Development and Implementation of an Interdisciplinary Telemedicine Clinic for Older Patients with Cancer- Preliminary Data

Abstract

BACKGROUND:

Frailty assessment is an important marker of the older adult’s fitness for cancer treatment independent of age. Pretreatment geriatric assessment (GA) is associated with improved mortality and morbidity outcomes but must occur in a time sensitive manner to be useful for cancer treatment decision making. Unfortunately, time, resources and other constraints make GA difficult to perform in busy oncology clinics. We developed the Cancer and Aging Interdisciplinary Team (CAIT) clinic model to provide timely GA and treatment recommendations independent of patient’s physical location.

METHODS:

The interdisciplinary CAIT clinic model was developed utilizing the surge in telemedicine during the Covid-19 pandemic. The core team consists of the patient’s oncologist, geriatrician, registered nurse, pharmacist, and registered dietitian. The clinic’s format is flexible, and the various assessments can be asynchronous. Patients choose the service method- in person, remotely, or hybrid. Based on GA outcomes, the geriatrician provides recommendations and arrange interventions. An assessment summary including life expectancy estimates and chemotoxicity risk calculator scores is conveyed to and discussed with the treating oncologist. Physician and patient satisfaction were assessed.

RESULTS:

Between May 2021 and June 2022, 50 patients from multiple physical locations were evaluated in the CAIT clinic. 68% were 80 years of age or older (range 67–99). All the evaluations were hybrid. The median days between receiving a referral and having the appointment was 8. GA detected multiple unidentified impairments. About half of the patients (52%) went on to receive chemotherapy (24% standard dose, 28% with dose modifications). The rest received radiation (20%), immune (12%) or hormonal (4%) therapies, 2% underwent surgery, 2% chose alternative medicine, 8% were placed under observation, and 6% enrolled in hospice care. Feedback was extremely positive.

CONCLUSIONS:

The successful development of the CAIT clinic model provides strong support for the potential dissemination across services and institutions.

INTRODUCTION

The care of older adults with cancer is complex. In addition to treatment side effects, there are a myriad of age-related potential causes of cancer treatment complications, and often, unexpected changes in the patient’s quality of life. The need for optimal delivery of cancer care is increasing as older adults constitute the fastest growing demographic in the US population.¹ The development of successful models of care must confront multiple implementation barriers such as limited time and space, and/or lack of geriatrics-trained health care providers.² Despite these challenges, several models of care in Geriatric Oncology have been published.³

Frailty assessment has become an important marker of the older adult’s fitness for cancer treatment independent of their age. The American Society of Clinical Oncology (ASCO),⁴ the International Society of Geriatric Oncology (SIOG),⁵ and the National Comprehensive Cancer Network (NCCN),⁶ recommend assessment of fitness by including Geriatric Assessment (GA) into the evaluation and care planning for older adults with cancer. These assessments must take place in a time sensitive manner to be useful for cancer treatment decision-making and avoid over/under treatment. A multicenter randomized controlled trial (RCT) showed that GA based interventions lead to better quality of life in older people receiving systemic anticancer treatment.⁷ Another RCT of older adults with hematological malignancies showed that geriatrician consultation embedded in the oncology clinic improved the odds of having end-of-life and goals-of-care discussions and better management of several geriatric syndromes.⁸ Multidisciplinary GA models and geriatric co-management in oncology and other specialties have received increased attention in recent years and are associated with improved mortality and morbidity outcomes.^9,10,11 Unfortunately, time and other constraints make GA difficult to perform in busy oncology clinics.¹²

The disruption in the traditional provision of clinical care brought on by the Covid-19 pandemic led to an unprecedented surge in telemedicine visits. Patients and healthcare providers report high satisfaction with telehealth visits.¹³ This context provided the opportunity to develop an alternative model of a geriatric oncology interdisciplinary clinic at our cancer center. Using the Theory of Change,¹⁴ we developed a model of geriatric oncology care to provide timely GA, accounting for institutional factors. In this paper we will describe how we developed the framework for the Cancer and Aging Interdisciplinary Team (CAIT) clinic, the facilitators, the barriers we faced and its potential for replication. We will discuss how this model addressed some of the ubiquitous chronic roadblocks to the development of such an interdisciplinary clinic, adding substantial feasibility to previous models of GA delivery.

