Abstract
Purpose
Elevated costs of cancer treatment can result in economic and psychological “financial toxicity” distress. This pilot study assessed the feasibility of a point-of-care intervention to connect adult patients with cancer-induced financial toxicity to telehealth-delivered financial counseling.
Methods
We conducted a three-armed parallel randomized pilot study, allocating newly referred patients with cancer and financial toxicity to individual, group accredited telehealth financial counseling, or usual care with educational material (1:1:1). We assessed the feasibility of recruitment, randomization, retention, baseline and post-intervention COmprehensive Score for Financial Toxicity (COST), and Telehealth Usability Questionnaire (TUQ) scores.
Results
Of 382 patients screened, 121 were eligible and enrolled. 58 (48%) completed the intervention (9 individual, 9 group counseling, 40 educational booklet). 29 completed follow-up surveys: 45% female, 17% African American, 79% white, 7% Hispanic, 55% 45–64 years old, 31% over 64, 34% lived in rural areas, 24% had cancer stage I, 21% II, 7% III, 31% IV. Baseline characteristics were balanced across arms, retention status, surveys completion. Mean (SD) COST was 12.4 (6.1) at baseline and 16.0 (8.4) post-intervention. Mean (SD) COST score differences were 6.3 (11.6) after individual counseling, 5.8 (8.5) after group counseling, and 2.5 (6.4) after usual care. Mean TUQ score among nine counseling participants was 5.5 (0.9) over 7.0. Non-parametric comparisons were not statistically meaningful.
Conclusion
Recruitment and randomization were feasible, while study retention presented challenges. Nine participants reported good usability and satisfaction with telehealth counseling. Larger-scale trials focused on improving participation, retention, and impact of financial counseling among patients with cancer are justified.
Keywords: Financial toxicity, Telehealth, Cancer, Financial counseling, Feasibility, SDOH
Introduction
Cancer is among the most financially burdensome conditions in the United States [1]. Costly technological advancements and the improved survival of people diagnosed with cancer contribute to exacerbating its public and private financial impact [2, 3]. While overall mortality rates from cancer continue to decline, incident rates are stable or increasing, consistent with a growing burden of cancer in the population [3]. In the U.S., net annualized out-of-pocket costs for medical services for adults aged 65 and above with cancer can average $2,200 to $3,823 across cancer sites [3]. Recently approved oral cancer treatments cost between $7,500 and $25,000 monthly [4]. In addition, cancer can result in loss of income and resources for patients and their caregivers [5]. With about 32% of adults unable to cover unexpected expenses of $400 with own savings in the U.S. [6], the cost of cancer represents a major financial concern.
The multidimensional psychological, material, and emotional stress that derives from the direct and indirect costs of cancer is often referred to as “financial toxicity” [7]. This concept encompasses economic and financial hardship, psychological distress, and destructive financial coping strategies and behaviors, such as the depletion of assets and indebtment, which increase the risk of insolvency and bankruptcy [5, 8, 9]. Financial distress is associated with delaying and foregoing medical care among patients [10] and their family members [11], quality of life losses [12], treatment non-adherence [13], and early mortality [14]. Financial constraints challenge participation in cancer clinical trials, leading to an under-representation of low socioeconomic status patients, with implications for representativeness and equity [15].
Interventions and services to identify, prevent, or mitigate cancer-related financial toxicity are not well-integrated into clinical care. Patients with financial toxicity have expressed interest in transportation vouchers, understanding the costs and burdens of treatment, access to complementary food during appointments, and assistance with or elimination of insurance deductibles [16]. While many cancer centers provide services to support patients in managing the costs of cancer treatment, additional resources for navigation services and systematic referral mechanisms are needed [17]. Barriers to the provision of financial counseling and advisory services in clinical settings include a lack of dedicated resources and training opportunities, limited price transparency, and the complexity of healthcare systems and health insurance mechanisms [18].
An emerging literature aimed at scoping ways to address financial toxicity identified promising targets for intervention, although evidence of effectiveness is still limited [5]. Among such interventions, financial counseling (FC) and navigation could be successful at providing and improving dedicated support to mitigate financial difficulties among patients with cancer [19, 20]. A prospective pilot study to improve patient knowledge, provide FC services, and support out-of-pocket expenses for patients with nonmetastatic solid tumors found that cost-related anxiety decreased in 33% of participants, whereas the analyses did not show significant changes in self-reported financial burden [21]. Another study provided adult patients with solid tumors with a financial education video, monthly counseling, and referral to assistance with unpaid non-medical bills for 6 months [22]. 33% of enrolled patients completed follow-up surveys and reported satisfaction with counseling and assistance but found the educational video challenging and overwhelming. Changes in financial toxicity measurements were not statistically significant. An analysis of Medicare claims in the Southeastern U.S. showed that participation in a lay navigation program was associated with a faster decline in Medicare spending and a decrease in the risk of emergency department visits [23]. Sadigh et al. did not find a significant difference in financial toxicity scores of patients who had financial navigation at three-month follow-up, possibly due to a small sample size [24]. Overall, the literature is still sparse and findings rely mostly on small samples, calling for additional evidence.
In response to this knowledge gap, the current study assessed the feasibility of a point-of-care intervention to connect patients with cancer facing financial toxicity with telehealth FC services. We conducted a pilot trial that delivered group or individual telehealth FC sessions in comparison to usual care with an educational booklet.
