Medical Manuscript
Preliminary Testing of a Patient-Reported Outcome Measure for Recurrent or Metastatic Head and Neck Cancer
American Journal of Hospice & Palliative Medicine® 1-8 ª The Author(s) 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1049909115569591 ajhpm.sagepub.com
Leanne K. Jackson, MD1, Jie Deng, PhD2, Sheila H. Ridner, PhD2, Jill Gilbert, MD1, Mary S. Dietrich, PhD3, and Barbara A. Murphy, MD1
Abstract Objective: We describe development and preliminary testing of Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) to assess residual symptoms, tumor-related symptoms, and side effects from therapy. Methods: Items were identified through patient and provider interviews. Card sort selected high-yield and high-impact items. The VHNSS-RM was administered to 50 patients with recurrent/metastatic head and neck cancer (RMHNC). Results: The VHNSS-RM includes 12 unique symptoms (diet change, tongue movement affecting speech/swallowing, face/neck swelling, neck/jaw cramping, bad breath, drooling, wound drainage/pain/odor, nasal congestion/drainage, eyes watering, face/tongue/ear/scalp numbness, headaches, and confusion) and 7 unique psychosocial issues (burden to family/friends, lost independence, fear, embarrassment, mood swings, stress, and boredom). Conclusions: The VHNSS-RM contains 35 physical and 12 psychosocial issues. The VHNSS-RM is feasible and not overly burdensome. Nineteen unique items may improve palliation to patients with RMHNC. Keywords recurrent head and neck cancer, metastatic head and neck cancer, patient-reported outcomes, symptom burden, toxicities, unique symptoms
Introduction Head and neck cancer (HNC) and its treatment are associated with substantial side effects resulting in symptom burden, functional deficits, and decreased quality of life. Acute toxicities due to surgical or radiation-based therapies for HNC are well described. Until recently, late effects have been poorly studied; however, increasingly there have been efforts to systemically identify late effects of toxicity and their impact.1-4 Conversely, data on the prevalence and severity of toxicities for patients with noncurable recurrent and/or metastatic (RM) HNC have been and remain limited. Efforts to document supportive care issues in this population remain scarce, and there are no patient-reported outcome (PRO) measures specifically developed for and tested in this population.5,6 A multitude of tools have been developed and tested, which focus on acute and/or late effects of primary therapy in HNC.6-10 Most of these tools are dominated by items directed at acute effects of radiation-based therapies for HNC. No PROs have been developed to assess symptom burden and functional deficits in patients with RMHNC. We postulate that the symptoms and functional problems experienced by patients in the RM setting are different than those noted during initial treatment. Thus, the use of PROs directed at the acute toxicities of surgery or radiation-
based primary therapy are inadequate. A PRO that can be used in the clinical and research settings to screen for and identify symptom control problems in this population is needed. The objective of this series of studies was to develop and undertake preliminary testing of a PRO specifically designed to identify symptom burden and functional issues in the population with RMHNC.
Methods Phase I: Tool Development Item generation. To identify appropriate content, a thorough and comprehensive literature review was performed, including 1
Department of Medicine, Vanderbilt University, Nashville, TN, USA School of Nursing, Vanderbilt University, Nashville, TN, USA 3 Department of Biostatistics, Schools of Medicine and Nursing, Vanderbilt University, Nashville, TN, USA 2
Corresponding Author: Leanne K. Jackson, MD, Department of Medicine, Vanderbilt University, 9402 The Vanderbilt Clinic, Nashville, TN 37232, USA. Email: [email protected]
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
2 consideration of items in other tumor-specific instruments, namely, Functional Assessment of Cancer Therapy—Head and Neck (FACT-H&N),7 MD Anderson Symptom Inventory8— Head and Neck (MDASI-HN), Vanderbilt Head and Neck Symptom Survey (VHNSS) version 2.0,6 University of Washington Quality of Life Questionnaire9 (UWQOL), European Organization for Research into the Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ)-C30, and EORTC QLQ-H&N35.10 To capture the symptom burden in this patient population, a focus group was conducted with 14 providers who care for patients with HNC including physicians (medical oncology and radiation oncology), nurses, lymphedema specialists, a physical therapist, nutritionist, case manager, and financial aid advisor. In addition, 17 one-on-one interviews were performed with patients having RMHNC. A comprehensive list of potential symptoms secondary to RM disease was generated which included 104 physical symptoms and 63 psychosocial items. To assess content validity of each item, the expert panel of providers rated items based on clinical importance allowing for deletion of low-priority items. Items that were rated ‘‘important’’ by more than half of the providers were retained. Additionally items that addressed similar constructs were combined. Finally, 46 physical symptoms and 13 psychosocial issues were retained. Assessment of patient item endorsement frequency and impact analysis. All retained items were reviewed included in a 2-step card sort. The goal of the card sort was to identify high frequency and high impact items. Twenty patients were given a stack of cards with a potential disease-related symptom written on each card. During the first pass, the patients were asked to sort the cards into 2 stacks, that is, one of the symptoms they had experienced as a result of RM disease and another of symptoms they had not experienced. During pass 2, patients were asked to place their endorsed symptoms in the order of severity. Patients were given the opportunity to write in additional symptoms not included in the original card sort. Questionnaire development. Items that were rated as either high impact (ranked in top 3 for severity) and/or were endorsed by 35% of patients on the card sort were included in the final questionnaire. Based on these criteria, the final questionnaire includes 35 physical symptoms and 12 psychosocial items. A scaling system was used that scored each item on a Likerttype scale from 0 (none) to 10 (severe).
