Implementation and Retrospective Examination of a Lung Cancer Survivorship Clinic in a Comprehensive Cancer Center

Abstract

Purpose:

The number of early-stage lung cancer survivors (LCS) is increasing, yet few survivorship programs address their specific needs. We developed a workflow to transition early-stage LCS to dedicated lung survivorship care and comprehensively identify and address their needs using electronic patient-reported outcomes (ePROs).

Methods:

A lung cancer multidisciplinary team developed a workflow (e.g., referrals, survivorship care plan delivery, documentation, orders, tracking, ePROs, and surveillance) for a survivorship clinic staffed by Advanced Practice Providers (APPs). ePROs included the NCCN Distress Thermometer, PROMIS-29, and investigator-developed patient satisfaction items. Patient characteristics, ePROs, and referrals are described; chi-square and t-tests examined ePRO completion by patient characteristics and compared PROMIS-29 domains by treatment modality and to a national sample.

Results:

From January 2020-March 2023, 315 early-stage LCS completed a survivorship orientation visit. Patient satisfaction was high; 75% completed ePROs. Females were overall less likely to complete ePROs than males; male, age 65+, Black or other race, and rural patients were more likely to complete ePROs in clinic vs. online. Patients reported lower symptom burden compared to a general population of early-stage LCS in the United States; scores were similar regardless of treatment modality. Rates of moderate-severe symptoms ranged from 6% (depression) to 42% (poor physical function); ≤20% had a referral placed.

Conclusions:

A referral-based, APP-staffed survivorship clinic model for early-stage LCS which includes ePROs to identify specific needs is acceptable to patients. Future work should include outreach to female LCS and increasing supportive care referrals and acceptability to further address early-stage LCS reported needs.

Micro Abstract:

The number of early-stage lung cancer survivors is increasing, necessitating survivorship programs to identify and address their needs. Of 315 patients seen for a survivorship orientation visit, the majority completed patient-reported outcome measures, but female patients may require greater outreach. Symptom burden was generally low regardless of treatment modality, but establishing and strengthening referral pathways may enhance outcomes.

Purpose:

As a result of advances in treatment and implementation of low-dose computed tomography (CT) screening, the number of individuals diagnosed with and treated for early-stage lung cancer is increasing.^1,2 Consistent with clinical practice guidelines, most patients with early-stage disease will receive curative treatment with either surgery or stereotactic body radiation therapy (SBRT).^3,4 Regardless of stage at diagnosis or treatment received, lung cancer survivors (LCS) report a complex array of unmet care needs, including limited functional capacity, symptoms and side effects (e.g., pain, dyspnea, insomnia, and fatigue), high rates of depression and anxiety, elevated risk for second primary cancers, and comorbid medical conditions.^5–8 These issues often persist for years post-treatment and are more prevalent among lung cancer survivors compared to other groups.^9–11

The rapid increase in early-stage LCS identified through lung cancer screening necessitates a plan for care after treatment. The Institute of Medicine and National Comprehensive Cancer Network (NCCN) recommend formal survivorship programs to address the needs of cancer survivors which include physical functioning, mental health, late and long-term treatment effects, smoking cessation, and ongoing surveillance.^12,13 Such programs are especially important in lung cancer given historically high symptom burden and low rates of service utilization in this population.¹⁴ There is a paucity of publications addressing sustainable care models for early-stage LCS care and an urgent need to identify efficient, cost-effective methods that balance the timely assessment of newly diagnosed patients with the needs of this exploding population.¹⁵ Lung cancer survivorship clinics staffed by advanced practice providers (APPs; e.g., nurse practitioners, physician assistants) and are one way to address these competing needs in a healthcare environment with limited resources. A previous non-randomized retrospective study found that early-stage LCS followed in a lung survivorship clinic after surgery had similar if not superior survival outcomes and rates of compliance with follow-up scans and visits compared to patients followed by their treating surgeon alone.¹⁵ While this previous work demonstrates feasibility of a survivorship clinic for early-stage surgical patients, to date, no studies have addressed lung survivorship needs in patients treated with SBRT, or determined whether SBRT patients are similar enough to surgical patients be followed in a combined clinic for early-stage LCS receiving single-modality treatment with curative intent. Since patients receiving SBRT are typically inoperable, in concordance with guidelines that advocate surgical resection for operable patients, there are notable and well-documented differences between these two groups in terms of pre-treatment characteristics. SBRT patients are more likely to have risk factors such as older age, greater overall comorbidity burden, greater prevalence of hypertension or pulmonary comorbidities, worse pulmonary function, and worse performance status,^16–20 however, it is unknown whether these differences necessitate a different model or level of survivorship care. The lack of published literature describing models of lung cancer survivorship care inclusive of SBRT patients, who may have worse outcomes and greater survivorship care needs, represents a critical knowledge gap. Informed by extant literature on the value of routinely-collected patient-reported outcome measures, ^21–24 we integrated the use of electronic patient reported outcome measures (ePROs) into survivorship visits to characterize symptom management needs and inform referrals.

