Lay Healthcare Worker Financial Toxicity Intervention: A Pilot Financial Toxicity Screening and Referral Program

Purpose Financial toxicity is a source of signi�cant distress for patients with urologic cancers, yet few studies have addressed �nancial burden in this patient population. Methods We developed a �nancial toxicity screening program using a lay health worker (LHW) and social worker (SW) to assess and mitigate �nancial toxicity in a single academic medical clinic. As part of a quality improvement project, the LHW screened all newly diagnosed patients with advanced stages of prostate, kidney, or urothelial cancer for �nancial burden using 3 COST tool questions and referred patients who had signi�cant �nancial burden to a SW who provided personalized recommendations. The primary outcome was feasibility de�ned as 80% of patients with �nancial burden completing the SW consult. Secondary outcomes were patient satisfaction, change in COST Tool responses, and qualitative assessment of �nancial resources utilized.


INTRODUCTION
Financial toxicity, or the nancial burden related to cancer care, is a source of signi cant distress for urologic oncology patients. 1Studies show that patients diagnosed with advanced stages of cancer are at increased risk for nancial toxicity due to the characteristics of the disease and associated treatments. 2 Financial toxicity is associated with increased psychological distress, 3 decreased compliance with cancer therapy, 4 and increased mortality. 5Despite the signi cant implications of nancial toxicity, there remain few interventions that address nancial burden among patients with advanced stages of cancer.
Cost-of-care conversations between patients and their oncology clinicians are one approach to identify and mitigate nancial toxicity. 6In 2019, the American Society of Clinical Oncology (ASCO) released a guidance statement recommending that physicians openly discuss cost with patients. 7However, as demonstrated in our prior work, there remain signi cant gaps in oncologists' knowledge and resources available to effectively address nancial burden. 8,9Approaches that have been developed to date and applied to the clinical setting include mobile applications, 10 nancial toxicity tumor boards 11 or nancial navigation programs 12,13 to assist oncologists in addressing and mitigating nancial toxicity.While these approaches are associated with decreased out of pocket costs and increased utilization of nancial assistance programs, they require signi cant resources including technology and nancial navigatorsresources that may not be accessible or feasible to implement in low-resourced and disparate patient care settings.Thus, oncologists and patients agree that novel approaches are needed such as utilization of other members of the care team to assist who may be better positioned to discuss these needs. 14y-healthcare workers (LHW), or nonclinical personnel, have been part of the US healthcare system for decades 15 and expand workforce capacity in the support of patients with cancer.LHW programs are associated with improved patient education regarding cancer topics, cancer screening, appointment adherence, clinical trial retention. 16,17In prior work, such programs have also been shown to improve advance care planning, symptom management, and patient experience and reduce acute care use and total costs of care. 18,19,20To our knowledge this is the rst report of a LHW-led nancial toxicity screening intervention to mitigate nancial toxicity among patients with cancer.This intervention aimed to screen for and address nancial toxicity among patients with advanced prostate, kidney, and urothelial cancer.

Study Design
We piloted a LHW-led quality improvement program in the urologic oncology clinics at a single academic medical center from August 2019 to January 2020.Adults older than 18 years of age with either metastatic or recurrent prostate, kidney or urothelial cancer were eligible to and offered participation in the program.Excluded patients 1) did not have advanced, recurrent, or metastatic cancer, 2) did not have a diagnosis of a urologic cancer, and/or 3) did not receive their primary oncology care at the medical center.

Ethics Statement
This quality improvement study was conducted retrospectively from data obtained for clinical purposes.We consulted with the Stanford University Institutional Review Board (IRB) who determined that our study was quality improvement and did not need ethical approval.An IRB o cial waiver of ethical approval was granted from the Stanford University IRB.

Lay Healthcare Worker Intervention
A LHW who was part of usual care in the clinics screened all patients with a scheduled visit in urologic oncology clinic for eligibility.The LHW then contacted all eligible patients by telephone and assessed nancial burden using an assessment adapted from the validated COST Tool, a patient-reported outcome measure developed to assess nancial toxicity in patients with cancer, 21 and informed by prior work at SCI. 22 The assessment included three statements rated on a Likert scale from 0 to 4 with 0 representing "Not at all" and 4 representing "Very much."The three statements included: "I worry about the nancial problems I will have in the future because of my illness or treatment", "My cancer or treatment has reduced satisfaction with my present nancial situation", and "I feel nancially stressed.""Signi cant nancial burden" was de ned a priori as a response of 3 or greater to all three assessment statements.
For all participants reported "signi cant nancial burden," the LHW encouraged patients to discuss any medication or cancer care-related nancial concerns with their clinical care teams at an upcoming visit and offered a referral to the clinic social worker (SW) who was trained on available nancial resources.
The SW consultation included a review of patients' nancial information, insurance status, and out-ofpocket costs, provided personalized recommendations to patients after review of this nancial information, and referred patients to nancial support services including those offered by the hospital, government, nonpro ts, and private corporations.