Methods
Study procedures
We conducted a single-institution pilot parallel randomized control trial with three arms with a 1:1:1 allocation procedure. Two intervention arms provided either (1) group or (2) individual telehealth FC. A control arm provided usual care and an informational booklet. Clinical staff introduced eligible patients to a research assistant who offered information about the study. If interested, the patient received a waiver of documentation to consent to participate and completed a questionnaire to assess financial toxicity using the COmprehensive Score for Financial Toxicity (COST) instrument [25] via the Research Electronic Data Capture (REDCap) platform. If interested and eligible, participants completed an electronic informed consent form. Those who expressed interest in the study, but did not want to participate immediately, were asked to complete a consent to recontact form. The study team followed-up with potential participants by telephone and provided a waiver of documentation to consent via email and a link to the COST questionnaire if interested. Eligible participants voluntarily consented via remote videoconferencing or telephone as preferred. Informed consent and participants’ protection procedures were approved by the University of Florida’s Institutional Review Board (IRB202001894). Participating patients were randomly assigned to one of the three arms using the RED-Cap’s randomization procedure.
Patients in the control group received usual care and the American Society of Clinical Oncology’s “Managing the Cost of Cancer Care” booklet [26]. The booklet provided an overview of direct and indirect cancer-related costs, employment, legal, and financial issues. It suggested strategies to discuss financial concerns with the medical team and insurance companies and further contact information for cancer support resources, governmental programs, insurance providers such as Medicare and Medicaid, and transportation and housing services. Three months after the first interaction, the study team contacted the patients via telephone or email and asked to complete another COST questionnaire via REDCap.
Participants in both counseling arms were referred to the FC service and required to complete a minimum of two sessions with a financial counselor from the University of Florida’s Institute of Food and Agricultural Sciences (UF/ IFAS) Extension program. The counselors in this project were University County Extension Agents, who are field faculty members with a human science Master’s Degree and have obtained an Accredited Financial Counselor (AFC®) designation. This designation is obtained after completing accrediting study and passing the comprehensive exam, as well as meeting an experience requirement. Cooperative Extension faculty are part of the partnership between the United States Department of Agriculture and state governments to provide outreach and engagement from their Land Grant Universities. The UF/IFAS Extension program has offices located in all counties across the state, with over a century of education and engagement activities which resulted in social trust with the communities of outreach. The counseling team was part of the UF/IFAS Cooperative Extension program and directed by the Associate Dean for Extension, Multiple Principal Investigator in this study. Extension Agents were not part of the study team.
At baseline, patients in the treatment arms completed an intake assessment and a COST questionnaire. Then, they were invited to participate in two FC sessions, approximately two weeks apart. The sessions involved diagnostics to determine financial standing, a review of employment status, income, assets, billing, and insurance, and referrals to financial and social services offered by the health system, the government, and nonprofit and private organizations. Examples included transportation benefits, charity services, Supplemental Security Income benefits, housing assistance, and Medicaid/Medicare coverage. The counselor provided a list of resources and the required forms, assisted with the application process, and tracked its status. After the second session, three months after study enrollment, the study team contacted the participants and asked to complete a follow-up COST questionnaire and a telehealth usability survey through the REDCap platform. Participants were contacted up to three times by members of the study team if they did not respond and/or did not complete the follow-up questionnaire.
The sessions occurred via the Zoom videoconferencing platform. Individual sessions included the counselor and the participant. Group counseling included up to three randomly-selected participants.
Recruitment
The research team recruited the study participants among newly referred adult patients with cancer in UF Health clinics between February and December 2021, within three-months after the initial diagnosis. The catchment area’s population has high rates of cancer and cancer risk factors and social vulnerabilities, obesity, tobacco use, late-stage cancer diagnosis, and low cancer screening rates, in comparison with state and national averages [27]. Eligibility criteria to participate in the study included being 18 years or older, having a new cancer diagnosis, and a score of 22 or less points in the COST questionnaire, representing significant financial toxicity [25].
Study size calculation
Based on preliminary work [28], the study team assumed a 20% increase in COST score for the individual counseling arm, 10% increase in the group counseling arm, versus 5% increase in the control arm (mean COST score of 20 and within-group standard deviation of 3.0). Power calculations indicated that a sample of 33 participants was necessary to achieve 91% power to detect differences in COST scores using an F test with a 0.017 significance level (after Bonfer-roni correction). Considering an expected drop-out rate of 20%, the study team aimed at enrolling a total of 126 study subjects (42 per group) for this pilot study.
Study measures: outcomes
Feasibility measures
The main outcomes for evaluating the feasibility of the pilot intervention [29, 30] were descriptive measures of eligibility and recruitment (number of eligible participants and number of participants recruited with respect to recruitment goal), randomization balance (absence of statistically significant differences in participants’ demographic and clinical characteristics across study arms), study retention (number of participants retained, dropped out, or lost to follow up by intervention arm and study phase, and absence of statistically significant differences by study retention status), baseline and post-intervention financial toxicity by intervention arm, and participants’ perceptions of usability of telehealth counseling (Telehealth Usability Questionnaire scores).
Financial toxicity
We measured patient-reported financial toxicity using COST, a survey instrument with validated psychometric properties [25, 31]. The COST includes eleven questions regarding concerns with current and future financial resources, expenses, direct costs of care, and indirect consequences of cancer, such as a limited ability to work and generate income. Responses to these questions are aggregated in a single score ranging between 0 and 44. Higher scores correspond to lower financial concerns.
Telehealth counseling usability
To assess participants’ experience with telehealth FC, we used the Telehealth Usability Questionnaire (TUQ). The TUQ includes twenty-one questions that elicit feedback on six usability domains: usefulness, ease of use and learnability, interface, interaction quality, reliability, and satisfaction. Each question is assessed over a seven-point Likert scale. The value of 1 corresponds to “strongly disagree” and 7 corresponds to “strongly agree” [32]. Overall and domain-specific scores are computed as the arithmetic mean of all non-missing responses.
