Introduction

ASCO guidelines for Exercise, Diet, and Weight Management During Cancer Treatment state that clinicians should refer their patients to exercise during chemotherapy to address common symptoms and side effects [1]. The benefits of exercise for people living with and beyond cancer are well described and include improvements in physical function, body composition, fatigue, anxiety, depression, sleep, bone health, and quality of life [1,2,3]. However, published evidence, also from ASCO, suggests that only 15% of patients recall being referred to exercise by their oncologist [4]. People living with and beyond cancer are less likely to be active than their healthy counterparts and published evidence suggests that advice from an oncology clinician results in more physical activity [5].

We know that exercise has benefits when patients participate in exercise oncology programming. What we know less about is how to make referrals to exercise oncology programming. Taken together, new methods are needed for implementing exercise referrals. One complication of this is the recognition that patients vary as to the level of care needed to successfully participate in appropriate exercise programming [6]. Some patients could exercise on their own, some have sufficient impairments that professional guidance would be advisable, and for yet others, there are documentable impairments for which outpatient rehabilitation would be the best course of action [7]. A system is needed to triage patients that can be broadly implemented in oncology care. This is a cancer care delivery problem that can be addressed through research [8].

The Exercise in Cancer Evaluation and Decision Support (EXCEEDS) triage tool was developed to support patients and clinician’s decision-making to participate in or refer to appropriate exercise oncology or rehabilitation services [9]. The tool has been validated using a Delphi process [9] including high satisfaction among oncology clinicians [10]. We undertook a quality improvement (QI) project to make triage and referral to exercise oncology standard of care, by embedding the EXCEEDS tool into clinical care. Goals of the QI initiative were to evaluate whether integrating a rehabilitation navigator and applying the EXCEEDS tool would be acceptable and feasible from the perspective of patients and clinical staff, as well as to assess the extent to which this approach would improve the proportion of patients referred to appropriate exercise or rehabilitation programming. The project was guided by the EPIS framework [11] (Exploration, Preparation, Implementation, and Sustainment) and prior efforts to map exercise oncology implementation strategies onto a compendium of implementation strategies [12].

Methods

This project was reviewed by the QI committee of the University of Pittsburgh Medical Center (UPMC) and deemed not to be human subjects research, but to be a QI initiative, according to review of the common rule (45 CFR 46.102(d)). SQUIRE 2.0 guidelines were followed [13]. So that we could build on the learnings of this QI project, we aligned the methods and measures used with the approaches promoted by the science of implementation, especially as noted in the cancer prevention and control field [14]. Through this initiative, our goals were to generate practice-based evidence to inform future research and practice-changing activities.

Implementation framework

The EPIS framework guided this QI initiative. The EPIS framework highlights key phases that guide and describe the implementation process and clarifies common and unique factors within and across levels of system and organizational context across phases, factors that bridge the variety of contexts specific to implementation, and the nature of the innovation being implemented, as well as the role of the innovators [11]. There are four phases to the EPIS Framework: Exploration, Preparation, Implementation, and Sustainment. In this project, we accomplished three of these phases; sustainment remains to be studied. After 5 months, we expanded our program to the 3rd floor of the Hillman Cancer Center. All implementation strategies were repeated for this expansion.

Triage tool

We chose to use the EXCEEDS tool [9] for our triage and referral program. EXCEEDS has 23 questions, in 3 sections, including a section regarding difficulty completing daily activities (positive responses lead to rehabilitation services referrals); a section about recent falls, recent cancer treatments, and recent symptoms (e.g., fatigue, neuropathy, memory, dizziness, nausea, lymphedema), and a section about catheters, current exercise, and confidence with exercise. It was developed with the help of a multidisciplinary team of experts and has been documented to be acceptable to oncology clinicians [10]. The EXCEEDS tool is intended to result in referral to four possible levels of intervention: unsupervised exercise, supervised cancer-specific community-based exercise, clinically supervised exercise, and cancer rehabilitation. EXCEEDS was designed to be adaptable to the needs of local systems. As such, based on available programming at our site, we collapsed into three levels of intervention: community-based exercise oncology programming, clinically supervised exercise, and cancer rehabilitation. The instrument was loaded into a RedCAP® database and delivered to patients on a computer tablet.

