Implementing a nurse‐led screening clinic for symptom distress with community‐based referral for cancer survivors: A feasibility study

This prospective, single‐arm, pragmatic implementation study evaluated the feasibility of a nurse‐led symptom‐screening program embedded in routine oncology post‐treatment outpatient clinics by assessing (1) the acceptance rate for symptom distress screening (SDS), (2) the prevalence of SDS cases, (3) the acceptance rate for community‐based psychosocial support services, and (4) the effect of referred psychosocial support services on reducing symptom distress.

assessing the fidelity of SDS protocols at tertiary cancer treatment centers showed that only 1-in-2 had documentation of psychosocial referral in patients with high distress. 14en with referral, low patient uptake also can impair successful symptom management programs, 7 and even if accepted, subsequent retention is often poor.Distressed patients appear more receptive to referrals for information support services (89%) than to psychological services (30%) 15 and to services providing physical (e.g.dietetics and physiotherapy) rather than emotional-support services. 16Tailored interventions that address amenable underlying problems rather than solely addressing psychological reactions to these (including psychotropic medication prescriptions) are crucial components for acceptable and, hence, successful SDS programs.
Currently, most studies on the implementation of SDS have either focused on cancer patients receiving active treatment or a mix of patients receiving, and patients following active treatment.However, referral offers may differ by treatment status.Distressed patients who completed active treatment were less likely to receive referral for supportive care services. 17Given many post-treatment patients continue to experience persistent-or late-cancer/treatment effects but have less-frequent follow-up clinic visits, routine SDS with appropriate and acceptable follow-up referral for these patients is essential.A successful screening program thus requires both a tailored symptom-distress screen and a referral pathway tuned to adequatelyresourced cost-effective interventions within the health system.
Here we report a prospective, single-arm, pragmatic implementation study to assess the feasibility of a nurse-led symptom screening program embedded in routine oncology post-treatment surveillance outpatient clinics.The objectives of this study were to determine (1) the acceptance rate for SDS, (2) the prevalence of symptom distress cases, (3) the acceptance rate for community-based psychosocial support services, and (4) the effect of referred psychosocial support services in reducing symptom distress.community-based centers.In developing this program, we considered the HKSAR government cancer strategic report published in 2019, 18 which included the aim of helping cancer survivors stay healthy in the community.Hence, the policy climate strongly supports the transfer of cancer supportive care services from tertiary to community settings after patients complete active curative treatments.Furthermore, we considered resourcing for and staff attitudes toward distress screening and readiness for implementation of distress screening, and sustainability.Crucial to successful implementation was that sites chosen were first and foremost those whose unit heads supported implementing symptom distress screening in their units.

| METHODS
Over the past 5 years at local health conferences, we have presented evidence gathered over 20 years on local cancer populations to argue for local distress screening program.This built credibility for when we met with oncology service heads to "sell" the idea.The first clinic was initiated in 2017 with breast cancer groups and experience was gained on resolving implementation and operational issues.Thereafter the service widened gradually and now includes several oncology surveillance clinics.

| Setting and patients
For this feasibility study, four HKU-affiliated teaching hospital outpatient surveillance clinics, held once-or-twice weekly, from the clinical oncology department, breast center, colorectal surgical oncology unit, and gynecological oncology units collaborated in this implementation program.For each collaborating unit, we met with the Chief-of-Service as well as the Nurse-in-Charge to discuss strategies for the adoption of the screening program.All the collaborating units recognized the feasibility of the proposed screening and referral program.

| Inclusion criteria
Patients who had completed primary cancer treatment within 2 years attending targeted outpatient surveillance clinics for follow-up.This study aimed to capture residual symptom distress in the early posttreatment phase and offer timely interventions.Ethical approval was obtained for this program (UW 18-496).

| Procedure
Within Outpatient oncology clinic, using clinic lists, a clinical assistant identified eligible patients who were then approached by a JCICC oncology-specialist nurse/case manager, to introduce the program and undergo screening.Eligible patients agreeing to join the program completed a standardized stepped symptom assessment administered by the nurse/case manager.