Ease of use and learnability indicate that the system provides intuitive navigation and facilitates the assimilation of information. Interface quality refers to video and audio quality. Interaction quality refers to resemblance of the features and comfort of in-person encounters. Reliability refers to the system’s ability to allow for easy recovery when users experience an error (e.g., wrong button clicks). The last domain assesses satisfaction with and willingness to use the system again [32].
Statistical analyses
To report measures of study eligibility, recruitment, and retention, we computed numbers and percentages. We assessed statistical balance in participants’ characteristics across the three arms with Fisher’s exact tests of differences in proportions. Further, we used Fisher’s exact tests to evaluate differences in baseline characteristics of participants that (a) completed baseline and follow-up questionnaires, (b) completed the intervention but not follow-up questionnaires, and (c) consented but dropped out before any FC session. We calculated mean, median, and standard deviations (SD) of baseline and follow-up COST scores by arm and compared paired pre- and post-intervention median scores (i) within and (ii) between study arms, using, respectively, a Wilcoxon exact signed rank test and a Kruskal–Wallis rank sum test. We estimated post-intervention mean and median scores, SD, and inter-quartile ranges of the telehealth usability survey responses among patients in the individual and group counseling arms. Results and statistical tests were produced using RStudio, Version 4.1.3. A P-value equal or less than 0.05 was selected as statistical significance threshold.
Results
Recruitment, randomization, retention, and study participants
Of 382 newly referred patients with cancer were screened for eligibility, 121 (32%) met eligibility criteria and signed an informed consent to participate, 220 (58%) did not meet the inclusion criteria, and 41 (11%) declined to participate. Figure 1 displays the flow and retention of patients throughout the study. Recruitment goals were achieved, with 121 patients randomly allocated to one of the three intervention arms. Of the forty patients in the usual care arm, twenty (50%) completed the follow-up questionnaires. Among the forty-one participants in group counseling, nine (22%) completed both counseling sessions and six of them (15%) completed follow-up questionnaires. Of the forty patients in the individual counseling arm, nine completed both counseling sessions (22%) and three of them (7%) completed follow-up questionnaires. Six participants in the individual counseling arm dropped out between the first and second counseling session, which resulted in a completion rate of the counseling intervention of 60%. All nine participants in the group counseling arm who did the first session completed also the second session, corresponding to a completion rate of 100%. The combined completion rate was 75%. After the intervention, the study team was unable to contact one half of the 18 participants who completed their assignment to fill in follow-up surveys. With respect to the number of patients enrolled in the study, participants’ responsiveness at three-months follow-up (completion of follow-up questionnaires) was higher in the usual care arm (50%) that in individual (7%) and group (15%) counseling arms. However, with respect to those who completed the FC sessions, if assigned to the intervention arm, responsiveness to follow-up questionnaires was the same in the intervention arms combined and in the control arm (50%).
Fig. 1.
CONSORT study flow diagram [33]
Table 3 in the Appendix reports the characteristics of the 121 patients with breakdowns by treatment arm. The data along with p-values of Fisher’s exact tests of differences in demographic and clinical characteristics among patients across the treatment arms showed patients did not present considerable differences between intervention arms, confirming that randomization was feasible and balanced.
Table 3.
Demographic and clinical characteristics of the consented study population by intervention arm (individual counseling, n = 40; group counseling, n = 41; usual care, n = 40)
| Characteristic | Overall population | Treatment arm | p-value | ||
|---|---|---|---|---|---|
| Group counseling | Individual counseling | Usual care: booklet | |||
| Sex | |||||
| Female | 70 (58%) | 22 (31%) | 29 (41%) | 19 (27%) | 0.065 |
| Male | 51 (42%) | 19 (37%) | 11 (22%) | 21 (41%) | |
| Race | |||||
| African American | 24 (20%) | 6 (25%) | 8 (33%) | 10 (42%) | 0.5 |
| Other | 3 (2.5%) | 2 (67%) | 0 (0%) | 1 (33%) | |
| White | 94 (78%) | 33 (35%) | 32 (34%) | 29 (31%) | |
| Ethnicity | |||||
| Hispanic or Latino | 5 (4.2%) | 1 (20%) | 2 (40%) | 2 (40%) | > 0.9 |
| Not Hispanic or Latino | 113 (96%) | 38 (34%) | 37 (33%) | 38 (34%) | |
| Unknown | 3 | 2 | 1 | 0 | |
| Age, in years | |||||
| Less than 45 | 21 (18%) | 8 (38%) | 5 (24%) | 8 (38%) | 0.2 |
| 45–64 years | 59 (49%) | 20 (34%) | 16 (27%) | 23 (39%) | |
| More than 64 | 40 (33%) | 13 (32%) | 19 (48%) | 8 (20%) | |
| Unknown | 1 | 0 | 0 | 1 | |
| Rurality | |||||
| Not identified | 6 (5.0%) | 2 (33%) | 2 (33%) | 2 (33%) | > 0.9 |
| Not rural | 77 (64%) | 26 (34%) | 25 (32%) | 26 (34%) | |
| Rural | 38 (31%) | 13 (34%) | 13 (34%) | 12 (32%) | |
| Cancer stage | |||||
| I | 20 (17%) | 8 (40%) | 5 (25%) | 7 (35%) | 0.3 |
| II | 15 (12%) | 8 (53%) | 4 (27%) | 3 (20%) | |
| III | 24 (20%) | 10 (42%) | 8 (33%) | 6 (25%) | |
| IV | 39 (32%) | 12 (31%) | 12 (31%) | 15 (38%) | |
| Missing/unknown | 23 (19%) | 3 (13%) | 11 (48%) | 9 (39%) | |
| Total | 121 (100%) | 41 (100%) | 40 (100%) | 40 (100%) | |
Demographic characteristics and baseline cancer stage of the 121 patients recruited and consented, overall and by treatment arm. Colum 5 reports the p-value of the Fisher’s exact test of differences in proportions between treatment arms
Table 1 reports the baseline characteristics of the 121 patients that consented to participate in the study. Of those, 29 completed individual, group counseling, or usual care and also completed the follow-up questionnaires, 25 completed the intervention (counseling or booklet) but did not complete the follow-up questionnaire, and 57 patients dropped out before participating in group or individual counseling sessions. The statistical tests indicated that participants who did not complete the end-line questionnaire, those who did, and those who did not participate in the assigned counseling sessions did not present significantly different characteristics along the observed traits.