Data Collection
In addition to data collected by the LHW during the 10-minute screening telephone interview, within three months of the SW referral, for all participants had reported "signi cant nancial burden" and were referred to the SW, a research assistant conducted a 10-minute telephone interview.The interview assessed postintervention responses to the adapted COST Tool and patient satisfaction with the SW intervention (assessed post-SW consultation by the statement "The SW provided nancial resources that were bene cial to me").Interview assessments also allowed participants to provide any additional feedback or comments regarding the program.

Outcomes
The primary outcome of feasibility was de ned as the percent of patients with "signi cant nancial burden" who completed a SW consultation assessed by clinical chart review by a investigator (DP).The objective was to achieve a completed SW nancial consultation among at least 80% of patients screened by the LHW as having "signi cant nancial burden."Secondary exploratory outcomes included among those who screened as having "signi cant nancial burden": 1) change in nancial burden scores as measured by the COST Tool from pre-intervention to post-intervention, 2) post-intervention patientreported satisfaction with the SW consultation, and 3) assessment of nancial resources and services offered to patients after SW consultation.

Statistical Analysis
We used descriptive statistics to report demographics (age, gender, race and ethnicity, marital status, and income status) and clinical characteristics (tumor type) of participants who were eligible for the program and agreed to participate.We used a paired t-test to compare the mean differences of the results of the pre-and post-intervention COST tool responses.We measured patient satisfaction with the SW consultation as the proportion of patients that answered "yes" to the patient satisfaction statement "The SW provided nancial resources that were bene cial to me" after the SW consultation.We assessed the nancial resources and services offered by the SW to patients with "signi cant nancial burden" who received a SW consultation in a three-step qualitative process.First, an investigator (DP) abstracted SW documentation in the electronic health record of nancial resources that were provided to the patients with "signi cant nancial burden" who received a SW consultation.Then, investigators (DP, GR) categorized nancial services or resources that the SW provided to these patients into the following major categories and subcategories based on literature review: governmental programs (Supplemental Security Income, Social Security Disability Insurance), pharmacy referrals (medication discount), health insurance enrollment and optimization (Medicare, Medicaid, private or state speci c insurance), private nancial resources (non-pro t foundation), institution speci c resources ( nancial counselor referral, nancial assistance program), psychological support (emotional support to patient, emotional support to family members) and non-medical expenses (housing, transportation, food). 23,24,25Then investigators (DP, GR) independently and consecutively coded de-identi ed copies of the SW documentation in the electronic medical record and discussed discrepancies in the resources that were identi ed.

RESULTS
Of 354 eligible patients in the urologic oncology clinic during the project period, 185 (52%) patients agreed to participate in the program.Reasons for not participating included opting out of communication from non-clinical staff (n=45, 21%), receiving primary oncology care outside of the academic medical center (n=26, 12%), deceased prior to contact by LHW (n=15, 7%), and inability to reach the patient by the LHW after 3 attempts (n=84, 59%) as depicted in Figure 1.
Table 1 depicts the baseline demographic and clinical characteristics of the 185 patients who opted to participate in the program.Most were male (n=152, 82%), non-Hispanic White (n=120, 65%), married (n=139, 75%) and with annual household income greater than $100,000 US Dollars (n=125, 68%).The majority had prostate cancer (n=114, 60%), followed by kidney cancer (n=39, 21%) and urothelial cancer (n=23, 12%).A total of 36 participants (19.5%) rated the statement "I worry about the nancial problems I will have in the future because of my illness or treatment" with a 3 or above.A total of 30 participants (16.2%) rated the statement "My cancer or treatment has reduced satisfaction with my present nancial situation" with a 3 or above.A total of 18 participants (9.7%) rated the statement "I feel nancially stressed" with a 3 or above.Figure 2 depicts the responses to the COST Tool questions by percent of patients.Across all 3 COST Tool statements, a total of 18 participants (9.7%) screened as having "signi cant nancial burden." Figure 2 displays responses to the COST Tool screening tool.Signi cant nancial burden was classi ed as "Quite a bit" or "Very Much" responses to all 3 questions.Of the 185 patients screened a total of 18 participants (9.7%) screened as having "signi cant nancial burden."