First, patients completed the modified Edmonton Symptom
Assessment System (ESAS-r) to screen for symptoms and any related distress. 19,20ESAS-r symptoms scores 0-3 are considered mild; 4-6 mild-to-moderate, while ≥7 indicate moderate-to-severe symptom distress. 21Second, for patients with ESAS-r symptom scores ≥7, the nurse administered a further symptom-specific assessment (Supplementary Material S1).Third, using a pre-defined triage system (Supplementary Material S2), the recommended clinical pathway directs the nurse which stepped strategy is optimal to manage the problematic symptom(s).Patients with mild and mild-to-moderate symptom distress were provided with self-management advice and, where needed, further symptom management and JCICC hotline contact in-

| Data collection
Eligible patients identified from clinic lists meeting the inclusion criteria were approached, and, if consenting, underwent SDS.Data were collected between August 2018 and June 2021.Data were collected on referral (Baseline), and at 3-months and 9-months thereafter (Post-baseline).

| Measures
ESAS-r assesses presence and intensity of 11 symptoms commonly encountered in routine cancer care 19,20 : pain, fatigue/tiredness, nausea, depression, anxiety, drowsiness, lack of appetite, well-being, shortness of breath, numbness, sleep difficulty, and "other (specify)".A numerical rating scale from 0 (no symptom) to 10 (the worst possible) measures each symptom's intensity, taking around 2 min to complete.
Any ESAS-r item scored ≥7 triggers a corresponding focused symptom assessment (Supplementary Material S1).

LAM ET AL.
Each patient also self-reported socio-demographic information and completed a patient-reported functional status (PRFS) tool, 22 comprising the activities/function domain of the Patient-Generated Subjective Global Assessment measure.Patients rated their own activities and function on a five-point scale, where 0 is "Normal, no limitations", 1 is "Not as normal, but able to be up and about with fairly normal activities", 2 is "not feeling up to most things, but in bed/ chair less than half the day", 3 is "able to do little activity and spend most of the day in bed or chair" and 4 is "pretty much bed-ridden, rarely out of bed".
ESAS-r case status also differed by gender, marital status, education, cancer site and occupation (Tables 1 and 3).Patients with breast cancer were more likely than patients with cancers of other sites to be cases, as were unemployed and divorced patients (Table 3).4).one fifth of all screened patients, all of whom received some form of intervention, mostly low-cost psycho-educational advice.The 80% acceptance of SDS was higher than other reported studies using ESAS; a 70% screening rate reported in an Ontario province-wide study, 23,24 77% reported in another Canadian study focusing on advanced cancer 25 and 69% reported in an Australia general hospital. 20Evidence suggests that implementation strategies could improve the screening rate to >90%. 267][28][29][30][31] Patients who were unemployed, divorced, had breast cancer, and reported poorer functional status were more likely to meet case criteria for symptom distress.