Flow of clinical activities

A rehabilitation navigator identified patients coming in for their second chemotherapy infusion visit at Hillman Cancer Center. The clinical encounters reported on herein occurred between March 21 and October 6, 2023. A list of patients was provided to the medical assistants. At the point of seating the patient in their chemotherapy chair, the medical assistants provided an iPad to the patient and asked that they complete the survey, which was provided in RedCAP® [15], [16]. It took between 2 and 5 min per patient for medical assistants to complete the task of providing the iPad and explaining the survey. Upon completion, the scored results became immediately available to the rehabilitation navigator, who approached the patient for a brief counseling session to discuss the results and provide the referral. These counseling sessions were usually 5 to 10 min long but could be as long as 90 min. Patients had the option to accept or deny any recommendation made.

Program staff

The Moving Through Cancer Triage Program staff includes a licensed physical therapist with specialty training in Exercise Oncology (the rehabilitation navigator) and five Medical Assistants, who provided the iPads to the patients at the start of the second infusion visit.

Programs offered

When a patient was identified as having the symptom and disease profile consistent with referral to community-based exercise programming (e.g., minimal symptoms or comorbidities), we offered those who were UPMC health plan members the option to connect with the UPMC Prescription for Wellness (https://www.upmcmyhealthmatters.com/prescription-for-wellness-for-upmc-health-plan-members/) a customized care management system for healthy lifestyle changes. Other options offered included local cancer exercise offerings (Cancer Bridges: https://cancerbridges.org/), or a link to online programming available through the Moving Through Cancer Directory (https://www.exerciseismedicine.org/eim-in-action/moving-through-cancer/). After 4 months during which acceptance of these options was close to zero, we began to offer these patients the option for a few sessions of clinically supervised exercise with the rehabilitation navigator. These sessions provided instruction that could be carried out within unsupervised home-based exercise sessions. There are ongoing and completed exercise oncology interventions that have offered brief in person counseling that have documented benefits [17,18,19]. Patients had the option to refuse this offering as well.

When patients had sufficient symptoms to warrant more attention, but insufficient for referral to outpatient rehabilitation, the rehabilitation navigator offered clinically supervised exercise sessions. These sessions could occur prior to the infusion visit or at a separate time, and could occur chair side, in the chemotherapy infusion center, in the 200 square-foot gym on the second floor of Hillman Cancer Center, or virtually.

If a patient identified sufficient impairments that the EXCEEDS triage tool suggested referral to outpatient rehabilitation, the rehabilitation navigator offered this referral. If the referral was accepted by the patient, the rehabilitation navigator facilitated the referral by providing supporting materials to the appropriate nursing staff. If patients refused the referral, they were offered clinically supervised exercise sessions with the rehabilitation navigator.

Statistical analysis

All analyses were conducted in R (version 4.3.2). Descriptive statistics were developed for tables. Acceptability was defined as the percentage of patients who were willing to engage in the triage process. Feasibility was defined as the percentage of triaged patients who accepted a referral. Our a priori threshold for establishing feasibility was 30% (double the published background rate) [4]. Differences between those who accepted the triage process and those who did not were examined using t-tests and chi-square tests. Differences in exercise confidence across EXCEEDS triage tool levels were examined using a chi-square test.

Results

Implementation results

The exploration phase of implementation focused on building the coalition for this program with key administrators, decision makers, implementers, and leadership champions. In this setting, iterative conversations with staff and leadership led to meetings with the cancer center director, the CEO of the cancer center, the Division Chief of Hematology Oncology, the Senior Director of Hillman Operations, 2nd floor Hillman Nursing Unit Director, Nursing Lead for Treatment Side, 2nd floor Hillman, and the managers for the Medical Assistants and Collaborative Nurses on the 2nd floor of Hillman. Similar meetings occurred when we expanded the program to the 3rd floor clinic. This phase of implementation took approximately 3 months. In the preparation phase, we made presentations in the settings of medical oncology grand rounds and during multiple meetings of nursing leadership and staff. We obtained formal commitment for the project from senior nursing staff. We developed educational training materials that were shared at all meetings and conducted numerous educational outreach visits within the infusion clinic. We further gained the buy-in of nursing staff by demonstrating adaptability to their workflow. We were interested in finding a time during the 2nd infusion visit encounter that had the lowest likelihood of interfering with the flow of clinic activities. The champions chose the interactions of medical assistants with patients as they are being positioned in the chemotherapy chair as the best possible time point. This phase of implementation lasted approximately 6 weeks.