METHODS

Memorial Sloan Kettering Cancer Center (MSK) is an academic cancer treatment and research institution. Its network spans a wide geographic range across two states (New York and New Jersey) and, in 2021 alone, over 78,000 patients 65 and older received cancer care. The Geriatrics Service at MSK was established within the Department of Medicine in 2009.¹⁵ It provides inpatient and outpatient care, consultations for GA, preoperative surgical risk assessment, treatment of geriatric syndromes, patient co-management, and long term follow up care. Despite these extensive services, the department has not been able to offer a structured, cohesive, interdisciplinary clinic that would provide timely, effective, and efficient GA with individualized recommendations to older patients scheduled to start medical oncology treatment. Notable barriers to the development of such a clinic have been the geographic fragmentation of the cancer center’s clinical sites and the lack of available physical space to house an interdisciplinary team within each site. The oncology specialties practice in different locations from the Geriatric outpatient clinic space. In addition, MSK has several regional centers outside New York City where physical distances and staff limitations have prevented a Geriatrics presence. The increased availability of telemedicine provided an opportunity to overcome these barriers and develop and implement the CAIT clinic- a hybrid clinic model that included in-person and/or telehealth visits. The study has been approved by MSK institution review board committee as exempt.

Development of the CAIT clinic model

MSK is a member of the Institute for Healthcare Improvement’s Leadership Alliance. In 2020–2021 MSK participated in a workgroup aimed at redesigning healthcare delivery systems called “Office Visit is a Dinosaur (OVID)”. In this space, the CAIT clinic model was developed using the “Theory of Change”. The theory provides a comprehensive description of how and why a desired change is expected to happen in a particular context. The first step to applying the theory is to define the desired long-term goals and then work back from these goals, identifying the conditions that must be in place for the goals to be achieved. A strength of the theory is its focus on “filling in” what has been described as the “missing middle” between what a program does (activities or interventions) and how these lead to desired goals being achieved.¹⁴

We first set up our goal to provide interdisciplinary GA, treatment recommendations, and optimization of deficits to older patients with cancer about to start cancer treatment in a timely manner and independent of their physical location. A driver diagram was developed by “backwards mapping” based on our assumptions as to what would be required to achieve the goal. (Fig 1) The diagram listed primary and secondary drivers of goal achievement. Concrete actionable change ideas were developed from the drivers, followed by the identification of the activities and interventions needed to achieve the intermediate objectives.

Figure 1.

Driver Diagram

The goal of the Cancer and Aging Interdisciplinary Team clinic, and the reverse mapping identifying primary and secondary drivers that would lead to achieving the goal

Team:

The core Interdisciplinary team (IDT) consists of the patient’s Oncologist, a Geriatrician, RN, geriatric pharmacist, and registered dietitian (RD). Physical (PT) and Occupational therapy (OT), Social Worker (SW) and Psychology services are included on an as-needed basis. Members of the team were identified through well-established collaborations between the GS and other health care providers.

Format of service delivery:

The clinic can be delivered as either:

1. Telemedicine only- where each of the clinicians and patient/caregivers are in different locations.

2. Hybrid - where the patient/caregiver is located at a physical clinical location (oncology or geriatrics clinic) for an in-person visit. All other clinicians join remotely from other locations.

Clinic Structure and Workflow:

The first touchpoint for most patients considered for the CAIT clinic is the medical oncology clinic (or surgical oncology if patient is being considered for neo-adjuvant chemotherapy) (Fig 2). Triggers for referral were age ≥ 65, with any of the following: oncologists’ impression of frailty, multimorbidity and/or cognitive impairment. Strict criteria were not established in this initial phase to facilitate the development of the model. When the patient is scheduled for the CAIT clinic, all clinicians on the team are notified and the patient receives an online questionnaire through the patient portal called the electronic Rapid Fitness Assessment (eRFA),¹⁶ a patient-reported, electronic version of the GA developed by the MSK GS that has been used since 2015. Patients without computer access complete the eRFA on an iPad during an in-person CAIT clinic visit or at an MSK regional center with assistance from staff when needed. Only one domain of the eRFA (the Timed-up-and go test) needed to be adapted to allow for a remote functional performance evaluation. It was substituted by the 30 second chair-stand test which can be administered remotely.¹⁷

Figure 2.