| DISCUSSION
JCICC's nurse-led symptom clinic has several unique components that warrant emphasis: a stepped-assessment approach using ESAS-r to screen and identify patients with symptom-related distress.Only those showing moderate-to-severe symptom distress then completed symptom-specific measures to more reliably determine the symptom intensity.Then, an assessment score-driven algorithm based on a stepped-care approach determined intervention: patients with mild and mild-to-moderate symptoms received self-management advice; those with moderate-to-severe symptom distress received recommendation and referral to community-based groups, and/or for more In contrast to other studies of psychosocial support services, which report low uptakes of 11%-30%, 6,30 our referral uptake was high at 69%.Our stepped-assessment approach allows symptom clinic nurses the opportunity to have a conversation about symptom distress and provide motivational coaching to encourage patients to recognize the benefits of supportive services.Also, beneficial is having community-based nurses case managers who deliver or coordinate all interventions.Previous studies showed highly-distressed patients often decline referral to mental health specialist-delivered psychosocial services 31,33,34 ; patients seemingly prefer interventions from nurses. 33We adopt a multi-disciplinary approach where a community nurse case manager coordinates a team of allied health professionals to deliver a broad spectrum of psychosocial services.This approach both minimizes the stigma of mental health or psychosocial services 31 and broadens potential solutions.
Referral to psychosocial support services was associated with observing significant declines in symptom distress over time.Being an uncontrolled study, we cannot attribute this unequivocally to interventions, our study does support the premise that SDS with appropriate support service referral probably helps in reducing symptom distress. 35Half of the cases received nurse-delivered selfmanagement-based psycho-education, via phone, suggesting that many patients with moderate-to-severe symptom distress can be effectively managed with modest resource investment. 19Nevertheless, 1-in-3 patients who were rescreened demonstrated persistent symptom distress, and 1-in-9 reported fluctuating symptom distress over time.Psychosocial distress seen in cancer can have multiple origins. 36Our findings reiterate the importance of focusing SDS to detect more severe cases.Stepped-assessment/care accommodates the criticism that routine screening in acute settings is resource intensive and inefficient as many patients have mild symptom distress. 36Early SDS in clinic patients who have recently completed cancer treatment to detect significant symptom distress with wellstructured optimal referral pathways to community-based services and continued monitoring, reduces staffing burden, patient distress and health expenditure.
Study limitations.First, as a prospective, single-arm implementation study that only addressed how well the screening program is working, it could not assess the specific effects of screening/care on symptom distress compared to usual care.Second, the a priori feasibility level was not set.However, the screening uptake rate was 80% with a 69% referral uptake rate (22% of all screened patients), demonstrating its feasibility.Third, ongoing hospital staffing shortages meant JCICC nursing staff ran the nurse-led symptom clinic embedded in public hospital outpatient oncology surveillance clinics.
While the current study demonstrates the feasibility and acceptability of in-clinic routine screening and community referral, future

of 11 -
Intervention background: Hong Kong's oncology care system lacks routine systematic symptom distress screening and management protocols.Currently, in patients receiving active cancer treatment, treatment-related symptoms are monitored and managed by oncology nurses who serve as case managers.However, once patients complete active cancer treatment, they are transferred to surveillance clinics where patients are seen periodically by oncologists to monitoring for cancer recurrence.No structured symptom management is given.Despite many cancer patients' continued experience of persistent or late effects from cancer treatments, no routine screening, referral or treatment is available to address patient needs.Locally, limited oncology psychosocial support services are predominately offered through community-based non-government organizations and sometimes, general in-hospital psychologists.Because broad demand for psychosocial support overwhelms available psychological services in tertiary care settings, community-based psychosocial support services addressing patient needs are a sensible alternative.However, Hong Kong currently lacks systematic referral pathways linking hospital and community oncology services.Consequently, psychosocial care support is not routinely offered to cancer patients.Patients who do access psychosocial support are likely to be the more resourceful and motivated.In contrast, patients with fewer resources lose out in what is effectively a self-referral service.In view of these shortcomings, the HKU-Jockey Club Institute of Cancer Care (JCICC), (established in 2018 with a donation from The Hong Kong Jockey Club Charities Trust) developed and is evaluating a nurse-led SDS program utilizing an evidence-based, structured stepped-assessment care model for routine cancer care involving recruitment at hospital out-patient clinics and service delivery at2 LAM ET AL.
formation.Patients with moderate-to-severe symptom distress and symptom-specific assessments were referred by the nurse to either group or individualized community-based JCICC symptommanagement programmes.All patients are offered symptom management leaflets.Patients accepting referral ("cases") are later contacted by the community-based JCICC nurses case managers who deliver and/or coordinate with other JCICC specialists.Based on the assessment scores, the triage system (Supplementary Material S2) refers patients to receive either self-management psycho-education (normally one face-to-face or phone consultation according to the patient's preference), group therapeutic intervention, or individualized targeted symptom-management input from nurses, clinical psychologists, social workers, dieticians, and/or exercise specialists, all current best evidence-based.Individual therapeutic interventions are personalized as far as possible according to the nature of the symptom distress and patient needs.The referring nurse remains the patient case manager, and carries out regular 3-monthly follow-up assessments.JCICC emphasizes self-management and avoids service duplication, sometimes referring to other providers, to ensure that patients receive timely interventions.This referral arrangement allows efficient collaboration between clinical institutions and other community service providers.
Standard descriptive analyses assessed sample characteristics, SDS acceptance rate, the ESAS-r (SDS) case prevalence, and the acceptance rate for community-based psychosocial support services.Ttests and Chi-squared tests assessed differences (1) between ESAS-r cases versus Non-ESAS-r cases, and (2) ESAS-r cases consenting to receive JCICC psychosocial support services versus those declining services.Logistic regression confirmed factors associated with ESASr case status.To test if symptom distress declined following psychosocial support referral, Paired-sample t-test compared ESAS-r score changes from baseline to follow-up assessments.Repeated Measures General Linear Modeling assessed changes in ESAS-r scores over time.Multinomial regression examined factors that differentiated changed ESAS-r score across the three time-points.SPSS software was used for statistical analysis.All proportions are rounded to the nearest whole number.
Of 673 patients consenting to referral, 259 (9% of screened, 38% of referred) received JCICC-delivered individualized therapeutic interventions; 54, (2% of screened, 8% of referred) received group interventions delivered by JCICC or other community provider, and 360, (12% of screened, 53% of referred) received phone-based psycho-education only for symptom self-management, delivered by the nurse case manager.Hence, one-in-five of all screened, and all of the referred patients received some form of intervention.Patients were reassessed at 3-months and 9-months post-baseline: 445/673 (