Phase II: Pilot Testing of the VHNSS Recurrent/ Metastatic The final version of the VHNSS Recurrent/Metastatic (VHNSS-RM) underwent pilot testing in a cross-sectional cohort of 50 consecutive patients with noncurable RMHNC at the Vanderbilt Ingram Cancer Center. Quantitative outcomes that were assessed included completion time, missed items, scoring range, frequency and severity of individual items for the purpose of item reduction as well as to provide preliminary
data on the prevalence and severity of symptom burden in this cohort of patients. Qualitative outcomes included ease of use, clarity, and content validity. Study data were collected and managed using Research Electronic Data Capture (REDCap) electronic data capture tools hosted at Vanderbilt University.11 The REDCap is a secure, Web-based application designed to support data capture for research studies, providing (1) an intuitive interface for validated data entry, (2) audit trails for tracking data manipulation and export procedures, (3) automated export procedures for seamless data downloads to common statistical packages, and (4) procedures for importing data from external sources. Approval for both phases was obtained from the Vanderbilt institutional review board before study recruitment. Patient eligibility included noncurable recurrent, or metastatic patients with HNC, age >18 years, English speaking, and able to provide written informed consent. All the participants signed informed consent. Patients were instructed to report any concerns or questions regarding the tool and/or its administration to the study staff. Patients participating in phase 2 completed a demographic questionnaire. Data analysis. Demographic and clinical nominal or ordinal characteristics of the sample were summarized using frequency distributions. Median and 25th to 75th interquartile range (IQR) summarized continuous variables. Frequency distributions of the responses to each of the VHNSS-RM items were generated for clinically derived groupings of severity: none (0), mild (1-4), moderate (5-6), and severe (7-10). Conceptually derived domains of items and single-item symptom severity values were summarized using median and IQR.
Results Phase I: Tool development From July 2012 to March 2013, a literature review, a focus group, one-on-one patient interviews, and expert panel discussion identified 104 physical symptoms and 63 psychosocial issues. Items with similar constructs were combined. Subsequently, the expert panel reviewed and scored each item based on clinical importance for inclusion in the tool. Forty-six physical symptoms and 13 psychosocial issues were retained and advanced to the next phase of development. Twenty patients participated in the 2-step card sort. Items endorsed by 35% of patients and those rated as high impact was retained yielding 35 physical symptoms and 12 psychosocial issues. Comparison with existing tools6-9 highlighted 12 unique physical symptoms and 7 unique psychosocial issues. Each item was converted to a question that was phrased to yield information with clinical impact, to capture patients’ self-reported symptom burden as well as enrich providers’ understanding of the symptom burden, and to allow for better clinical decision making. Unique physical symptoms included diet change due to mouth swelling, tongue movement affecting speech or eating/swallowing, swelling of face or neck, cramping in neck or jaw, bad breath,
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
3
drooling, wound complications: drainage/pain/odor; nasal congestion or sinus drainage, eyes watering, numbness of face/tongue/ear/scalp, headaches, and confusion. Unique psychosocial issues included perception of burden to family/friends, lost independence, fear, embarrassment, mood swings, stress, and boredom.
Phase II: Preliminary Testing The primary outcomes were assessment of patient acceptability, feasibility of administration, and completion time. Secondary goals included determination of response frequency and range for each item with the intent of deleting items with low severity or response frequency. Demographic and treatment characteristics of the study sample are included in Table 1. There were 32 (64.0%) males, most were white (n ¼ 42, 84.0%), with a median age of 59.0 years. Patients were accrued between April 2013 and March 2014. Sixty-three percent reported a smoking history. The most common primary tumor sites were oropharynx (n ¼ 19, 38.0%), followed by oral cavity and salivary gland, each with 8 patients, 16%. There were no barriers to completion of the tool via computer interface. Patients found the tool acceptable, with high feasibility and readability. Completion time was a median of 9 minutes (IQR: 6-11.5 minutes, n ¼ 49). A full range of possible responses (0-10) was noted for all questions. Five items had a ‘‘not applicable (N/A): percutaneous endoscopic gastrostomy tube only’’ option. The maximum frequency with which N/A was chosen was n ¼ 4 for questions pertaining to taste, swallowing, and diet changes. The frequency and severity of physical symptoms and psychosocial issues are shown in Table 2. As noted, symptom burden was high. Based on prior work with the VHNSS version 2.0, as well as expert opinion, groups of items for hypothesized domains of symptoms were combined into a single score and evaluated for internal consistency using the Cronbach a statistic (see Table 3 for domain component items). Three of the domains, nutrition/ swallowing, systemic, and psychosocial, had excellent internal consistency (a 0.9) and 5 had good internal consistency (0.7 a < 0.9), speech/communication, musculoskeletal, secretions, respiratory/nasal, and neurologic. Of the 9 symptom domain scores, 8 demonstrated acceptable levels of internal consistency (0.75). The wound domain was slightly lower at 0.69 however considered acceptable for this preliminary small sample study. Pain, headache, xerostomia, and lacrimation were individual items not thought fit into any other domains.