Table 1.
Demographic and clinical characteristics of the study population by study retention group: participants who provided consent (n = 121), completed the intervention (usual care or counseling) and follow-up questionnaires (n = 29), completed the intervention but not the follow-up questionnaires (n = 35), and were assigned to counseling but dropped out before the sessions (n = 57)
| Consented (all) | Treated and completed follow-up questionnaire |
Treated and no follow-up questionnairea |
Only consented (individual and group counseling) |
p-value | |||||
|---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
||||||
| n | % | n | % | n | % | n | % | ||
| Sex | 0.219 | ||||||||
| Female | 70 | 58 | 13 | 45 | 20 | 57 | 37 | 65 | |
| Male | 51 | 42 | 16 | 55 | 15 | 43 | 20 | 35 | |
| Race | 0.782 | ||||||||
| African American | 24 | 20 | 5 | 17 | 9 | 26 | 10 | 18 | |
| Other | 3 | 2 | 1 | 3 | 1 | 3 | 1 | 2 | |
| White | 94 | 78 | 23 | 79 | 25 | 71 | 46 | 81 | |
| Ethnicity | 0 | 0 | 0 | 0 | 0.722 | ||||
| Hispanic | 5 | 4 | 2 | 7 | 1 | 3 | 2 | 4 | |
| Not Hispanic | 113 | 93 | 27 | 93 | 33 | 94 | 53 | 93 | |
| Unknown | 3 | 2 | 0 | 0 | 1 | 3 | 2 | 4 | |
| Age | 0.720 | ||||||||
| Less than 45 | 21 | 17 | 4 | 14 | 5 | 14 | 12 | 21 | |
| 45–64 years | 59 | 49 | 16 | 55 | 19 | 54 | 24 | 42 | |
| More than 64 | 40 | 33 | 9 | 31 | 10 | 29 | 21 | 37 | |
| Unknown | 1 | 1 | 0 | 0 | 1 | 3 | 0 | 0 | |
| Rurality | 0.642 | ||||||||
| Not identified | 6 | 5 | 2 | 7 | 1 | 3 | 3 | 5 | |
| Not Rural | 77 | 64 | 17 | 58 | 26 | 74 | 34 | 60 | |
| Rural | 38 | 31 | 10 | 34 | 8 | 23 | 20 | 35 | |
| Cancer stage | 0.190 | ||||||||
| I | 20 | 17 | 7 | 24 | 6 | 17 | 7 | 12 | |
| II | 15 | 12 | 6 | 21 | 3 | 9 | 6 | 11 | |
| III | 24 | 20 | 2 | 7 | 6 | 17 | 16 | 28 | |
| IV | 39 | 32 | 9 | 31 | 10 | 29 | 20 | 35 | |
| Unknown | 23 | 19 | 5 | 17 | 10 | 29 | 8 | 14 | |
| Total | 121 | 100 | 29 | 100 | 35 | 100 | 57 | 100 | |
p-values were calculated from a Fisher’s exact test of the difference between (i) participants who completed the assigned intervention and follow-up questionnaire, (ii) participants that completed the assigned intervention but not the follow-up questionnaire, and (iii) participants that provided informed consent but did not do the counseling sessions. aNote that “Treated and no follow up questionnaire” includes also six participants in the individual group who attended only one counseling session
Of the 29 participants that completed counseling or usual care and the follow-up questionnaires, 45% were female, 17% African American, 79% white, and 7% were of Hispanic ethnicity. Of these 29 participants, 55% were between 45 and 64 years old, 14% less than 45, and 31% 64 or above, and 34% lived in rural areas. Most patients presented cancer stages I (24%) and IV (31%), followed by stage II (21%) and III (7%). Five patients had missing cancer stage information (17%).
Financial toxicity COST scores
Table 2 reports the mean and standard deviations (SD) of pre- and post-intervention COST scores and their differences for the entire sample (“All”), each study arm, and individual counseling (“IC”) and group counseling (“GC”) arms together (“IC + GC”).
Table 2.