For our implementation phase, we continued regular meetings with appropriate nursing staff to report on our progress. An audit and feedback process with our nursing leadership champions led to increased likelihood of medical assistants complying with the new clinical workflow. The principal investigator (an exercise physiologist) met with the director of operations, who then conveyed the audit and feedback information to nursing leadership. The first occasion of audit and feedback was a phone call discussion; two subsequent occasions were email exchanges. We intervened with medical assistants in another educational meeting when there was a drop in implementation. At present, all 5 participating medical assistants are regularly getting the iPad into the hands of patients at the second infusion visit. The intervention was delivered with fidelity to the planned implementation of this triage and referral process with 100% of medical assistants complying with the necessary actions over the entire 6 month observation period reported herein.

Figure 1 shows the flow of participants through the triage and referral process. A total of 587 patients were seen for a second infusion visit during the time period of this QI project. Of these, 501 (85.3%) were offered the triage survey by medical assistants. Of these, 78% accepted the triage survey. A total of 391 patients completed the survey and, of these, 176 (45%) were connected to an exercise or rehabilitation intervention (feasibility). The proportion of patients for whom responses to the EXCEEDS triage survey suggested that a community-based exercise intervention was appropriate was 27.9%. Of these, 9.2% accepted the referral. Given the low uptake of this referral, we added the option of a consultation with the rehabilitation navigator after a few months. Of the 78 patients offered this option, 12 (15%) accepted the referral. The proportion of patients triaged to clinical exercise supervision was 18.7%. Of these patients, 37% agreed to the referral. The proportion of triaged patients for whom responses to the EXCEEDS triage tool suggested that it would be appropriate to refer to cancer rehabilitation was 53.7%. Of these, 23.3% accepted this referral. Of those who did not accept the referral to cancer rehabilitation, we offered clinical supervised exercise. Of the 161 offered this option, 78 or 48.4% accepted the referral.

Fig. 1
figure 1

Flow of participants

Full size image

For the remaining 127 patients, a total of 173 total consultations and full exercise sessions were carried out. Most of the patients received 1–3 exercise sessions or consultations (N = 108). Nine patients received between 4 and 6 sessions.

Table 1 presents a description of the overall patients approached for the triage and referral system, as well as those who agreed versus those who refused the triage instrument. There were no significant statistical differences between the patients who accepted versus those who did not accept the triage instrument.

Table 1 Description of patients
Full size table

Table 2 presents the results of the EXCEEDS triage process overall, as well as by the final category in which the patient was placed. Of the 391 respondents, 25% reported doing aerobic activity three times in the past week and 14% reported doing muscle-strengthening activities in the past week. In this group of respondents, 36% reported being highly confident they would exercise regularly without support from an exercise professional. Exercise confidence was associated with EXCEEDS triage category, with greater confidence among those whose answers indicated appropriate referral to community-based exercise as compared to those who reported symptoms consistent with a referral to rehabilitation services (χ2 = 42.96, p < 0.0001). There was a similar significant association between engaging in resistance exercise (yes/no) and EXCEEDS category (χ2 = 12.93, p = 0.002). There was also a significant association between engagement in aerobic exercise (yes/no) and EXCEEDS, such that those with fewer impairments reported greater confidence (χ2 = 32.35, p < 0.0001).

Table 2 EXCEEDS responses (N = 391)
Full size table

Discussion

Recent ASCO guidelines indicate that medical oncologists should refer patients to exercise to address symptoms and side effects [1]. A major challenge to achieving this goal lies in the need to triage patients to the appropriate kind of exercise or rehabilitation programming [6]. This process requires time and expertise that may be beyond the scope of medical oncologists. Oncologists self-report that they agree that patients should be referred to appropriate exercise programming, but that they do not have the time or training to do so [20]. As a result, 15% of patients report recalling being referred to an exercise program by their oncologist [4]. Systems are needed that include the buy-in of oncologists but require minimal time and effort. We implemented a clinical workflow to address this cancer care delivery challenge. This system is acceptable to clinicians (100% of medical assistants participated) and patients (78% of infusion patients completed the triage survey). The system appears to be feasible as well, given that 45% the patients who completed the triage survey and who were offered an intervention accepted the offer. The success of this approach, from the perspective of patients, nursing leadership, and oncology clinicians, suggests that there may be value to broader implementation of this approach at other cancer centers.