The objective of the present study is to describe the process of transitioning early-stage LCS to comprehensive, long-term care and to report patient needs identified via ePROs administered in our dedicated survivorship clinic inclusive of both surgical and SBRT patients. While this clinic shares similarities to others previously described,²⁵ the Atrium Health Wake Forest Baptist Comprehensive Cancer Center (AHWFB-CCC) Lung Cancer Survivorship Clinic was developed with a specific focus on including all patients receiving single-modality treatment and using tools within the electronic health record (EHR) to streamline workflows and improve patient access. These foci were informed by the substantial proportion of rural patients served by the cancer center and the program’s development in the context of the COVID-19 pandemic. The Lung Cancer Survivorship Clinic is ongoing; this report uses data gathered from the first three years of implementation for the following objectives: 1) describe the clinic workflow developed through stakeholder feedback; 2) report on patient satisfaction with the clinic; 3) identify potential differences in ePRO completion by patient characteristics; and 4) describe the needs of our LCS for symptom management and referrals (e.g., primary care, smoking cessation, psychological support), examining potential differences by treatment modality. This study provides a model for a dedicated lung survivorship clinic in a cancer center setting and identifies specific survivorship care needs.

Patients and Methods:

Survivorship Clinic Planning Process

In January 2019, stakeholders from the lung cancer multidisciplinary team at AHWFB-CCC formed a multidisciplinary working group, including cardiothoracic surgeons, radiation oncologists, medical oncologists, nurse and patient navigators, advance practice providers (APPs), implementation science researchers, the survivorship medical director, and administrative staff. Representatives from Psychosocial Oncology (counselors and social workers) were also consulted regarding psychosocial issues. This group met to collaboratively develop a workflow for a referral-based lung survivorship clinic (objective 1) that would achieve the following: delivering a survivorship care plan (SCP) to every early-stage lung cancer survivor; providing a 1-hour in-person survivorship orientation visit to every patient, during which APPs meet with patients to review their SCP and address specific concerns; and provide appropriate surveillance and follow-up care, including referrals to institutional and community resources for patient-reported concerns. Issues addressed by this group included: identifying staff needed to run the clinic, determining criteria and a standard pathway for e-referring patients to the clinic, developing a standardized clinician workflow (documentation, orders, and SCP delivery), selecting patient-reported assessment tools, developing a real-time tracking system for clinic volumes, demographics, and ePROs, and developing standard follow-up care for LCS based on national guidelines.^3,26

Data Collection and Measures

Demographics and clinical characteristics (e.g., age, gender, race, residence according to 2010 rural-urban commuting area (RUCA) codes),²⁷ tobacco use, type of treatment received for lung cancer (surgery or SBRT), and items from the clinic flowsheet (e.g., referrals placed) are automatically extracted from the EHR into a database and visualized using a custom real-time dashboard. Additional data (e.g., stage, histology) were retrieved from the AHWFB-CCC cancer registry. The use of retrospective clinical data for the present study was approved by the institutional review board at Wake Forest University School of Medicine.

Patient-reported assessments included pre-visit overall distress using the NCCN Distress Thermometer (DT), the standard distress screening tool utilized by AHWFB-CCC,²⁸ and specific symptoms using the PROMIS-29.^29,30

NCCN Distress Thermometer (DT).

The NCCN DT is a single-item Likert-style scale from 0 (no distress) to 10 (extreme distress)³¹ that is widely-used as part of distress screening due ease of administration, adequate psychometric properties, and cross-cultural validity.^32,33 A cutoff score of 4 has been recommended for maximizing sensitivity and specificity in identifying patients who are likely experiencing psychological distress.³³

PROMIS-29.