Results of the Comprehensive Score for Financial Toxicity (COST) questionnaire at baseline (pre-intervention) and follow-up (post financial counseling sessions) and differences between post and pre-intervention COST scores by intervention arm (n = 29)
| Study arm | n | Mean | SD | Median |
p-value (pre vs post, within arm) |
|---|---|---|---|---|---|
| Group counseling (GC) | |||||
| Post | 6 | 18.5 | 7.3 | 16 | 0.156 |
| Pre | 6 | 12.7 | 6.6 | 13.5 | |
| Difference | 6 | 5.8 | 8.5 | 7 | |
| Individual counseling (IC) | |||||
| Post | 3 | 18.3 | 12.5 | 18 | 0.5 |
| Pre | 3 | 12 | 5.6 | 13 | |
| Difference | 3 | 6.3 | 11.6 | 12 | |
| IC + GC | |||||
| Post | 9 | 18.4 | 8.5 | 17 | 0.074 |
| Pre | 9 | 12.4 | 5.9 | 13 | |
| Difference | 9 | 6 | 8.9 | 9 | |
| Usual care and booklet | |||||
| Post | 20 | 14.9 | 8.8 | 14 | 0.136 |
| Pre | 20 | 12.4 | 6.4 | 11.5 | |
| Difference | 20 | 2.5 | 6.4 | 1 | |
| All | |||||
| Post | 29 | 16 | 8.7 | 15 | |
| Pre | 29 | 12.4 | 6.1 | 13 | |
| Difference | 29 | 3.6 | 7.3 | 3 | |
p-values of the pre vs post-intervention difference within study arm (col. 4) were calculated from Wilcoxon’s exact signed rank test. “IC + GC” corresponds to the aggregation of individual and group counseling populations
At baseline, the study population had a mean (SD) COST score of 12.4 (6.1), indicating that, on average, participants had more financial concerns than the eligibility threshold of 22 points. At follow-up, the mean score increased by 3.6 (7.3), with differences between arms. The IC arm presented the highest absolute point-estimate increase (by 6.3 points, 11.6 SD), for a post-intervention mean score of 18.3 (12.5), followed by the GC arm by 5.8 (8.5), with a post-intervention mean score of 18.3 (7.3). The mean score in the usual care arm increased by 2.5 points (6.4), leading to 14.9 (8.8). The differences within and between arms did not reached the statistical significance threshold, partially due to the small sample sizes. The p-values of a Kruskal–Wallis rank sum test for differences in pre, post, or changes in COST scores between groups were, respectively, 0.998, 0.576, and 0.498.
Feasibility of telehealth counseling and satisfaction
Figure 2 reports the distribution of Telehealth Usability Questionnaire (TUQ) mean scores by intervention arm (individual versus group counseling) and domain. Higher scores indicate higher satisfaction and feasibility. Average TUQ domain-specific mean scores ranged between 5.6 (1.2 SD) and 6.1 (1 SD) over a maximum of 7. Participants in the IC arm reported higher mean scores, on average, that those in the GC arm across all domains, suggesting an overall better experience in terms of usefulness and feasibility. However, p-values from the Wilcoxon exact rank sum test for differences between IC and GC’s TUQ scores were all above the significance threshold (except a 0.05 for reliability), indicating that the observed differences in median values between treatment arms cannot be interpreted as statistically significant. Table 4 in the Appendix reports mean and median values, SDs, and p-values by TUQ domain and counseling arm.
Fig. 2.
Feasibility of the telehealth financial counseling sessions. Distribution of post-intervention Telehealth Usability Questionnaire scores by domain (Usefulness, Ease of Use, Interface Quality, Interaction Quality, Reliability, Satisfaction and Future Use) and by intervention arm (n = 9. Group financial counseling, n = 6; Individual financial counseling, n = 3)
Table 4.
Results of the Telehealth Usability Questionnaire (n = 9)
| Domain | Study arm | n | Median (IQR) | Mean (SD) | Min | Max | p-value |
|---|---|---|---|---|---|---|---|
| Telehealth Usability Questionnaire | |||||||
| Usefulness | All | 9 | 6.3 (5.3–6.7) | 5.9 (1.1) | 4 | 7 | 0.152 |
| Group counseling | 6 | 5.7 (4.6–6.2) | 5.5 (1.2) | 4 | 7 | ||
| Individual counseling | 3 | 6.7 (6.5–6.8) | 6.7 (0.3) | 6.3 | 7 | ||
| Ease of use | All | 9 | 6.3 (5.7–6.7) | 6.1 (1) | 4 | 7 | 0.118 |
| Group counseling | 6 | 5.8 (5.4–6.2) | 5.7 (1) | 4 | 7 | ||
| Individual counseling | 3 | 6.7 (6.7–6.8) | 6.8 (0.2) | 6.7 | 7 | ||
| Interface quality | All | 9 | 6.5 (5.2–7) | 6 (1.1) | 4 | 7 | 0.354 |
| Group counseling | 6 | 5.4 (5.2–6.6) | 5.7 (1.2) | 4 | 7 | ||
| Individual counseling | 3 | 6.5 (6.5–6.8) | 6.7 (0.3) | 6.5 | 7 | ||
| Interaction quality | All | 9 | 6.8 (5.5–7) | 6.1 (1) | 4 | 7 | 0.114 |
| Group counseling | 6 | 5.8 (5.3–6.6) | 5.8 (1.1) | 4 | 7 | ||
| Individual counseling | 3 | 7 (6.9–7) | 6.9 (0.1) | 6.8 | 7 | ||
| Reliability | All | 9 | 5.7 (4.3–6.7) | 5.6 (1.2) | 4 | 7 | 0.05 |
| Group counseling | 6 | 5 (4.1–5.7) | 5.1 (1.1) | 4 | 6.7 | ||
| Individual counseling | 3 | 7 (6.7–7) | 6.8 (0.4) | 6.3 | 7 | ||
| Satisfaction and future use | All | 9 | 6 (5.5–7) | 6 (1) | 4 | 7 | 0.086 |
| Group counseling | 6 | 5.8 (5.1–6) | 5.6 (1) | 4 | 7 | ||
| Individual counseling | 3 | 7 (6.8–7) | 6.8 (0.3) | 6.5 | 7 | ||
A p-value equal or below the significance level of 0.05 indicates that one of the group’s median is significantly different from the other group’s median for the relative usability domain. p-values are based on a Wilcoxon’s Exact Rank Sum Test
Discussion
Summary
The goals of this study were to assess the preliminary feasibility of recruitment, randomization, retention, and usability of a point-of-care pilot intervention to connect newly diagnosed patients with cancer with financial toxicity to individual and group telehealth FC, and to provide preliminary estimates of intervention outcomes. The results showed that recruitment and randomization goals were feasible and achievable, but intervention uptake and study retention presented more challenges than anticipated.