Primary and Secondary Outcomes
The intervention was feasible.Of the 18 participants who screened positive to having "signi cant nancial burden" and referred for a one-on-one SW nancial consultation, 100% completed the consultation.All participants reported satisfaction with the SW consultation.Speci cally, of the 18 participants who received the SW consultation, all 18 (100%) reported that the SW "provided nancial resources that were bene cial to me."Overall, participants' mean scores were 0.83 points lower on the COST Tool post-intervention than pre-intervention (95% con dence interval [0.26, 1.41]).
All 18 patients received at least one form of nancial resource or service.Seven participants (38.9%) received 1 nancial resource or service, 5 (27.8%) participants received 2 nancial resources or services, 3 (16.7%)participants received 3 nancial resources or services, 2 (11.1%) participants received 4, and 1 (5.6%) participant received 5 nancial resources or services.Figure 3 depicts the major categories of nancial resources or services provided to each patient with "Signi cant nancial burden" that completed a SW consultation.Figure 3 includes the major categories of nancial resources and services.The major categories of nancial resources and services and their respective subcategories are as follows: Assistance with governmental program (Supplemental Security Income, Social Security Disability Insurance), pharmacy referral (medication discount), health insurance enrollment and optimization (Medicare, Medicaid, private or state speci c insurance), private nancial resources (Non-pro t foundations), institution speci c nancial resources ( nancial counselor referral, nancial assistance program), psychological support (emotional support to patient, emotional support to family members), non-medical expenses (housing, transportation and food).

DISCUSSION
In this quality improvement project, we developed and implemented a novel intervention in which a lay health worker screened patients with either metastatic or recurrent prostate, kidney or urothelial cancer for nancial toxicity and referred patients with signi cant nancial burden to a SW who provided personalized nancial resources.The results demonstrate the feasibility of such an approach and the associated reductions in nancial burden and patient-reported satisfaction with the program.
7][28][29] Financial toxicity is increasingly associated with psychological distress, 3 medical nonadherence, 4 and poor quality of life. 30As such, identifying and addressing nancial toxicity for patients with cancer is essential.2][33][34][35] These approaches are associated with improved patient satisfaction, 35 decreased patient anxiety about costs, 33 improved receipt of nancial assistance 36 , and improved COST scores. 35Unlike this intervention, in which we created a stepped approach where a LHW conducted the screening and referred patients with signi cant nancial burden to a SW to conduct the necessary intervention, these prior studies rely on highly skilled workers such as nancial navigators or even professional clinicians such as nurse navigators or oncology clinicians themselves.As was seen in this intervention, attaining the group of 18 patients with signi cant nancial burden that bene tted from a SW consultation required a very labor-intensive screening process: 354 eligible patients were contacted by phone, of which a substantial number declined to participate, with ultimately approximately half agreeing to program participation.In the setting of workforce shortages across oncology clinicians and nursing staff as well as the dearth of availability of skilled nancial counselors and/or nancial navigators universally, the use of highly skilled team members may be suboptimal, especially in settings where nancial toxicity may not be as prevalent or where patients may be less likely to utilize such resources.Regardless of setting, however, it is crucial to consider development of approaches that can increase capacity.Employing a stepped approach, like in our intervention, and screening patients through a LHW before referring them to a SW, may lead to increased capacity for SW, nurse navigators and nancial navigators to more e ciently and effectively deliver tailored services for patients with nancial needs.Our ndings show that utilizing a LHW for this role may be an e cient and sustainable solution.
In this program, patients were satis ed with the SW intervention and found that the services provided were bene cial to them.Patients reported better post-intervention COST Tool scores than pre-intervention scores indicating an associated decrease in nancial distress.These ndings are likely a re ection of the personalized services and recommendations provided by the SW to address each patients' speci c needs.
Furthermore, unique to this LHW intervention, we assessed for nancial toxicity using three of the eleven COST Tool statements which were more closely correlated with signi cant nancial burden in our prior work with a similar patient population. 21We hypothesize that utilizing this adapted COST Tool could contribute to a faster, more effective screening process.However, it is unknown if the three statements we selected from the COST Tool would be appropriate for use in other populations.Further studies are needed to support this hypothesis.
There are limitations to our work.First, this project was conducted at a single institution academic center where the majority of individuals reported annual household incomes of greater than $100,000 US Dollars.As noted in this study, many eligible patients opted out, indicating a potentially lower need for nancial resources among this population.Thus, the results may not be applicable to other settings where nancial toxicity may be more prevalent.Second, the majority of the participants identi ed as White, thus, results may not be applicable to other populations, as it has been shown that racial and ethnic minority populations often experience more severe nancial toxicity. 37Third, non-response bias could have occurred given that one third of the participants recruited for enrollment in the LHW intervention either opted out of the intervention or the LHW was unable to reach them.Fourth, we selected only three of the twelve COST Tool statements.By not using the entire validated tool, we may have excluded participants who could have bene ted from the program.
In conclusion, this program in which a lay health worker screened for nancial toxicity and referred patients with signi cant nancial burden to a social worker was feasible, acceptable, and associated with reductions in nancial distress among participants with advanced genitourinary cancers.Future studies should evaluate the use of such programs as a potential approach for screening for nancial toxicity screening in low resource settings where it may not be feasible to employ clinical workforce for this important care service.

Table 1 :*
Demographic and Clinical Factors DemographicInitial stage at diagnosis I-III however with recurrent disease at the time of enrollment COST Tool Responses

Flow
Chart

Figure 2 COST Tool Participant Responses Figure 3
Figure 2