Cancer and Aging Interdisciplinary Team (CAIT) clinic workflow

The initiation of the Cancer and Aging Interdisciplinary Team clinic consult, the process of the clinical evaluations permitting asynchronous visits, compilation of information and discussion between the referring oncologist and geriatrician to arrive at a treatment decision

*- electronic Rapid Fitness Assessment

Each team member completes a discipline-specific assessment (Table 1), the results of which are compiled by the Geriatrician. Based on deficits identified, occupational and/or physical therapy referrals, social work referral, nutritional supplements, falls prevention measures, and other geriatric interventions are implemented to correct potentially reversible deficits and optimize chronic diseases. A summary of the recommendations and interventions is conveyed to the treating oncologist together with life expectancy estimates¹⁸ and results obtained from the chemotherapy toxicity risk calculator.¹⁹ The goal of this communication is to trigger a discussion and to arrive at a treatment plan that incorporates not only oncological best practices but also patients’ goals, needs, and ability to tolerate treatment. A 3-month follow up is scheduled with the Geriatrics Service. Earlier or additional follow up visits are left to the discretion of the Geriatrician.

Table 1.

CAIT CLINICAL ASSESSMENTS

Clinician	Domain/Disease assessed	Tools	Intervention when appropriate
Geriatric Pharmacist	Medications	Medication Appropriateness Index36 Lexicomp37 Beer’s Criteria23	Discontinuation, changes, or addition. Identification of significant interactions and red flags for non-adherence.
Registered Dietitian	Nutrition	Mini Nutritional Assessment38	Dietary modifications. Optimization strategies Nutritional Supplements.
Registered Nurse	Patient reported GA items	eRFA*16	Identify gaps in the reporting and complete. Social service referral. Initiate home care/nursing services.
Clinical Psychologist	Distress Mood, support		Coping strategies. Psychological support for patients and caregivers.
Geriatrician	Function	ADLs*39 IADLs*40	Identify source of support. Optimize adequacy of compensations.
	Cognition	MoCA41 MiniCog42	Strategies for compensation. B12, Folate, TSH levels. Neuropsychological testing referral. Caregiver education.
	Performance Measures	TUG*43 30 second chair stand17	Referral to Occupational and/or Physical therapy.
	Medical Comorbidities		Testing. Optimization. Referrals.
	Calculators:
	Chemotoxicity	CARG* score19 CARG-BC* score44
	Life expectancy	ePrognosis18‚45
	Surgical risk	NSQIP* surgical risk calculator46 (If planned for neoadjuvant chemotherapy)
Oncologist	Cancer diagnosis and treatment		Identify treatment plan based on standard of care, GA recommendations and patient goals.

^*eRFA= electronic rapid fitness assessment, ADL= activities of daily living, IADL= instrumental activities of daily living, TUG= timed up and go, CARG= cancer and aging research group, CARG-BC= cancer and aging research group- Breast Cancer, NSQIP= National surgical quality improvement program

The clinic format permits a flexible structure. Usual visit components (e.g., vitals check, blood draws) are unbundled, and the various clinical assessments can be asynchronous, reducing the time needed for clinic visits. The eRFA can be done by the patients from home prior to the visit through the patient portal. The pharmacist, RD and other clinical assessments can occur at any time over a span of one week based on the patient’s and clinicians’ availability.

Technology innovation and support:

We initially used existing online services such as Doximity^™ or Cisco Jabber^™. As telehealth became more established, MSK’s Information Technology (IT) team developed an in-house telemedicine software that better integrated with the existing electronic medical records. After a few iterations, the current platform we use is a Microsoft^™ Teams application and can be accessed by the patients through the MSK patient portal.

Implementation Timeline:

The basic framework for the CAIT clinic model and initial workflow was launched in January 2021. The flow was designed in a collaborative effort with representation from the Geriatric Service, Medical Oncology, Nursing, hospital operations, administration, and IT. The patient encounters were initially synchronous, scheduled once a week. The initial collaborating oncologist was from the Lymphoma Service with clinic on the same day as the geriatrician, so patients would be transferred from one provider to the other in a virtual telemedicine room.

Three Plan-Do-Study-Act (PDSA) cycles were utilized to identify and correct workflow issues.²⁰

1. The original workflow called for the oncologist and geriatrician to meet on the audiovisual platform at the end of the evaluation, to discuss findings and develop a care plan. It proved difficult for clinicians to meet given their need to continue seeing patients after assessing the CAIT clinic patient. Therefore, the process was changed, and the Geriatrician would send a summary to the Oncologist via email and discuss the best time and method to further communicate.

2. A second collaborating oncologist from the Gynecology Service referred patients to the CAIT clinic. The clinic schedule of this provider and the geriatrics service did not overlap so patients could not be seen by both providers on the same day. To address this issue, the CAIT clinics were changed to asynchronous. The visits were allowed to be scheduled on any clinic day of the four geriatricians to allow for maximum flexibility. The model was then opened to all GYN medical oncologists.