Value of intervention (recruitment/satisfaction)
The enrollment rate among eligible potential participants (75%) was higher than similar interventions in which eligibility criteria did not formally include baseline financial toxicity (e.g., 57.5% in Sadigh et al. [24]). Our high enrollment rate suggests that patients with financial toxicity are interested in receiving FC and value this opportunity. The value of FC interventions is further supported by the fact that 75% of participants who had a first counseling session completed the intervention. The relatively high completion rates among intervention participants suggest that patients with cancer who managed to schedule and enroll in the counseling sessions valued the counseling intervention and its availability was convenient. Additionally, our assessment of telehealth usability showed that participants in individual and group arms who completed counseling sessions and follow up questionnaires gave positive feedback. This finding is in line with similar studies that have found good participant satisfaction with FC interventions, despite different eligibility criteria and designs [21, 34].
Retention: general
Although we did not encounter challenges in recruiting patients with financial toxicity for our study, our overall retention was low. Most participants who dropped out of the study were lost to follow-up prior to initiating counseling sessions, indicating that primary challenges may lie in logistical and communication obstacles. Our results are similar to previous studies that have reported low participant retention throughout the course of the intervention and follow-up assessments [21, 22, 34]. In our study, two thirds of participants in the counseling arms were lost to follow-up before beginning any counseling session. This indicates that most retention issues did not relate to the intervention content. Unfortunately, the study team was not able to assess reasons for dropout among participants who were lost to follow-up.
Other studies have found that patients feeling overwhelmed is a common reason for dropout [21]. There are several other possible explanations for low retention. Health concerns and treatment planning may outweigh financial and material concerns, especially for patients recently diagnosed with cancer. The salience of financial concerns for patients with cancer may emerge later than three-months after a diagnosis, which was our time window for enrollment. In addition, patients with advanced cancer may find participating in counseling sessions particularly challenging.
Strategies to improve retention: delivery
Given the recurring retention struggles across studies, strategies are needed to improve engagement and retention of participants. Tailoring intervention delivery modalities to patient preferences and needs may serve as one method to enhance patient engagement with financial counseling interventions. Previous research shows higher participation rates for telephone and asynchronous components of financial counseling [21, 34], suggesting that reducing travel and scheduling barriers may increase the reach of FC interventions. While the use of video conferencing in our study removed travel requirements, it may have also introduced additional barriers related to technological literacy and internet availability. Challenges related to internet access may have been especially salient for participants in the study who resided in rural communities. Sadigh et al. found that all study participants opted for phone sessions instead of optional video-conferencing [24], suggesting that providing telephone options may enhance engagement.
Retention: format
Intervention delivery format may also affect patient engagement. Most previously tested FC interventions were delivered in an individual format. One study initially delivered a financial education course in an in-person group format, but patient feedback led to the inclusion of an option to view an online video of the course. Subsequent participation in other components of the intervention was higher among patients who viewed the online video compared to those who attended the in-person group session (69% vs. 50%) [21]. In our study, group sessions presented additional logistical difficulties in coordinating multiple participants’ availability and perhaps some were unwilling to discuss financial stress around strangers. Personalized communication between the patient and the research team may represent an opportunity to improve monitoring and participants’ retention in future studies. While the use of group intervention delivery may be a more efficient use of resources, future work should consider the relative tradeoffs between cost-savings and potential for reduced patient engagement.
Another promising approach to enhancing engagement in financial counseling is the inclusion of caregivers. While only 59% of patients completed at least one intervention component in a study conducted by Shankaran et al., in a subsequent trial that included caregivers, this rate rose to over 90% [22]. We did not include caregivers in our study. Including family caregivers in FC interventions may lessen some of the burden associated with participating in the intervention while undergoing treatment.
Effectiveness
Mean financial toxicity COST scores improved in post-intervention assessments compared to baseline for each study arm, suggesting that FC could be associated with a relief in financial concerns in larger clinical trials. A breakdown by intervention arm showed that changes in point estimates of mean scores were higher for three participants in the individual counseling arm than six participants in the group counseling arm and twenty participants in the usual care arm, but the three groups had overlapping confidence intervals. Due to small sample size, the results cannot be compared at statistically meaningful levels and require further validation in future larger studies. Lower observed financial toxicity scores after counseling sessions among nine FC participants are consistent with other studies in this emerging field of research. A recent pre-post assessment feasibility study of a financial navigation and counseling service for patients with cancer found that participants’ average COST scores increased by 6 points, from 10.0 (9.6 SD) to 16.9 (8.1 SD) [24]. Another study found that, adjusting for gender and nodal status, patients undergoing radiation treatment who utilized FC reported a decrease in financial hardship (− 0.204 ± 0.096), representing a 0.2 units reduction compared with matched patients treated before the integration of FC in the health care delivery model [19]. A financial coaching intervention with 12 patients with cancer and 18 patient-caregiver dyads showed that financial standing improved, while COST and Caregiver Strain Index scores did not show statistically significant changes at follow up [22].
