for assessing the quality of health mobile apps. JMIR Mhealth Uhealth 2015;3:e27. [Crossref] [PubMed]
29. Fitbit API. [Accessed on 04/27/2022]. Available online: https://dev.fitbit.com/build/reference/web-api/
30. Gilley KN, Baroudi L, Yu M, et al. Risk Factors for COVID-19 in College Students Identified by Physical, Mental, and Social Health Reported During the Fall 2020 Semester: Observational Study Using the Roadmap App and Fitbit Wearable Sensors. JMIR Ment Health 2022;9:e34645. [Crossref] [PubMed]
31. Faust L, Purta R, Hachen D, et al. Exploring compliance: Observations from a large scale Fitbit study. Proceedings of the 2nd International Workshop on Social Sensing. 2017:55-60.
32. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986;51:1173-82. [Crossref] [PubMed]
33. Kenny DA, Ledermann T. Detecting, measuring, and testing dyadic patterns in the actor-partner interdependence model. J Fam Psychol 2010;24:359-66. [Crossref] [PubMed]
34. Ledermann T, Kenny DA. Analyzing dyadic data with multilevel modeling versus structural equation modeling: A tale of two methods. J Fam Psychol 2017;31:442-52. [Crossref] [PubMed]
35. Kenny DA, Kashy DA, Cook WL. Dyadic Data Analysis. Guilford Publications; 2020.
36. Nahum-Shani I, Smith SN, Spring BJ, et al. Just-in-Time Adaptive Interventions (JITAIs) in Mobile Health: Key Components and Design Principles for Ongoing Health Behavior Support. Ann Behav Med 2018;52:446-62. [Crossref] [PubMed]
37. Nahum-Shani I, Shaw SD, Carpenter SM, et al. Engagement in digital interventions. Am Psychol 2022;77:836-52. [Crossref] [PubMed]
38. Radovic A, Badawy SM. Technology Use for Adolescent Health and Wellness. Pediatrics 2020;145:S186-94. [Crossref] [PubMed]
39. Prior HHS poverty guidelines and federal register references [Internet]. ASPE. [cited 2022 Apr 29]. Available online: https://aspe.hhs.gov/topics/poverty-economic-mobility/poverty-guidelines/prior-hhs-poverty-guidelines-federal-register-references
40. Kid Count Data Center. Children below 200 percent poverty [Internet]. [cited 2022 Jun 7]. Available online: https://datacenter.kidscount.org/data/tables/47-children-below-200-percent-poverty?loc=1&loct=2#detailed/2/24/true/867/any/329,330
41. Thompson T, Ketcher D, Gray TF, et al. The Dyadic Cancer Outcomes Framework: A general framework of the effects of cancer on patients and informal caregivers. Soc Sci Med 2021;287:114357. [Crossref] [PubMed]
42. Gutiérrez-Colina AM, Lee JL, VanDellen M, et al. Family Functioning and Depressive Symptoms in Adolescent and Young Adult Cancer Survivors and Their Families: A Dyadic Analytic Approach. J Pediatr Psychol 2017;42:19-27. [PubMed]
43. Lyons KS, Lee CS. The Theory of Dyadic Illness Management. J Fam Nurs 2018;24:8-28. [Crossref] [PubMed]
44. Juth V, Silver RC, Sender L. The shared experience of adolescent and young adult cancer patients and their caregivers. Psychooncology 2015;24:1746-53. [Crossref] [PubMed]
45. Shin JY, Kedroske J, Vue R, et al. Design considerations for family-centered health management. In: Proceedings of the 17th ACM Conference on Interaction Design and Children - IDC ’18 [Internet]. New York, New York, USA: ACM Press; 2018. Available online: 10.1145/3202185.321078110.1145/3202185.3210781
46. Shin JY, Kang TI, Noll RB, et al. Supporting Caregivers of Patients With Cancer: A Summary of Technology-Mediated Interventions and Future Directions. Am Soc Clin Oncol Educ Book 2018;38:838-49. [Crossref] [PubMed]
47. Salsman JM, Butt Z, Pilkonis PA, et al. Emotion assessment using the NIH Toolbox. Neurology 2013;80:S76-86. [Crossref] [PubMed]
48. Lyubomirsky S, Layous K. How Do Simple Positive Activities Increase Well-Being? Curr Dir Psychol Sci 2013;22:57-62. [Crossref]
49. Nguyen J, Maio V, Handley N. Improving Cancer Care Through Digital Health Coaching. Am J Med Qual 2022;37:369-70. [Crossref] [PubMed]
50. Janssen SHM, van der Graaf WTA, van der Meer DJ, et al. Adolescent and Young Adult (AYA) Cancer Survivorship Practices: An Overview. Cancers (Basel) 2021;13:4847. [Crossref] [PubMed]