The PROMIS-29 is a rigorously-developed, widely-used disease non-specific tool from the PROMIS suite of measures. The PROMIS-29 covers 7 health-related domains, each of which demonstrate strong reliability and validity: depression, anxiety, fatigue, pain interference, sleep disturbance, social functioning, and physical functioning.³⁰ First, each of the 7 domains is summed to yield a raw score, which is paired with a corresponding T-score. All PROMIS subscales are standardized to a T-score metric with a mean of 50 and a standard deviation of 10; this metric is anchored to a general United States (US) population sample.²⁹ A score of greater than 50 indicates a worse outcome compared to the general population for symptoms (e.g., anxiety, fatigue, pain interference, sleep disturbance) and better than the general population for functioning (e.g., social functioning, physical functioning). ²⁹ Additional work has established PROMIS-29 norms specific to cancer survivors in the US, since T-scores tend to be lower for these groups than the general US population; we used T-score reference values specific to early-stage (I-II) LCS in the US for the purpose of this study.³⁴ While normative comparisons provide helpful context, they do not necessarily indicate clinical meaningfulness. Although the standard deviation of each T-score is ~10, previous psychometric work with cancer patients has established a difference of 3 points on PROMIS-29 domains as clinically meaningful;^34,35 therefore we report on both clinically-significant and statistically-significant differences in all comparisons. Severity levels for each symptom (mild, moderate, severe) were determined based on established cut points for individuals with cancer.^34,36

Patient Satisfaction.

Patients also completed an optional clinician-developed after-visit survey assessing their perceptions of the visit and feelings of empowerment and satisfaction on a scale from 0-10. This survey was offered at the end of their visit in clinic via paper and pen and was also sent to them via the EHR patient portal.

Statistical Analyses

Following a pilot period from June-December 2019, ePRO data were collected for patients who completed lung cancer survivorship orientation visits from January 2020 to March 2023. All patients completing survivorship orientation visits during this data collection period were included in analyses; there were no additional eligibility criteria to be included in this retrospective EHR study. Data were exported from medical records and all analyses were performed using R Statistical Software v4.2.1.³⁷ Descriptive statistics were used to summarize patients’ demographic and clinical characteristics (e.g., age, gender, race/ethnicity, residence according to RUCA codes, tobacco use, lung cancer treatment modality, and patient satisfaction scores (for study objective 2). Chi-square tests were used to investigate the relationships between patient characteristics, ePRO completion, and method of completion (objective 3). Fisher’s exact test was used where variables had small cell sizes (e.g., race/ethnicity). For objective 4, referrals and ePROs (DT distress score and PROMIS-29 domains by severity level) were summarized using descriptive statistics. T-tests were also used to compare survivorship clinic patients’ average T-Scores to known United States population averages for early-stage LCS³⁴ and to compare PROMIS-29 domains by treatment type (surgery vs. SBRT). Rates of missing data were <1% for key sociodemographic and clinical characteristics, except for histology (31.7%) and stage (41.0%). Among individuals who completed ePROs, no DT or PROMIS-29 items were missing, therefore all analyses used complete cases.

Results:

1). Lung Survivorship Clinic Model of Care and Procedures

The Lung Cancer Survivorship Clinic opened in June 2019. Similar to previously reported “Oncology Specialist” designs^25,38 and informed by stakeholder input, the clinic is staffed by APPs who are embedded in the lung cancer multidisciplinary care team and also involved in the initial evaluation and treatment of lung cancer patients at AHWFB-CCC.

Clinic Appropriateness and Referral Process.

Based on cancer registry data, approximately 450 new cases of lung cancer are seen per year at AHWFB-CCC, 1 (0.2%) of which are stage 0, 113 (25.1%) stage I, and 23 (5.1%) stage II. The lung cancer multidisciplinary team identified the target clinic population as any patient: 1) treated with curative intent, 2) treated with a single modality (e.g., surgery or SBRT), and 3) post-treatment. Due to these criteria, most patients suitable for the clinic have stage I disease although some have low-risk stage II disease. Patients are not excluded from the clinic if they had a cancer diagnosis prior to their current lung cancer diagnosis. Unlike other models of survivorship care that are limited to surgical patients or require patients to be disease-free for 1 year before referral to survivorship,²⁵ SBRT patients are able to be referred to this clinic and there is no requirement that patients be disease-free for a certain amount of time prior to referral. Patients are referred by their treating physician at the time of their first post-operative visit or at their final session of SBRT. After receiving the referral, survivorship clinic staff schedule patients for a survivorship orientation appointment in addition to a chest CT approximately 6 months after the patient’s post-operative or post-SBRT visit.

Clinic Workflow.