3. The third cycle resulted in the addition of a Geriatrics follow up visit 3 months later, to reevaluate patient’s needs. We also realized that for patients who were planned for neoadjuvant chemotherapy, the surgeons were their first touch point. To facilitate timely consultations, the GYN surgeons were added to the list of referring providers.

In line with the Theory of Change framework, we developed indicators to measure the effectiveness of the model. One performance indicator was the timeliness of scheduling and conducting the CAIT clinic evaluation; the benchmark for success is that the CAIT clinic evaluation was completed prior to the next oncology visit achieving the goal of providing the oncologist with the opportunity to include this crucial information in the treatment decision-making process. Another performance indicator was patient and clinician satisfaction.²¹ An Oncologist Satisfaction Survey was sent once to each referring oncologist by email; a Patient Satisfaction Survey was sent through the patient portal following completion of the CAIT evaluation.

We then opened the CAIT clinic in a stepwise manner to several MSK services located in and outside of Manhattan. We identified key stakeholders and champions from the middle management and upper leadership. Top management support is considered one of the critical success factors in project management. It facilitates easier spread of information regarding the model to potential participants and quicker adoption.²² The project was initially presented to nursing as well as the Department of Medicine leadership to introduce the model and obtain their buy-in. Educational presentations on the importance of GA in this patient population were made at individual divisional service meetings. We worked with each division’s champion to develop specific referral criteria based on characteristics common to the cancer type. This stepwise expansion allowed the Geriatrics Service to accommodate increased referral volume while tailoring the CAIT clinic to disease-specific needs.

RESULTS

Between May 2021 and June 2022, 50 patients were evaluated in the CAIT clinic. Patient characteristics are shown in Table 2. The majority (68%) were 80 years of age or older (range 67–99 years). All patients were evaluated using a hybrid model. Patients were referred from the Gynecology (26%), Gastrointestinal (20%), Breast (16%) and other services (38%). The median days between receiving a referral and the patient being seen by the Geriatrician was 8 days (quartiles 5, 13).

Table 2:

Patient Demographics And Other Characteristics (N=50)

Variable	Value
Age, median (Interquartile range, IQR)	83 years (76, 89)
Female, n (%)	34 (68%)
Race, n (%)
White	42 (84%)
Black	3 (6.0%)
Asian	1 (2.0%)
Other	2 (4.0%)
Missing	2 (4.0%)
Education Status, n (%)
Less than high school diploma	1 (2.0%)
High school diploma	6 (12%)
Some college	7 (14%)
College graduate	10 (20%)
Advanced degree	13 (26%)
Missing	13 (26%)
Marital Status, n (%)
Not Partnered	28 (56%)
Partnered	22 (44%)
Living Conditions, n (%)
Alone	14 (28%)
Living with Family	33 (66%)
Living with 24/7 Aide	3 (6.0%)
Residence, n (%)
New York City (NYC)	13 (26%)
NY State, non-NYC	21 (42%)
NJ State	10 (20%)
Other	6 (12%)
Type of Cancer, n (%)
Breast	8 (16%)
Gastrointestinal	10 (20%)
Gynecological	13 (26%)
Head and Neck	3 (6.0%)
Hematological	5 (10%)
Hepato-pancreato-biliary	4 (8.0%)
Melanoma	1 (2.0%)
Urological	6 (12%)
Metastatic Cancer (among non-Hematological cancers), N=45, n (%)	22 (49%)
5-year mortality risk, N=43, Median (IQR)	50% (37%, 70%)
Cancer and Aging Research Group (CARG) risk score, N=38, n (%)
Low (0–5)	2 (5.3%)
Medium (6–9)	8 (21%)
High (10–19)	28 (74%)
Montreal Cognitive Assessment (MoCA), n (%)
Normal (26–30)	16 (32%)
Mild Cognitive Impairment (18–25)	18 (36%)
Moderate Cognitive Impairment (10–17)	4 (8.0%)
Severe Cognitive Impairment (0–9)	3 (6.0%)
Attempted, but unable to complete	2 (4.0%)
Missing	7 (14%)
Abnormal 30-second Chair Stand, N=34, n (%)	18 (53%)
Mini nutritional assessment (MNA), n (%)
Normal nutrition status (12–14 points)	24 (48%)
At risk of malnutrition (8–11 points)	18 (36%)
Malnourished (0–7 points)	8 (16%)
Polypharmacy (≥ 10 medications), n (%)	30 (60%)

GA data obtained through the patient reported components of the eRFA are shown in Figure 3. Common medical comorbidities identified using ICD-9 and ICD-10 codes were coronary artery disease (20%), Atrial fibrillation (22%), stroke or transit ischemic attacks (12%), chronic obstructive pulmonary disease (24%), diabetes mellitus (30%). 20% of patients had a history of cognitive impairment or dementia at baseline.