Mean post-intervention COST scores in our study did not surpass the threshold value of 22 of considerable financial toxicity, suggesting that participants maintained substantial financial concerns, despite improvements from pre-intervention values. Similarly, another FC pilot study found that post-intervention COST scores remained below the threshold value of 22 [24]. However, confidence intervals around the post-intervention mean in our study included also values above 22, suggesting that larger-scale randomized trials with adequate statistical power may be capable of demonstrating improvement in financial concerns for patients with cancer.
Limitations
High attrition among participants before completion of follow-up questionnaires can lead to several potential limitations. Small sample size and limited statistical power did not allow the research team to infer statistically meaningful differences at conventional levels for COST scores and feasibility scores between and within intervention arms. Most evaluations of preliminary efficacy of FC interventions in the literature are subject to similar limitations due to small sample sizes and retention issues [35]. High study dropout rates may also introduce selection bias if participants who remained in the study were characterized by unobservable traits, such as a preference for telehealth services or better potential for improvement in terms of financial toxicity, that differed from those who dropped out. However, a comparison between participants with different study retention profiles based on observable characteristics did not reveal the presence of statistically significant differences between participants who remained and those who dropped out (Table 1) along observable traits. Small sample sizes did not allow us to perform the same test within treatment arm. Nonetheless, initial enrollment rates and completion rates upon take up of the first FC session in our study suggest that FC services are valued by patients with cancer and may be capable of impacting financial toxicity. Larger studies that build on these findings and further improve participation and retention in financial counseling and enhance its impact are justified.
Conclusion
This pilot study indicated that recruitment and randomization goals were feasible, while retention in the pilot study was more difficult than anticipated. Future studies may build on these findings to explore different frequencies and forms of interaction to improve study participation and retention. Participants in individual and group telehealth counseling sessions who completed follow-up questionnaires reported positive feedback in terms of feasibility, usefulness, ease of use, interaction and interface quality, reliability, and overall satisfaction with telehealth. Future research could build upon these findings to assess the effectiveness of financial counseling through large-scale randomized control studies. The study size and statistical power calculations should consider large dropout rates. Future research could also adopt a narrower focus on the population that presents intersectional social determinants of health and on rural communities, among whom telehealth may be particularly useful.
Funding
Research reported in this publication was supported by the UF Health Cancer Center, supported in part by state appropriations provided in Fla. Stat. §381.915 and the National Cancer Institute of the National Institutes of Health under Award Number P30CA247796. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the State of Florida. Additional funding was provided by NCATS UL1 TR001427.
Appendix
Footnotes
Conflict of interest The authors declare no relevant conflict of interest.
Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the sponsors.
Data availability
The datasets generated and analyzed in this study are not publicly available due to participants’ privacy protection restrictions but are available from the authors upon reasonable request.
References
- 1.Soni A (2009) The five most costly conditions, 1996 and 2006: estimates for the US civilian noninstitutionalized population. Statistical brief# 248. https://www.bmedreport.com/archives/7231. Accessed 1 Feb 2023 [Google Scholar]
- 2.Shih YCT, Smieliauskas F, Geynisman DM, Kelly RJ, Smith TJ (2015) Trends in the cost and use of targeted cancer therapies for the privately insured nonelderly: 2001 to 2011. J Clin Oncol 33(19):2190. 10.1200/JCO.2014.58.2320 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Yabroff KR et al. (2021) Annual report to the nation on the status of cancer, part 2: patient economic burden associated with cancer care. J Natl Cancer Inst 113(12):1670–1682. 10.1093/JNCI/DJAB192 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Carrera PM, Kantarjian HM, Blinder VS (2018) The financial burden and distress of patients with cancer: understanding and stepping-up action on the financial toxicity of cancer treatment. CA Cancer J Clin 68(2):153–165. 10.3322/CAAC.21443 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lentz R, Benson AB, Kircher S (2019) Financial toxicity in cancer care: prevalence, causes, consequences, and reduction strategies. J Surg Oncol 120(1):85–92. 10.1002/JSO.25374 [DOI] [PubMed] [Google Scholar]
- 6.The Fed: Report on the economic well-being of U.S. households in 2021—May 2022. https://www.federalreserve.gov/publications/2022-economic-well-being-of-us-households-in-2021-executive-summary.htm. Accessed 29 Sep 2023 [Google Scholar]
- 7.Altice CK, Banegas MP, Tucker-Seeley RD, Yabroff KR (2017) Financial hardships experienced by cancer survivors: a systematic review. J Natl Cancer Inst 109:2. 10.1093/jnci/djw205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Asche CV et al. (2021) Society of behavioral medicine (SBM) position statement: support increased knowledge and efforts to address the financial burden associated with cancer treatment. Transl Behav Med 11(6):1289–1291. 10.1093/TBM/IBAA073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bentley C, Teckle P, McQuarrie L, Peacock S, El Adam S (2022) Impact of cancer on income, wealth and economic outcomes of adult cancer survivors: a scoping review. BMJ Open 12(9):e064714. 10.1136/BMJOPEN-2022-064714 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Buzaglo JS, Miller MF, Zaleta AK, Johnson J, Gupta NK (2017) Financial toxicity and cancer-related distress among melanoma survivors. J Clin Oncol 35(15):9588–9588. 10.1200/JCO.2017.35.15_SUPPL.9588 [DOI] [Google Scholar]