One week prior to the orientation visit, ePROs (described below) and an introductory video are automatically delivered to the patient via the EHR patient portal. Upon completion by the patient, ePRO responses are delivered to the responsible APP’s inbox for review prior to the visit; patients are given the option to complete PROs on paper in clinic prior to the visit if they are unable to complete them at home. The APP also completes a SCP prior to the patient visit, which is standardized with multiple auto-generated fields to minimize provider effort. The early-stage LCS orientation visit flowsheet comprises a standardized checklist for end-to-end visit documentation, including a review of treatment and SCP distribution, discussion of surveillance and primary care needs, update of family history and immunizations, and a focused review of symptoms. At the visit, staff complete the flowsheet while rooming the patient. Flowsheet data automatically populate a clinic note, minimizing the provider’s documentation time. The APP reviews the flowsheet information with the patient and discusses issues reported in the ePROs. Wellness recommendations (e.g., immunizations) are reviewed, a primary care provider (PCP) is identified, the SCP is reviewed, and the patient and PCP receive a copy of the SCP. Patients without a documented PCP are offered a referral. Appropriate referrals to specialists (e.g., cardiology, pulmonology) or for supportive care services (e.g., psychosocial oncology, pulmonary rehabilitation, physical therapy, nutrition, smoking cessation) are made based on ePRO review and shared decision-making. Lung cancer survivor-specific smart sets with prefilled standard orders prompt providers at the end of the visit. After the orientation visit, thirty-minute follow-up visits are scheduled at appropriate intervals alongside a surveillance chest CT and an after-visit survey is sent to the patient via the patient portal.

Electronic Patient Reported Outcomes (ePROs).

The ePRO battery consisted of two measures selected by lung multidisciplinary team stakeholders: the NCCN distress thermometer (DT),²⁸ which is part of routine distress screening at AHWFB-CCC, and the PROMIS-29.²⁹ Stakeholders with PRO expertise recommended the use of the PROMIS-29 measure in addition to the DT based on its ability to provide: 1) clinically-meaningful severity scores to inform referrals, 2) greater information about the clinic population distribution of specific needs, and 3) a point of comparison to established population norms. The PROMIS-29 became available in the EHR as of January 2020, and all APPs and clinic staff were trained on how to use the ePROs and other tools by EHR support staff prior to and during workflow adoption. The PROMIS-29 was also used to inform referrals. In May of 2021, an automated decision support alert was developed for this clinic. Moderate to severe scores on PROMIS-29 measures of fatigue, anxiety, and/or depression triggered the alert which prompts the clinician to address the symptom (typically via psychosocial oncology referral). Patients with severe scores are offered immediate assessment with an on-call psychosocial oncology team member, while patients with moderate scores receive a referral addressed in 1-3 business days. To close the alert, providers must document how the moderate to severe ePRO was addressed.

Developing a real-time patient tracking system.

A custom dashboard was created to display data extracted from the clinic workflow in real time. The dashboard is monitored monthly by the medical director of cancer survivorship. Annual summary reports are reviewed by the lung multidisciplinary team and the Cancer Committee in accordance with Committee on Cancer standards. APPs review and update their clinic workflow and resources annually.

Lung Survivorship Clinic Patients

From January 2020 to March 2023, 315 lung cancer survivorship orientation visits (including PCP identification and SCP delivery) were completed in-person; data were collected in real time. Sample characteristics are displayed in Table 1; numbers included in analyses are depicted in Figure 1. Briefly, 143 (45%) identified as female, more than 70% of patients were over the age of 65 (M=69, SD=10) and most identified as white (N=260, 83%) or Black/African American (N=45, 14%). A significant minority (N=52, 17%) lived in rural areas, and 46 (15%) reported current smoking. More patients received surgery than SBRT (64% vs 36%). Of patients with available data, 79% had adenocarcinoma and 74% had stage I disease.

Table 1.

Lung Cancer Survivorship Orientation Visit Patient Characteristics (N=315)