Preliminary Analysis of Symptom Frequency and Severity Based on Domains Nutrition/swallowing domain. Severe symptoms (VHNSS score 7-10) were noted in 20% to 36.2% of patients for all questions excluding difficulty swallowing thin liquids (n ¼ 9, 19.1%) and presence of nausea (n ¼ 4, 8%). However, 8 (16%) patients had moderate nausea (see Table 2).
Table 1. Demographics and Clinical characteristics.a Age, median (IQR) Gender, N (%) Male Female Race, N (%) White African American or Black Ethnicity (N ¼ 48), N (%) Hispanic or Latino Not Hispanic or Latino Marital status (N ¼ 49), N (%) Single Married Widowed Living with partner Other Highest grade of education completed (N ¼ 47), median (IQR) Number of people living in home (N ¼ 47), median (IQR) Employment (N ¼ 47), N (%) Employed full time Employed part time Unemployed Homemaker Retired On disability Other Yearly household income, N (%) US$10 000 or less US$10 001 to US$20 000 US$20 001 to US$30 000 US$40 001 to US$50 000 US$50 001 to US$60 000 Over US$60 000 Do not care to respond Primary financial support for household, N (%) No Yes Health insurance (N ¼ 49), N (%) No Yes Dental insurance (N ¼ 46), N (%) No Yes Currently drink alcohol, N (%) No, never drank No, previous drinker Yes Ever smoked (N ¼ 49), N (%) No Yes Years smoked (N ¼ 30), mean (SD) Currently smoke (N ¼ 48), N (%) No Yes Time since diagnosis (N ¼ 43), mean (SD), months Primary site, N (%) Oropharynx Oral cavity
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
59.0 (55-69) 32 (64.0) 18 (36.0) 42 (84.0) 8 (16.0) 1 (2.1) 47 (97.9) 10 (20.0) 29 (59.2) 6 (12.2) 2 (4.1) 2 (4.1) 12.0 (12-14) 2.0 (2-2)
6 (12.8) 3 (6.4) 3 (6.4) 2 (4.3) 14 (29.8) 16 (34.0) 3 (6.4) 11 (22.9) 4 (8.3) 3 (6.3) 2 (4.2) 2 (4.2) 12 (25.0) 14 (29.2) 21 (44.7) 26 (55.3) 7 (14.3) 42 (85.7) 29 (63.0) 17 (37.0) 15 (30.0) 26 (52.0) 9 (18.0) 18 (36.0) 31 (63.3) 27.4 (14.2) 41 (85.4) 7 (14.6) 21.2 (12-93) 19 (38.0) 8 (16.0) (continued)
American Journal of Hospice & Palliative Medicine®
4 Table 1. (continued) Salivary gland Larynx Nasal cavity and paranasal sinuses Hypopharynx Nasopharynx Other Initial chemotherapy (N ¼ 48), N (%) None Induction Concurrent Induction plus concurrent Chemotherapy for recurrent disease (N ¼ 41), N (%) None Induction Concurrent Induction plus concurrent Initial radiation (N ¼ 47), N (%) No Yes Radiation for recurrent disease (N ¼ 41), N (%) No Yes
8 5 4 2 1 3
(16.0) (10.0) (8.0) (4.0) (2.0) (6.0)
14 7 20 7
(29.2) (14.6) (41.7) (14.6)
10 27 3 1
(24.4) (65.9) (7.3) (2.4)
Headache. Severe headaches were experienced by 8 (16%) of the patients. Neurologic. Severe numbness was seen most commonly in the face (n ¼ 13, 26%) and tongue (n ¼ 12, 24%), followed by the scalp (n ¼ 7, 14.3%) and ear (n ¼ 6, 12.2%). Nine (18.8%) patients experienced hearing loss as a result of their cancer. Systemic. Between 30% and 36.7% of patients had severe generalized weakness, decreased energy, and difficulty with sleep. Severe confusion was found in 8 (16%) patients, severe impairment in thinking/concentration in 9 (18%) patients, and memory deficiencies in 11 (22%) patients.