A
nurse-led symptom clinic embedded in outpatient oncology surveillance clinics with available community-based psychosocial interventions appears feasible.This prospective, single-arm, pragmatic implementation study revealed that in a Hong Kong Chinese population, among patients who have recently completed cancer treatment 80% of those approached were receptive to SDS and of those screened, two-thirds showed no significant distress.Sixty-nine percent of the remaining distress-symptomatic patients accepted onward referral to community-based psychosocial support services, studies need implementation strategies to incentivize clinical partners to deploy clinic staff to run the nurse-led symptom clinic.We are currently conducting a trial intervention to test if implementation strategies including training, audit, feedback, facilitation, and adaptable workflow will increase screening and referral compared to usual control under which no implementation strategies will be used to facilitate the adoption of the systematic symptom distress screening and referral.Fourth, about 40% of cases did not rescreen at 3-and 9months post-baseline screening.Patients who did not rescreen had higher initial SDS scores and were more likely to receive an individualized therapeutic intervention.We did not collect details of the individualized therapeutic intervention received by the patients.Moreover, 31% of positive screened patients declined referral.Understanding why patients may not engage even when symptomatic, or respond to follow-up warrants further investigation.Clinical implications.While some countries have included SDS in their national healthcare plans, implementation in clinical settings poses significant challenges.Our findings suggest that a nurse-led symptom clinic embedded in outpatient oncology surveillance clinics, supported by community-based services, appears feasible for post-treatment patients.In particular, a comprehensive screening program requires not only a tailored symptom-distress screen but also a well-designed and optimizing referral pathway to connect patients to adequately resourced, effective interventions within the healthcare system.Integrating this nurse-led symptom clinic into clinical settings may help streamline the monitoring and management process, ensuring that patients receive comprehensive care within a familiar healthcare setting.LAM ET AL.
Comparison of patients consenting to versus patients declining support service.

Predictors Persistent ESAS cases Recurrent ESAS cases Delayed-recovery ESAS cases Odds ratio (95% CI) SE p-value Odds ratio (95% CI) SE p-value Odds ratio (95% CI) SE p-value
Logistic regression of factors associated with ESAS cases (Cases as referent).Functional status: 1, Normal with no limitations; 2, Not my normal self, but able to be up and about with fairly normal activities; 3, Not feeling up to most things, but in bed or chair less than half the day/Able to do little activity and spend most of the day in bed or chair/Pretty much bedridden, rarely out of bed.
Note: Functional status: 1, Normal with no limitations; 2, Not my normal self, but able to be up and about with fairly normal activities; 3, Not feeling up to most things, but in bed or chair less than half the day/Able to do little activity and spend most of the day in bed or chair/Pretty much bedridden, rarely out of bed.Abbreviations: NS, Not significant; SD, Standard deviation.LAM ET AL.T A B L E 3Abbreviations: NS, Not significant; SE, Standard error.T A B L E 4Multinomial logistic regression of predictors on ESAS trajectory status (Recovery as referent).