Characteristic	N (%)
Age [mean (SD)]	69 (10)
Age Group
<65	89 (28%)
≥65	226 (72%)
Gender
Male	172 (55%)
Female	143 (45%)
Race/Ethnicity
White/Caucasian	260 (83%)
Black/African-American	45 (14%)
Other/Refused	10 (3.2%)
Residence (RUCA)
Urban (1-3)	263 (83%)
Rural (4-10)	52 (17%)
Treatment Modality
SBRT	113 (36%)
Surgery	202 (64%)
Smoking Status
Every day	36 (11.4%)
Some Days	10 (3.2%)
Former	196 (62.2%)
Never	72 (22.9%)
Missing	1 (0.3%)
Smokeless Tobacco Use
Current	4 (1.3%)
Former	20 (6.3%)
Never	290 (92.1%)
Missing	1 (0.3%)
Method of ePRO Completion
In-Clinic	155 (65%)
Self-Administered via Patient Portal	82 (35%)
Histology
Adenocarcinoma	160 (50.8%)
Large Cell	9 (2.9%)
Squamous Cell Carcinoma	46 (14.6%)
Missing/Unknown	100 (31.7%)
Clinical Stage
0	4 (1.3%)
I	148 (47.0%)
II	34 (10.7%)
Missing/Unknown	129 (41.0%)
ePRO Results (N=237)
Distress Thermometer [mean (SD)]	1.71 (2.33)
Distress Thermometer (non-zero) [mean (SD)]	3.40 (2.26)
PROMIS-29 Health Domains [mean (SD)]
Anxiety	48.0 (9.27)
Depression	46.3 (7.73)
Fatigue	47.1 (10.00)
Pain Interference	51.6 (9.53)
Physical Function	45.7 (10.14)
Sleep Disturbance	47.9 (8.91)
Social Function	53.5 (9.67)
After Visit Survey Results (N=101) *
Q1: Did you find your appointment with your survivorship provider helpful?	77 (76%)
Q2: Do you find your Treatment Summary helpful?	70 (71%)
Q3: Did your Survivorship Visit today help you understand your cancer treatment and side effects?	70 (74%)
Q4: Did your Survivorship Visit today help you feel better or able to cope better with side effects of your cancer treatment? a	66 (69%)

ePRO= electronic patient reported outcome measure; RUCA=rural-urban commuting area; SBRT= stereotactic body radiation therapy

^*Response options for all questions range from 0-10 with 0 being not helpful at all and 10 being extremely helpful. Due to skewness of responses, all after visit survey results dichotomized as 10/10 “extremely helpful” vs. <10/10

Figure 1.

Numbers included in analyses

2). Patient Satisfaction with the Clinic

Among the patients who completed the after-visit survey (N=101), overall satisfaction was high and >69% found their visit “extremely” helpful (Table 1).

3). ePRO Completion by Patient Characteristics

Most of the 315 patients with orientation visits completed ePROs (n=237, 75%); 82 (35%) completed them at home via the patient portal prior to the visit; 155 (65%) completed them in clinic. Most missed ePROs happened early in the implementation process (2020). Female LCS were less likely to complete ePROs using either method (p=.005; Table 2). Patients aged 65 or older (p =.04), male (p =.009), Black/African American or other non-white race (p <.001), or from rural areas (p =.02) were less likely to complete the ePROs via the patient portal (Table 2)

Table 2.

Patient Reported Outcome Measure Completion and Method by Demographic Characteristics (N=315)

ePRO Completion Status
	Completed ePROs N=237	Did Not Complete ePROs N=78	P-Value
Age Group			0.991
<65	67 (28)	22 (28)
≥65	170 (72)	56 (72)
Gender			0.005
Male	140 (59)	32 (41)
Female	97 (41)	46 (59)
Race/Ethnicity *			0.738
White/Caucasian	193 (81)	67 (86)
Black/African-American	36 (15)	9 (12)
Other/Refused	8 (3.4)	2 (2.6)
Residence (RUCA)			0.147
Urban (1-3)	202 (85)	61 (78)
Rural (4-10)	35 (15)	17 (22)
Treatment Modality			0.279
SBRTc	89 (38)	24 (31)
Surgery	148 (62)	54 (69)
ePRO Completion Method
	In-Clinic N=155	Patient Portal/Self-Administered N=82	P-Value
Age Group			0.039
<65	37 (24)	30 (37)
≥65	118 (76)	52 (63)
Gender			0.009
Male	101 (65)	39 (48)
Female	54 (35)	43 (52)
Race/Ethnicity *			<0.001
White/Caucasian	116 (75)	77 (94)
Black/African-American	33 (21)	3 (3.7)
Other/Refused	6 (3.9)	2 (2.4)
Residence (RUCA)			0.019
Urban (1-3)	126 (81)	76 (93)
Rural (4-10)	29 (19)	6 (7.3)
Treatment Modality			0.055
SBRT	65 (42)	24 (29)
Surgery	90 (58)	58 (71)

Note. ePRO= electronic patient reported outcome measure; RUCA=rural-urban commuting area; SBRT= stereotactic body radiation therapy