- 11.Kazzi B et al. (2022) Shared burden: the association between cancer diagnosis, financial toxicity, and healthcare cost-related coping mechanisms by family members of non-elderly patients in the USA. Support Care Cancer 1:1–13. 10.1007/S00520-022-07234-9/TABLES/5 [DOI] [PubMed] [Google Scholar]
- 12.Yousuf Zafar S, McNeil RB, Thomas CM, Lathan CS, Ayanian JZ, Provenzale D (2015) Population-based assessment of cancer survivors’ financial burden and quality of life: a prospective cohort study. J Oncol Pract 11(2):145–150. 10.1200/JOP.2014.001542 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Knight TG et al. (2018) Financial toxicity in adults with cancer: adverse outcomes and noncompliance. J Oncol Pract 14(11):120. 10.1200/JOP.18.00120 [DOI] [PubMed] [Google Scholar]
- 14.Ramsey SD et al. (2016) Financial insolvency as a risk factor for early mortality among patients with cancer. J Clin Oncol 34(9):980–986. 10.1200/JCO.2015.64.6620 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nipp RD et al. (2019) Addressing the financial burden of cancer clinical trial participation: longitudinal effects of an equity intervention. Oncologist 24(8):1048–1055. 10.1634/THEONCOLOGIST.2019-0146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Aviki EM et al. (2022) Patient-reported benefit from proposed interventions to reduce financial toxicity during cancer treatment. Support Care Cancer 30(3):2713–2721. 10.1007/S00520-021-06697-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.de Moor JS et al. (2021) Delivery of financial navigation services within national cancer institute-designated cancer centers. JNCI Cancer Spectr 5:3. 10.1093/JNCICS/PKAB033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Khera N et al. (2020) Current practices for screening and management of financial distress at NCCN member institutions. J Natl Compr Canc Netw 18(7):825–831. 10.6004/JNCCN.2020.7538 [DOI] [PubMed] [Google Scholar]
- 19.Farrugia M et al. (2021) Financial counseling is associated with reduced financial difficulty scores in head and neck cancer patients treated with radiation therapy. Cancers 13(11):2516. 10.3390/CANCERS13112516 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wheeler SB et al. (2022) Lessening the impact of financial toxicity (LIFT): a protocol for a multi-site, single-arm trial examining the effect of financial navigation on financial toxicity in adult patients with cancer in rural and non-rural settings. Trials 23(1):1–16. 10.1186/S13063-022-06745-4/FIGURES/2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shankaran V et al. (2017) Pilot feasibility study of an oncology financial navigation program. J Oncol Pract 14(2):e122–e129. 10.1200/JOP.2017.024927 [DOI] [PubMed] [Google Scholar]
- 22.Watabayashi K et al. (2020) A pilot study of a comprehensive financial navigation program in patients with cancer and caregivers. J Natl Compr Canc Netw 18(10):1366–1373. 10.6004/JNCCN.2020.7581 [DOI] [PubMed] [Google Scholar]
- 23.Williams CP et al. (2022) Impact of lay navigation on utilization and medicare spending for cancer survivors in the ‘deep south.’ J Cancer Surviv 16(4):705–713. 10.1007/S11764-021-01064-0/FIGURES/2 [DOI] [PubMed] [Google Scholar]
- 24.Sadigh G, Coleman D, Switchenko JM, Hopkins JO, Carlos RC (2022) Treatment out-of-pocket cost communication and remote financial navigation in patients with cancer: a feasibility study. Support Care Cancer. 10.1007/S00520-022-07270-5/TABLES/5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.De Souza JA et al. (2014) The development of a financial toxicity patient-reported outcome in cancer: the COST measure. Cancer 120(20):3245–3253. 10.1002/CNCR.28814 [DOI] [PubMed] [Google Scholar]
- 26.Managing the cost of cancer care guidance and resources for patients and families from the American society of clinical oncology ABOUT ASCO. www.conquer.org. Accessed 23 Oct 2019
- 27.University of Florida Health Cancer Center (2020) Cancer population research program pilot project program. https://cancer.ufl.edu/files/2020/03/CPS-Pilot-Project-RFP-FY20.pdf. Accessed 21 Sep 2022
- 28.Gutter M, Navarro G, Mountain T, Cao X, Salloum R (2019) Determining the financial well-being and influencing factors in cancer patients. [Google Scholar]
- 29.Eldridge SM et al. (2016) CONSORT 2010 statement: extension to randomised pilot and feasibility trials. Pilot Feasibility Stud 2(1):2016. 10.1186/S40814-016-0105-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Pilot Studies: Common Uses and Misuses∣NCCIH. https://www.nccih.nih.gov/grants/pilot-studies-common-uses-and-misuses. Accessed 9 April 2023. [Google Scholar]
- 31.de Souza JA et al. (2017) Measuring financial toxicity as a clinically relevant patient-reported outcome: the validation of the comprehensive score for financial toxicity (COST). Cancer 123(3):476–484. 10.1002/CNCR.30369 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Parmanto B, Lewis AN Jr, Graham KM, Bertolet MH (2016) Development of the telehealth usability questionnaire (TUQ). Int J Telerehabil 8(1):3. 10.5195/IJT.2016.6196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Moher D, Schulz KF, Altman D (2001) The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA 285(15):1987–1991. 10.1001/JAMA.285.15.1987 [DOI] [PubMed] [Google Scholar]
- 34.Kircher SM et al. (2019) Piloting a financial counseling intervention for patients with cancer receiving chemotherapy. J Oncol Pract 15(3):E202–E210. 10.1200/JOP.18.00270 [DOI] [PubMed] [Google Scholar]
- 35.Doherty MJ, Thom B, Gany F (2021) Evidence of the feasibility and preliminary efficacy of oncology financial navigation: a scoping review. Cancer Epidemiol Biomark Prev 30(10):1778–1784. 10.1158/1055-9965.EPI-20-1853/671120/AM/EVIDENCE-OF-THE-FEASIBILITY-AND-PRELIMINARY [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.
Data Availability Statement
The datasets generated and analyzed in this study are not publicly available due to participants’ privacy protection restrictions but are available from the authors upon reasonable request.