7 (14.9) 40 (85.1)
Psychosocial. In all, 18% to 26.5% of patients experienced severe burden for 10 of the 12 questions. The least troubling psychosocial issues were severe embarrassment (n ¼ 7, 14%) and severe problems relating to family/friends/partner (n ¼ 5, 10%).
31 (75.6) 10 (24.4)
Discussion
Abbreviations: IQR, interquartile range; SD, standard deviation. a N ¼ 50.
Speech/communication domain. Severe symptoms were found in 30% to 36.7% of patients for all questions excluding decreased tongue movement making it hard to speak for which 12 (24%) of patients had severe difficulty and 8 (16%) had moderate difficulty. Musculoskeletal. Swelling in the face/neck and decreased movement of the neck, shoulders, and jaw were severe in 30.6% to 34% of patients, while severe cramps in the neck were seen in 8 (16%) patients and 7 (14%) patients had severe cramps in the jaw. Wound. Our sample contained 8 of the 50 patients who answered in the affirmative to at least 1 wound question. One (2%) had a wound that had severe drainage/bleeding and 3 (6%) had wounds that were severely painful or malodorous. Eleven (22%) of the patients reported severe bad breath. Secretions. Of the patients, 11 (22%) had severe drooling and 17 (34.7%) had increased mucus/saliva. Xerostomia. In all, 19 (38%) had severe dry lips/mouth. Respiratory/nasal. Of the patients, 10% to 14% had cough, shortness of breath, nasal stuffiness/congestion, or sinus drainage. Mild symptoms were reported for 26% to 34% of patients for this domain of questions. Lacrimation. Seven (14%) of the patients had severe eye watering. Pain. Severe pain was experienced by 15 (30%) of the patients
We identified a need for a tool to capture symptom burden and functional impairment in patients with RMHNC. The tool was developed for use by researchers and providers treating and managing patients with metastatic or noncurable recurrent HNC. One-on-one patient interviews and a focus group with providers experienced in the care of this specific patients with HNC population helped ensure relevant content selection. The card sort method helped ensure that all items were either highfrequency or high-impact symptoms. All questions responses are scored on a Likert-type scale from 0 (none) to 10 (severe) making it easy for patients to complete and for health care providers to interpret. The tool was completed in an online format in less than 11.5 minutes. Thus, our results support that the VHNSS-RM is a feasible, brief, and easy-to-use tool to identify symptom burden in this cohort of patients. The number of patient in the pilot test was small but sufficient to allow a preliminary domain analysis to provide a conceptual understanding of the symptom domains in this cohort of patients. The results of this analysis demonstrated 9 domains and 3 single items. The domains demonstrated high internal consistency. The exception was the wound domain, which was composed of 2 items: ‘‘I have bad breath’’ and ‘‘I have a wound that drains or bleeds/is painful/smells bad.’’ Originally, it was hypothesized that bad breath in this cohort of patients was secondary to necrotic or superinfected recurrent cancer involving the mucosa and soft tissues. The preliminary results of the domain analysis bring up the possibility that bad breath and wound care issues address distinct issues and may need to be considered single items. That being said there were only 8 patients who answered affirmative to the wound questions. Thus a larger sample size is needed to better understand how these items relate. Further testing is planned to confirm the results of the conceptual domain analysis. We conducted a content analysis comparing commonly used PROs (FACT, EORTC, VHNSS, MDASI, and UWQOL) in the
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
5
Table 2. Vanderbilt Head and Neck Symptom Survey Recurrent Metastatic: Frequency and Severity.a Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) question