^*Fisher’s exact test used due to small sample size in racial/ethnic groups

4). Symptom Management and Referral Needs

Forty-two (18% of 237) reported a DT score ≥4 and 44 (19%) reported moderate-severe anxiety, depression, or fatigue on the PROMIS-29. On the PROMIS-29, fewer than 6% of patients reported a severe score for any symptom; moderate to severe scores ranged from 6% (depression) to 42% (physical function; Figure 2). Patients reported a higher level of functioning and lower symptom burden relative to norms for early-stage (I/II) LCS in the United States³⁴ in all areas (Figure 3). Specifically, this population reported statistically and clinically-significantly (≥3 points) better scores in the areas of social functioning, sleep disturbance, physical function, fatigue, and depression. Patients in this population also reported statistically (but not clinically) better anxiety and pain interference compared to early-stage lung cancer population norms. There were no statistically- or clinically-significant differences in PROMIS-29 domains between patients treated with surgery versus SBRT (all p’s >.05, Figure 4). Following ePRO discussion and shared decision-making, ≤20% had a referral placed (N=64, Figure 5). Common referrals included: tobacco cessation (12; 26% of currently smoking population referred), pulmonology (8; 2.5% of total population), pain clinic (5; 1.6% of total population), psychosocial oncology (4; 9.1% of patients with moderate-severe anxiety, depression, or fatigue referred), internal medicine and urology (4 each, 1.3% of total population), and physical therapy, ear nose and throat, family practice, swallowing/speech therapy (3 each; 1% of total population). Developing a method to track completion of the referral by LCS is ongoing.

Figure 2.

PROMIS-29 T Scores by Severity Cut Point (N=237)

Figure 3: Comparison of PROMIS-29 T-Scores for AHWFB-CCC (N=237) and US Early-Stage (I/II) Lung Cancer Population (N=308).

Note. AHWFB-CCC= Atrium Health Wake Forest Baptist Comprehensive Cancer Center

Mean values and 95% CI by PROMIS Domain AHWFB-CCC vs. US Stage I/II Lung Cancer Populations:

Social Function: 53.5 [52.2, 54.7] vs. 47.2 [46.0, 48.4], p<.001

Sleep Disturbance: 47.9 [46.8, 49.0] vs. 51.1 [49.9, 52.3], p=.003

Physical Function: 45.7 [44.4, 47.0] vs. 40.2 [39.2, 41.2], p<.001

Pain Interference: 51.6 [50.4, 52.8] vs. 54.2 [53.0, 55.4], p<.001

Fatigue: 47.1 [45.8, 48.3] vs. 54.6 [53.4, 55.8], p<.001

Depression: 46.3 [45.3, 47.3] vs. 50.1 [48.7, 51.5], p<.001

Anxiety: 48.0 [46.8, 49.2] vs. 50.8 [49.4, 52.2], p=.01

Figure 4: Comparison of PROMIS-29 T-Scores for AHWFB-CCC Early-Stage Lung Cancer Patients Receiving Surgery and SBRT (N=237).

Note. AHWFB-CCC= Atrium Health Wake Forest Baptist Comprehensive Cancer Center

Mean values and 95%CI by PROMIS Domain Surgery vs. SBRT Populations:

Social Function: 52.8 [51.2, 54.4] vs. 54.6 [52.5, 56.6], p=0.18

Sleep Disturbance: 48.5 [47.0, 50.0] vs. 47.0 [45.3, 48.8], p=0.22

Physical Function: 45.0 [43.3, 46.6] vs. 47.0 [44.8, 49.2], p=0.14

Pain Interference: 51.7 [50.1, 53.3] vs. 51.4 [49.5, 53.2], p=0.77

Fatigue: 47.4 [45.8, 49.0] vs. 46.5 [44.4, 48.6], p=0.51

Depression: 46.8 [45.5, 48.1] vs. 45.5 [44.0, 47.1], p=0.21

Anxiety: 48.4 [46.8, 49.9] vs. 47.1 [45.3, 49.0], p=0.31

Figure 5:

Clinical Referrals Resulting from 315 Survivorship Orientation Visits (N=64)

Discussion and Conclusion:

We successfully implemented an Early-Stage Lung Cancer Survivorship Clinic and provided an SCP, survivorship orientation visit, and specialty referrals for 315 survivors. Using an integrated APP model with ePROs, we were able to screen for care needs of early-stage LCS and adhere to clinical practice guidelines for follow-up and monitoring. Satisfaction scores show that patients find survivorship care separate from their treating surgeon or oncologist acceptable. Our “Oncology Specialist” clinic model,³⁸ in which survivorship APPs are situated within the cancer center and integrated with the patient’s oncology treatment team, may have eased provider and patient concerns about the survivorship transition. In the present study, we were able to ensure patients had a PCP and felt empowered to manage symptoms related to their cancer treatment. Survivorship care visits offer benefit over SCP delivery alone, including increased referrals to supportive care.³⁹ Thus, programs like ours which combine SCP delivery with a survivorship orientation visit may add value by facilitating timely surveillance and emphasizing other important health maintenance behaviors. Furthermore, by focusing on preparation for survivorship visits, education on survivorship recommendations, and assessment and treatment of symptom concerns (e.g., pain and stress management), our model meets key survivorship needs identified by rural LCS.⁴⁰