Figure 3.

Prevalence (and 95% confidence interval) of patient-reported impairments captured on the eRFA (N=38).* Domains are ordered based on descending prevalence.

*For clarity, prevalence of polypharmacy as reported on the eRFA is not shown in this figure since details related to medication was also available among patients who did not complete the eRFA and is reported in Table 2 for the full cohort.

Informed by the GA, 36% patients were recommended to start PT/OT. 62% of patients were given nutritional recommendations, and 18% of patients were asked to liberalize unnecessary diet restrictions. Patients’ medication lists ranged between 2 and 22 prescription and non-prescription medications (median 10, quartiles 8, 13). Polypharmacy (defined as ≥10 medications) was present in 60% of patients. The clinical pharmacist evaluated 31 (62%) patients. The pharmacist interventions included patient education (100%), recommendations for deprescribing (22%), for non-adherence (16%), for potentially inappropriate medications (16%) [23] and significant drug-drug interaction (3%). For patients not evaluated by the pharmacist, medication reconciliation was done by the RN.

Following the CAIT clinic assessment, 52% of the patients went on to receive chemotherapy (24% standard dose, 28% with dose modifications), 20% received Radiation therapy, 12% Immune therapy, 4% hormonal therapy. In addition, 2% underwent surgery, 2% chose to pursue alternative medicine, 8% were placed under observation, and 6% enrolled in hospice care.

Patient and oncologist satisfaction with the program was an important aspect over this period. Nearly half (14/29) of oncologists responded to the survey. 100% of the Oncologists who responded reported that the appointment was scheduled in a timely manner and that it was easy to communicate with the geriatricians. The majority (93%) agreed that their goals for referring the patients to the CAIT clinic were met and that the clinic helped define an optimal treatment plan. 100% responded that they would refer patients to the CAIT clinic again.

Patients’ feedback was insufficient (16%, 8/50 responded) though responses were favorable. 100% responders agreed with the importance of a thorough review of their health and fitness level, and 88% agreed about understanding the goals of the evaluation.

DISCUSSION

This paper describes the development and implementation of an interdisciplinary, hybrid, asynchronous, geriatric oncology clinic, taking advantage of the surge in telemedicine during the COVID-19 pandemic. The clinic aims to provide frailty risk assessment with timely recommendations so that geriatric syndromes are appropriately addressed, and GA variables become part of the decision-making process in the cancer treatment of the older adult. Most clinicians reported high satisfaction with the process.

There are multiple published models for providing GA such as self-administered questionnaires, screening tools, and abbreviated or lengthy evaluations.^3‚24 Common roadblocks (e.g., lack of time, limited geriatrician availability) and institution-specific barriers have limited implementation of these models. The proposed model addresses some of these barriers with the potential for widespread implementation.

Notable barriers at MSK are the geographic fragmentation of clinic locations, as well as the lack of space; both were mitigated using telemedicine. To address oncologists’ potential reluctance to accepting the model, we provided educational meetings about Geriatric Oncology and the benefits of incorporating GA and prognostic calculators into decision making. Scaling the program in stages and performing PDSA cycles to identify problems resulted in a smooth implementation process without major scheduling delays.

Patient willingness and comfort is another important barrier. Traditionally, healthcare is an exception among service industries in that it limits consumer choice in the method of service utilization. This model innovatively removes that restriction and allows the patient to choose how they utilize the service (i.e., in person, remotely, or hybrid). Patients are often unwilling to undergo time consuming multidisciplinary visits. Many patients have difficulty finding caregivers who can take time off work to accompany them.²⁵ The remote option overcame these barriers. However, some patients expressed difficulty navigating telemedicine. These patients were able to involve a caregiver, come into the Geriatrics clinic, or go to the nearest MSK regional center to be set up for the telemedicine visit. Therefore, access to and knowledge of technology were not significant barriers for our patient population.