N
None: 0, n (%)
Mild: 1-4, n (%)
Moderate: 5-6, n (%)
Severe: 7-10, n (%)
50 46 46 48 48 47 47 50 50 49 49 50 49
17 (34.0) 26 (56.6) 28 (60.9) 21 (43.8) 21 (43.8) 14 (29.8) 25 (53.2) 12 (24.0) 26 (52.0) 17 (34.7) 24 (49.0) 27 (54.0) 20 (40.8)
10 (20.0) 4 (8.7) 3 (6.5) 12 (25.0) 4 (8.3) 8 (17.0) 10 (21.3) 14 (28.0) 12 (24.0) 8 (16.3) 7 (14.3) 6 (12.0) 10 (20.4)
10 (20.0) 4 (8.7) 5 (10.9) 1 (2.1) 8 (16.7) 8 (17.0) 3 (6.4) 8 (16.0) 8 (16.0) 6 (12.2) 2 (4.1) 2 (4.0) 4 (8.2)
13 (26.0) 12 (26.1) 10 (21.7) 14 (29.2) 15 (31.3) 17 (36.2) 9 (19.1) 16 (32.0) 4 (8.0) 18 (36.7) 16 (32.7) 15 (30.0) 15 (30.6)
50
22 (44.0)
7 (14.0)
7 (14.0)
14 (28.0)
50 49 50
23 (46.0) 18 (36.7) 15 (30.0)
7 (14.0) 10 (20.4) 11 (22.0)
8 (16.0) 6 (12.2) 8 (16.0)
12 (24.0) 15 (30.6) 16 (32.0)
50
16 (32.0)
11 (22.0)
6 (12.0)
17 (34.0)
50 50 50 50 49 50 50 50 50 50 50 50 50 50 50 48 50 50 49 49 49 50 50 50 50 50 50 50
30 (60.0) 28 (56.0) 30 (60.0) 22 (44.0) 17 (34.7) 12 (24.0) 45 (90.0) 42 (84.0) 45 (90.0) 22 (44.0) 26 (52.0) 19 (38.0) 18 (36.0) 27 (54.0) 27 (54.0) 31 (64.6) 24 (48.0) 32 (64.0) 33 (67.3) 35 (71.4) 20 (40.8) 28 (56.0) 26 (52.0) 23 (46.0) 14 (28.0) 9 (18.0) 17 (34.0) 11 (22.0)
9 (18.0) 13 (26.0) 5 (10.0) 10 (20.0) 8 (16.3) 10 (20.0) 2 (4.0) 3 (6.0) 1 (2.0) 13 (26.0) 16 (32.0) 17 (34.0) 17 (34.0) 12 (24.0) 10 (20.0) 8 (16.7) 8 (16.0) 3 (6.0) 7 (14.3) 6 (12.2) 7 (14.3) 8 (16.0) 9 (18.0) 15 (30.0) 16 (32.0) 17 (34.0) 11 (22.0) 17 (34.0)
3 (6.0) 2 (4.0) 4 (8.0) 7 (14.0) 7 (14.3) 9 (18.0) 2 (4.0) 2 (4.0) 1 (2.0) 9 (18.0) 3 (6.0) 7 (14.0) 8 (16.0) 4 (8.0) 5 (10.0) 0 (0.0) 5 (10.0) 3 (6.0) 3 (6.1) 1 (2.0) 4 (8.2) 5 (10.0) 4 (8.0) 4 (8.0) 5 (10.0) 8 (16.0) 7 (14.0) 4 (8.0)
8 (16.0) 7 (14.0) 11 (22.0) 11 (22.0) 17 (34.7) 19 (38.0) 1 (2.0) 3 (6.0) 3 (6.0) 6 (12.0) 5 (10.0) 7 (14.0) 7 (14.0) 7 (14.0) 8 (16.0) 9 (18.8) 13 (26.0) 12 (24.0) 6 (12.2) 7 (14.3) 18 (36.7) 9 (18.0) 11 (22.0) 8 (16.0) 15 (30.0) 16 (32.0) 15 (30.0) 18 (36.0)
50 50 50 50 50
28 (56.0) 30 (60.0) 27 (54.0) 23 (46.0) 20 (40.0)
9 (18.0) 12 (24.0) 11 (22.0) 12 (24.0) 17 (34.0)
2 (4.0) 3 (6.0) 2 (4.0) 6 (12.0) 0 (0.0)
11 (22.0) 5 (10.0) 10 (20.0) 9 (18.0) 13 (26.0)
Symptoms 1 2a 2b 3 4 5 6 7 8 9 10 11a 11b 12a 12b 13 14a 14b 15a 15b 16 17 18 19 20a 20b 20c 21 22 23a 23b 24 25 26 27a 27b 27c 27d 28 29 30 31 32 33 34 35
My appetite is _____ My diet has changed because of PAIN in my mouthb My diet has changed because of SWELLING in my mouthb My taste has changedb My tumor makes it difficult to swallowb I have difficulty swallowing SOLID foodsb I have difficulty swallowing THIN liquidsb I have lost weight I have nausea My voice has changed It is difficult for me to communicate I cannot talk for a long time because of PAIN in my mouth or throat I cannot talk for a long time because of DRYNESS in my mouth or throat Decreased movement of my tongue makes it hard to EAT or SWALLOW Decreased movement of my tongue makes it hard to SPEAK I have swelling in my face or neck I have decreased movement or increased stiffness in the tissues of my NECK or SHOULDERS I have decreased movement or increased stiffness in the tissues of my JAW I get cramps in my NECK I get cramps in my JAW I have bad breath I have problems with drooling I have increased mucus or saliva My mouth or lips are dry I have a wound that drains or bleeds I have a wound that is painful I have a wound that smells bad I have a cough I feel short of breath I have nasal stuffiness or congestion I have sinus drainage My eyes water I have frequent headaches I have hearing loss from my cancer I have numbness of my FACE I have numbness of my TONGUE I have numbness of my EAR I have numbness of my SCALP I have difficulty sleeping or falling asleep I have problems with my thinking or concentration I have problems with my short or long term memory I sometimes get confused I feel tired all the time I have decreased energy My CURRENT pain level is: I feel weak in general Psychosocial issues: because of my disease . . .