The 75% ePRO completion rate similarly demonstrates feasibility and acceptability of our clinic procedures and compares favorably to previously-reported ePRO completion rates (≤60%) from clinics administering similar measures.^21,23,41 Notably, the majority of missed ePROs happened early in the implementation process and in the context of the COVID-19 pandemic. We successfully used the PROMIS-29 to identify and address the needs of early-stage LCS, which flagged individuals for psychosocial oncology follow-up at a rate similar to the Distress Thermometer (19% vs. 18%).⁴² Our data showed that early-stage LCS reported lower sleep disturbance, pain interference, fatigue, depression, and anxiety and better physical and social functioning at their orientation visit than a similar reference sample of early-stage LCS³⁴ and that these issues did not differ by treatment type (surgery vs. SBRT), indicating that a combined survivorship clinic for patients receiving single-modality treatment is feasible.

However, the PROMIS-29 also identified areas for improvement. Notably, a substantial proportion of PROMIS scores were in the moderate to severe range (42% for physical function, 22% for pain interference). These data suggest that establishing and strengthening pathways for connecting patients with resources to address these concerns is an important future direction. The number of referrals to supportive care and other specialties is lower than expected given the prevalence of specific issues (e.g., ~15% currently smoking but ~4% referred for smoking cessation, ~19% reporting moderate-severe anxiety, depression, or fatigue but <2% referred to psychosocial oncology). These rates are consistent with previous research indicating limited receipt of mental health care among LCS who screen positive for depression or anxiety using validated measures.^14,43 For example, among a population of mostly stage I patients receiving curative surgical resection, only 37.5% of those with clinically significant post-operative anxiety or depression were referred to mental health care, although all who had abnormal scores were encouraged to go.⁴³ Reasons for reluctance or refusal may be patient perceptions that mental health care is not needed, a lack of accessible resources, or a preference to manage issues independently.^14,43,44 However, rates of referrals to smoking cessation and psychosocial oncology services may also be lower because these issues are already routinely screened as a part of clinical care and patients smoking or experiencing psychological distress may have already been connected with services prior to their survivorship visit. Given time and resource constraints associated with retrospective EHR-based research, it was not feasible to describe outcomes of referrals or to determine whether patients were already connected with appropriate services at the time of their survivorship visit. Thus, we are unable to determine the reason for the lower-than-expected referral rates and whether the referrals made a meaningful difference to patients. The discrepancy between referrals and ePRO-identified needs may reflect a true unmet need, a patient- or provider-level belief that symptom management needs are already met or will be met through established clinical relationships, limited resources and access issues, or some other factor. It is also important to note that the population of patients who screen positive for anxiety, depression, and distress is heterogeneous; some have clinical anxiety or depressive disorders which may have preceded their cancer diagnosis, others have distress driven by financial toxicity, declining functional status, or other treatment side effects, and still others may experience prognostic uncertainty or fear of cancer recurrence.^8,32 Recent studies have identified an especially high prevalence of living with uncertainty and fear of recurrence as unmet emotional needs among lung cancer survivors irrespective of stage, indicating a need for greater attention to this issue in psychosocial care for lung cancer survivors.^8,45 Exploring strategies to improve tracking of survivorship needs (including outcomes of referrals) and to enhance acceptability, uptake, and impact of supportive care among early-stage LCS represent areas for further investigation.

Although overall completion rates were high, we also identified disparities in ePRO utilization. Patients who were female were less likely to complete ePROs overall, and those who were older, male, Black or other non-white race, and rural were less likely to complete ePROs through the patient portal prior to their orientation visit. While these differences in in-person vs. online completion are consistent with previous research on ePRO use and EHR engagement,^23,46–48 our finding that female patients were overall less likely to complete ePROs runs directly counter to previously published work.^46,49 Lower rates of ePRO completion among women may partially explain the lower symptom burden we found especially in the areas of sleep disturbance, anxiety, depression, and fatigue, since women tend to score higher on these measures.^50,51,52

Nevertheless, flexibility in format for ePRO completion may allow for more equitable assessment of patient needs. Addressing only issues reported by patients who completed ePROs via the patient portal would have missed key psychosocial issues among already-underserved groups. Indeed, only 34.6% of ePRO completers in our clinical population did so via the patient portal compared to 43-85% in previous ePRO studies.^21,23,41 These findings highlight the importance of considering patients’ access to and comfort with technology when developing lung survivorship clinic procedures, and the need to better engage female LCS and explore reasons for ePRO non-completion.