GA can detect previously unidentified problems in as many as 70% of patients with cancer.²⁶ In our cohort GA identified multiple undetected deficits such as functional dependency or cognitive impairment. While 20% of patients had a history of cognitive impairment, 50% of them had an abnormal cognitive screen at assessment. 53% of the patients showed impaired functional performance that could interfere with cancer treatment. It is now well established that GA results should be taken into consideration when making treatment decisions and that Geriatrician-led GA interventions are associated with improved chemotherapy tolerance and higher likelihood of treatment completion.^27‚28 Recent RCTs showed significant benefits of such interventions on clinical outcomes. The GAIN²⁹ and GAP70+¹¹ trials showed a reduction in grade 3 or higher chemotherapy related toxicities in older patients who had GA results-led interventions. The GAP70+ trial showed that reduced dose chemotherapy in their patient cohort did not compromise survival. In our cohort, of the patients that went on to receive chemotherapy (26/50), over half had regimen modifications.

Understanding life expectancy, risk of developing treatment toxicity and the potential impact of treatment on their perceived quality of life may help patients in their approach to shared decision making. If the outcome of treatment is survival but with severe functional or cognitive impairment, many older adults would choose to forgo such treatment.³⁰ A cluster-randomized clinical trial showed that the inclusion of GA in the oncology visit significantly improved patient-centered and caregiver-centered communication about aging-related concerns.³¹ In addition, integration of the GA results, life expectancy estimates and chemotoxicity risk calculators scores may prevent undertreatment based on chronological age in fit older adults who could tolerate standard therapies, as well as overtreatment in frail older adults who would derive a greater net benefit from less intensive therapy.³² 24% (n=12) of our patients went on to receive standard dose chemotherapy. For patients who were originally considered for chemotherapy but did not proceed (n=15), frailty was the reason in 93.3%. Findings on the GA can be crucial in end-of-life discussions³³ and 6% of our cohort of patients enrolled in hospice after the CAIT clinic assessment.

Despite the significant growth in the body of evidence supporting the need to risk-assess older adults with cancer, the implementation of such assessment remains limited. The top three barriers to the use of GA reported by the Association of Community Cancer Centers survey are lack of familiarity with available screening and/or assessment tool, limited time, and limited personnel.³⁴ We were able to take advantage of the telemedicine growth to develop a hybrid format that reduced the time and personnel needed to implement this clinic. Specifically, the same personnel were able to see patients referred from multiple locations, including sites with no geriatricians on staff.

Our study is not without limitations, including factors that limit the generalizability of the model. Our patient population is predominantly White, living with family and well educated. This may be different from the patient profiles in other organizations. Many cancer care institutions do not have the resources needed to develop a model such as the one described in this paper. To evaluate the “bottom line” elements as recommended in the ASCO guidelines (function, falls, depression, cognition, nutrition, comorbidities),⁴ the instruments needed are not complex and may be administered by allied health practitioners. Providers could incorporate one tool at a time in their routine and test feasibility. Asynchronous remote visits with an RN would allow for collection of medical history and medication lists and data on function, social support, and patients’ perspective on what is most important to them. The eRFA is a tool developed and used in MSK. However, it is for the most part, a patient-reported electronic instrument that includes all GA domains. It could be implemented on-line or in a paper-and-pen version depending on the clinician’s available resources. The small number of patients who responded to the satisfaction survey is another limitation of our project. MSK patients receive multiple questionnaires and survey fatigue is known to reduce response rates and data quality.³⁵ Overall, MSK survey completion rates among outpatients are 15– 25%. Our rate of response was 16% fell within this range but is still insufficient and needs to be improved. Future evaluation of the CAIT clinic will include a caregiver satisfaction survey. We should also emphasize that our study is about the process of developing a new model of care, and we did not measure patient outcomes such as chemotoxicity or emergency room visits.

Next steps for the CAIT clinic will focus on ensuring the sustainability of the program, expanding the CAIT clinic to all services, and rigorous evaluation of the impact of CAIT clinic recommendations on cancer treatment outcomes. During our study period close to 8000 patients 65 and older started chemotherapy at MSK. As we expand, eligibility criteria will need significant modifications in order to accommodate referrals in a timely manner with the available resources. The initial success of the CAIT clinic provides strong support for the potential dissemination across services and institutions.

Key Points:

• Frailty assessment can be done efficiently to inform cancer treatment decision making

• Telemedicine opens opportunities to overcome barriers and develop new care models

Why does this paper matter?

This flexible model permits unbundling of visit components to deliver timely, interdisciplinary, pre-cancer treatment geriatric assessment