1 2 3 4 5
I am a burden to family or friends I have problems relating to my partner/family/friends I have lost my independence I avoid social situations I feel depressed
(continued) Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
6 Table 2. (continued)
6 7 8 9 10 11 12
Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) question
N
None: 0, n (%)
Mild: 1-4, n (%)
Moderate: 5-6, n (%)
Severe: 7-10, n (%)
I feel anxious I feel sad I feel fearful I have mood swings I feel stressed I feel bored I feel embarrassed
50 48 50 50 50 49 50
22 (44.0) 19 (39.6) 25 (50.0) 18 (36.0) 18 (36.0) 22 (44.9) 30 (60.0)
12 (24.0) 15 (31.3) 11 (22.0) 13 (26.0) 14 (28.0) 8 (16.3) 6 (12.0)
6 (12.0) 5 (10.4) 5 (10.0) 8 (16.0) 7 (14.0) 6 (12.2) 7 (14.0)
10 (20.0) 9 (18.8) 9 (18.0) 11 (22.0) 11 (22.0) 13 (26.5) 7 (14.0)
Abbreviation: PEG, percutaneous endoscopic gastrostomy. a Bolded questions represent unique items. b Patients had the option to choose: not applicable: PEG tube only.
Table 3. Conceptual Domains: VHNSS-RM Domain Item Score Reliability and Summaries.a Domain Nutrition and swallowing Speech and communication Musculoskeletal Secretions Xerostomia Wound Respiratory and nasal Lacrimation Frequent headaches Neurologic Pain Systemic Psychosocial
Number items
Question numbers
Cronbach a
10 5 5 2 1 4 4 1 1 5 1 7 12
1-8, 12a 9-11, 12b 13-15 17, 18 19 16, 20 21-23 24 25 26, 27 34 28-33, 35
0.90 0.85 0.87 0.80 – 0.69 0.75 – – 0.88 – 0.94 0.94
Median [IQR] (min, max) 3.2 [1-6] 3.0 [0-7] 3.4 [0-6] 3.75 [0-7] 5.00 [0-9] 0.0 [0-3] 2.3 [0-4] 0.0 [0-4] 0.0 [0-5] 1.0 [0-4] 3.0 [0-7] 2.3 [1-6] 1.7 [0-5]
(0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 9) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10)
Abbreviations: IQR, interquartile range; min, minimum; max, maximum; VHNSS-RM, Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic. a N ¼ 50.
HNC population with the VHNSS-RM to assess the utility and relevance of available tools for patients with RMHNC. Many of the symptoms that were graded as moderate to severe are related to residual effects of primary therapy, thus, these symptoms are captured by currently existing tools directed at the acute and late effects of HNC therapy. However, the content analysis demonstrated that the VHNSS-RM included 19 items that are unique to the patient population with RMHNC (see Table 2) and are not included in the other tools that were examined. Furthermore, items found on other tools were not incorporated into the VHNSS-RM, as they likely reflect acute effects of primary therapy and have therefore resolved. Because the VHNSS-RM items were generated through a rigorous process of patient and expert interview, the item base is felt to be more reflective of patient experience and provides a comprehensive picture of patient symptom burden. Through this study, we were able to develop a conceptual framework for symptom burden in patients with RMHNC. There are 3 critical sources of symptoms in the population with RMHNC, namely, residual symptoms from prior therapy, tumor-related symptoms, and side effects from ongoing therapy (see Figure 1). The vast majority of patients who develop RM disease were initially treated with single modality or
combined modality therapy for localized disease. Most patients have some degree of residual toxicities from primary treatment which fail to resolve over time and may potentially increase in severity. In the setting of RM disease, the residual toxicities of primary treatment are compounded by symptom burden secondary to recurrent cancer. One of the primary goals of therapy in patients with incurable disease is to decrease tumor-related symptoms. For example, systemic therapy may result in tumor shrinkage and subsequent healing of an open wound. Unfortunately, treatment itself is associated with symptoms burden due to side effects. When treatment is effective at ameliorating tumor-related symptoms and side effects are minimal, the benefit to patients is clear. It also clear that treatment is not of benefit when there is both disease progression and significant toxicity from treatment. The major quandary for patients and clinicians is when there is a mixed response. For example, if the cancer responds to therapy but there is substantial treatment toxicity, the risks of therapy may outweigh the benefits. Under such circumstances, the decision to continue therapy must be carefully considered. Valid and reliable measures of symptom burden from patients’ perspectives may enhance decision making in these difficult circumstances. Furthermore, such tools may alert the clinician to potentially treatable symptom control
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
7
Figure 1. Conceptual framework: symptom burden in patients with recurrent–metastatic head and neck cancer.
issues. We hypothesize that the VHNSS-RM could play a role in patient education and early symptom control intervention. In addition, minimizing symptom burden might allow a patient to continue receiving therapy, possibly improving outcomes, including survival.12 Determining the prevalence of previously unrecognized symptoms such as watery eyes or nasal congestion could allow for better understanding of the cause of these symptoms and potentially improve symptom treatment strategies. For example, watery eyes and nasal congestion might be attributable to mass effect from tumor or to the presence of lymphedema.