The results of this study must be considered within the context of its limitations, namely, it was a retrospective study using data collected for clinical purposes from a single institution, and missing data may be a cofounding factor. Our target clinical population (early-stage lung cancer, completed single modality treatment) cannot be readily identified from discrete fields in the EHR. Therefore, we do not know the total number of patients who were potentially suitable for the clinic but who did not complete a survivorship orientation visit; these may be patients who continued to be followed by their treating oncologist or who were referred to the clinic but did not attend an orientation visit. It is possible that the population that attended a survivorship orientation visit may be higher functioning than the broader survivorship target population. Additionally, it was not feasible to obtain high quality data on patient characteristics requiring manual abstraction from clinical notes. This limitation precluded comparing surgical and SBRT patients by comorbidity burden, pulmonary functioning, or rates of specific comorbidities (including prior cancer diagnoses), thus, it is possible that our SBRT population is younger and healthier than the general population of early-stage patients treated with SBRT alone. However, our finding that surgical and SBRT patients have similar quality of life despite likely differences in pre-treatment risk profiles is consistent with several observational studies showing that while patients referred for SBRT have worse baseline health-related metrics, surgical resection is actually associated with a steeper decline in QOL immediately post-treatment, with both groups showing comparable outcomes at later follow-up.^20,53,54 Engaging stakeholders in pulmonology, understanding the prevalence of comorbidities (especially chronic obstructive pulmonary disease [COPD]), and collecting patient-reported respiratory functioning outcomes among patients in our early-stage lung cancer survivorship clinic are important next steps given the prevalence of these issues in the early-stage lung cancer survivor population generally⁵⁵ and the frequency of pulmonology referrals and impairments in physical functioning reported by this specific population. Additionally, although patient satisfaction was very high, we did not have data on patients’ health literacy or education level to understand if these factors may influence the survivorship clinic experience.

Future directions include engaging stakeholders and gathering qualitative data to identify methods to enhance equitable identification of patient needs via ePROs and referrals to appropriate services. Qualitative and mixed-methods research may also enhance our understanding of whether both surgical and SBRT patients feel that survivorship visits are sufficiently tailored to their specific needs. Other future directions include implementing alerts for moderate to severe physical functioning impairments and establishing clear pathways for managing common comorbidities of early-stage LCS that worsen physical function (e.g., COPD, sleep apnea, cardiovascular disease). Since grouping together surgery and SBRT was acceptable, another next step could be expanding the orientation visit model to include drop-in wellness consultations for higher-risk patients who might not meet the traditional definition of eligibility for “survivorship” care (e.g., patients on immunotherapy or maintenance chemotherapy with stable disease). Similarly, including stakeholders (e.g., patients, providers) in decisions regarding survivorship care for these higher-risk patients may identify optimal models of care with respect to timing of the survivorship orientation visit (e.g., following chemoradiation, following immunotherapy) and management of commonly undertreated comorbidities such as COPD.⁵⁶

In summary, we developed a referral-based, APP-staffed clinic for early-stage LCS that allowed us to identify and address patient-reported needs and efficiently deliver comprehensive survivorship care.

Clinical Practice Points:

• There is a paucity of published research describing efficient and acceptable models of post-treatment care for early-stage lung cancer survivors (LCS), especially models of care inclusive of SBRT patients that involve transitioning LCS soon after treatment.

• Our study described the process of survivorship clinic development and implementation and provides patient-reported outcome data from 315 early-stage LCS who received an orientation visit.

• The clinic was acceptable based on high patient satisfaction and higher rates of electronic patient-reported outcome measure (ePRO) completion compared to previous work.

• Counter to previous research, female LCS were less likely to complete ePROs and may require greater engagement efforts.

• Early-stage LCS may rely more heavily on in-clinic ePRO administration than other patient populations. Patients who were male, 65+, rural, and Black or another non-white race were more likely to complete ePROs in clinic.

• While symptom burden in this population was lower than a national sample of early-stage LCS, patients still experienced high rates of moderate-severe impairment especially in the areas of physical functioning and pain interference.

• Symptom burden did not differ between surgical and SBRT patients, suggesting management of both groups in a single clinic is feasible.

• Fewer referrals were placed than expected given patient-reported needs, suggesting room for improvement in establishing and strengthening referral pathways and uptake.

• Our study provides a roadmap for other institutions tasked with providing dedicated survivorship care to a growing population with unique needs.