Implementation of this triage and referral process was accomplished using well-documented approaches suggested by a recent review of exercise oncology implementation [12] and followed the EPIS framework [11]. Buy-in from clinical leadership, including physicians and nurses, was crucial, as was informing the staff, adaptability with regard to fitting into the clinical flow, and ensuring that all medical assistant staff were comfortable with the new processes before starting. Audit and feedback processes enabled nursing management to understand which medical assistants needed to be reminded of the new procedures and resulted in 85% of infusion patients being offered the triage instrument. Acceptability is underscored by the seamless expansion of the program to an additional floor of Hillman Cancer Center after 5 months of the QI project being in operation.

We observed that a low percentage of patients triaged to community-based exercise accepted any referrals (15%). One hypothesized explanation for this would be that these patients did not value the referral because they saw no problem to be fixed. The rate of acceptance for the other two categories is consistent with this hypothesis. A comparatively higher percentage of patients triaged to clinically supervised exercise accepted the referral (35%). Further, 62% of patients triaged to rehabilitation accepted either a referral to rehabilitation services or clinically supervised exercise. That said, only 23.3% of patients triaged to rehabilitative services accepted the referral. The reasons for refusing the referral included cost, time, distance, prior negative experiences with physical therapy, or unwillingness to add additional appointments while going through chemotherapy. The patients who refused rehabilitation services were offered onsite clinically supervised exercise and 48.4% accepted this referral. Hypotheses for why the patient acceptance rate was so much higher for the onsite clinically supervised exercise include the co-location of the program, the lack of cost (the program is provided to patients without cost), and the convenience regarding timing with infusion visits. The high acceptance of this new clinical workflow by the clinicians and clinical staff is hypothesized to be the result of our facilitation efforts, coalition building activities, technical support, and adaptability to the clinical workflow. These observations may be useful for others interested in implementing similar triage programs with regards to the proportion of patients likely to be referred to community-based programming, clinically supervised exercise, and rehabilitation services. It also validates our prior published documentation of the value of co-location of exercise oncology and rehabilitation services with infusion therapy [21].

One prior study used a shorter “Ask Advise Refer” approach to triage and referral of 1174 patients [22]. Patients identified as being inadequately active were further evaluated to discern whether their performance status suggested value of a referral to rehabilitation services (physical or occupational therapy). A total of 540 patients were evaluated for performance status, 168 were referred to rehabilitation services. Of these, 13 (8%) accepted and completed the referral. Multiple differences between this prior approach and the current approach explain differences. First, My Wellness Check was integrated into the electronic medical record. We are working toward this goal at Hillman Cancer Center. Further, unlike our Hillman Cancer Center program, the My Wellness Check project did not make any referrals to exercise programming. Finally, unlike the My Wellness Check project, we integrated a rehabilitation navigator to talk the patients through the results of the triage tool and personally explain the referral.

Limitations of our study include that we were unable to follow-up on the 49 patients who accepted referral to rehabilitation services to discern whether they were connected to those services because services outside of our health system were chosen. Further, the focus of this project was to evaluate implementation of triage and referral, and therefore, we are unable to comment on the effectiveness of the exercise and rehabilitation interventions.

We have documented high acceptability and feasibility of the implementation of a new clinical workflow into a busy academic cancer center, in two chemotherapy infusion clinics, as well as excellent implementation of the triage process by medical staff (85%), acceptability of the triage process by patients (78%) and feasibility of referrals (45%). This process holds promise as an approach to making exercise oncology and rehabilitation referrals standard of practice in the setting of medical oncology. Making exercise and rehabilitation triage and referral standard of care will assist with addressing documented disparities in physical function, quality of life, and mental health, and perhaps survival, by race, ethnicity, geography (urban versus rural), and socioeconomic status [23,24,25,26].