Further testing is planned in order for formally establishing the reliability, validity, and responsiveness to change over time. Authors’ Note The article was orally presented at the Multinational Association of Supportive Care in Cancer (MASCC), Miami, FL, June 2014.
Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
Limitations The number of patients included in the pilot testing was small, thus, we were unable to do a sufficiently powered cluster analysis. The analysis only allows for an initial conceptualization of the item domains. Formal analysis with larger numbers of patients is needed to confirm our preliminary results.
Conclusions The VHNSS-RM is a promising tool for use in patients with HNC having RM disease. It has potential to help improve our understanding of the frequency and severity of symptom burden and functional deficits in this population. It may also prove to be an effective screening and ongoing assessment tool to identify issues that require intervention and/or education.
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a fellows grant from Bristol-Myers Squibb Corporation (VUMC39969-R). REDCap is supported by Vanderbilt Institute for Clinical and Translational Research grant UL1 TR000445 from NCATS/NI.
References 1. Murphy BA. Advances in quality of life and symptom management for head and neck cancer patients. Curr Opin Oncol. 2009;21(3):242-247. 2. Kolnick L, Deng J, Epstein JB, et al. Associations of oral health items of the Vanderbilt Head and Neck Symptom Survey with a dental health assessment. Oral Oncol. 2014;50(2):135-140. 3. Murphy BA, Gilbert J, Cmelak A, Ridner SH. Symptom control issues and supportive care of patients with head and neck cancers. Clin Adv Hematol Oncol. 2007;5(10):807-822.
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
8 4. Murphy BA, Gilbert J, Ridner SH. Systemic and global toxicities of head and neck treatment. Expert Rev Anticancer Ther. 2007; 7(7):1043-1053. 5. Bentzen SM, Trotti A. Evaluation of early and late toxicities in chemoradiation trials. J Clin Oncol. 2007;25(26):4096-4103. 6. Cooperstein E, Gilbert J, Epstein JB, et al. Vanderbilt head and neck symptom survey version 2.0: report of the development and initial testing of a subscale for assessment of oral health. Head Neck. 2011;34(6):797-804. 7. List MA, D’Antonio LL, Cella DF, et al. The performance status scale for head and neck cancer patients and the functional assessment of cancer therapy-head and neck scale. A study of utility and validity. Cancer. 1996;77(11):2294-2301. 8. Rosenthal DI, Mendoza TR, Chambers MS, et al. Measuring head and neck cancer symptom burden: the development and validation of the M. D. Anderson symptom inventory, head and neck module. Head Neck. 2007;29(10):923-931.
9. Rogers SN, Gwanne S, Lowe D, Humphris G, Yueh B, Weymuller EA Jr. The addition of mood and anxiety domains to the University of Washington quality of life scale. Head Neck. 2002; 24(6):521-529. 10. Osthus AA, Aarstad AK, Olofsson J, Aarstad HJ. Prediction of survival by pretreatment health-related quality-of-life scores in a prospective cohort of patients with head and neck squamous cell carcinoma. JAMA Otolaryngol Head Neck Surg. 2013;139(1): 14-20. 11. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2): 377-381. 12. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med. 2010;363(8):733-742.
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Preliminary Testing of a Patient-Reported Outcome Measure for Recurrent or Metastatic Head and Neck Cancer
American Journal of Hospice & Palliative Medicine® 1-8 ª The Author(s) 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1049909115569591 ajhpm.sagepub.com
Leanne K. Jackson, MD1, Jie Deng, PhD2, Sheila H. Ridner, PhD2, Jill Gilbert, MD1, Mary S. Dietrich, PhD3, and Barbara A. Murphy, MD1
Abstract Objective: We describe development and preliminary testing of Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) to assess residual symptoms, tumor-related symptoms, and side effects from therapy. Methods: Items were identified through patient and provider interviews. Card sort selected high-yield and high-impact items. The VHNSS-RM was administered to 50 patients with recurrent/metastatic head and neck cancer (RMHNC). Results: The VHNSS-RM includes 12 unique symptoms (diet change, tongue movement affecting speech/swallowing, face/neck swelling, neck/jaw cramping, bad breath, drooling, wound drainage/pain/odor, nasal congestion/drainage, eyes watering, face/tongue/ear/scalp numbness, headaches, and confusion) and 7 unique psychosocial issues (burden to family/friends, lost independence, fear, embarrassment, mood swings, stress, and boredom). Conclusions: The VHNSS-RM contains 35 physical and 12 psychosocial issues. The VHNSS-RM is feasible and not overly burdensome. Nineteen unique items may improve palliation to patients with RMHNC. Keywords recurrent head and neck cancer, metastatic head and neck cancer, patient-reported outcomes, symptom burden, toxicities, unique symptoms
Introduction Head and neck cancer (HNC) and its treatment are associated with substantial side effects resulting in symptom burden, functional deficits, and decreased quality of life. Acute toxicities due to surgical or radiation-based therapies for HNC are well described. Until recently, late effects have been poorly studied; however, increasingly there have been efforts to systemically identify late effects of toxicity and their impact.1-4 Conversely, data on the prevalence and severity of toxicities for patients with noncurable recurrent and/or metastatic (RM) HNC have been and remain limited. Efforts to document supportive care issues in this population remain scarce, and there are no patient-reported outcome (PRO) measures specifically developed for and tested in this population.5,6 A multitude of tools have been developed and tested, which focus on acute and/or late effects of primary therapy in HNC.6-10 Most of these tools are dominated by items directed at acute effects of radiation-based therapies for HNC. No PROs have been developed to assess symptom burden and functional deficits in patients with RMHNC. We postulate that the symptoms and functional problems experienced by patients in the RM setting are different than those noted during initial treatment. Thus, the use of PROs directed at the acute toxicities of surgery or radiation-
based primary therapy are inadequate. A PRO that can be used in the clinical and research settings to screen for and identify symptom control problems in this population is needed. The objective of this series of studies was to develop and undertake preliminary testing of a PRO specifically designed to identify symptom burden and functional issues in the population with RMHNC.
Methods Phase I: Tool Development Item generation. To identify appropriate content, a thorough and comprehensive literature review was performed, including 1
Department of Medicine, Vanderbilt University, Nashville, TN, USA School of Nursing, Vanderbilt University, Nashville, TN, USA 3 Department of Biostatistics, Schools of Medicine and Nursing, Vanderbilt University, Nashville, TN, USA 2
Corresponding Author: Leanne K. Jackson, MD, Department of Medicine, Vanderbilt University, 9402 The Vanderbilt Clinic, Nashville, TN 37232, USA. Email: [email protected]
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
2 consideration of items in other tumor-specific instruments, namely, Functional Assessment of Cancer Therapy—Head and Neck (FACT-H&N),7 MD Anderson Symptom Inventory8— Head and Neck (MDASI-HN), Vanderbilt Head and Neck Symptom Survey (VHNSS) version 2.0,6 University of Washington Quality of Life Questionnaire9 (UWQOL), European Organization for Research into the Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ)-C30, and EORTC QLQ-H&N35.10 To capture the symptom burden in this patient population, a focus group was conducted with 14 providers who care for patients with HNC including physicians (medical oncology and radiation oncology), nurses, lymphedema specialists, a physical therapist, nutritionist, case manager, and financial aid advisor. In addition, 17 one-on-one interviews were performed with patients having RMHNC. A comprehensive list of potential symptoms secondary to RM disease was generated which included 104 physical symptoms and 63 psychosocial items. To assess content validity of each item, the expert panel of providers rated items based on clinical importance allowing for deletion of low-priority items. Items that were rated ‘‘important’’ by more than half of the providers were retained. Additionally items that addressed similar constructs were combined. Finally, 46 physical symptoms and 13 psychosocial issues were retained. Assessment of patient item endorsement frequency and impact analysis. All retained items were reviewed included in a 2-step card sort. The goal of the card sort was to identify high frequency and high impact items. Twenty patients were given a stack of cards with a potential disease-related symptom written on each card. During the first pass, the patients were asked to sort the cards into 2 stacks, that is, one of the symptoms they had experienced as a result of RM disease and another of symptoms they had not experienced. During pass 2, patients were asked to place their endorsed symptoms in the order of severity. Patients were given the opportunity to write in additional symptoms not included in the original card sort. Questionnaire development. Items that were rated as either high impact (ranked in top 3 for severity) and/or were endorsed by 35% of patients on the card sort were included in the final questionnaire. Based on these criteria, the final questionnaire includes 35 physical symptoms and 12 psychosocial items. A scaling system was used that scored each item on a Likerttype scale from 0 (none) to 10 (severe).
Phase II: Pilot Testing of the VHNSS Recurrent/ Metastatic The final version of the VHNSS Recurrent/Metastatic (VHNSS-RM) underwent pilot testing in a cross-sectional cohort of 50 consecutive patients with noncurable RMHNC at the Vanderbilt Ingram Cancer Center. Quantitative outcomes that were assessed included completion time, missed items, scoring range, frequency and severity of individual items for the purpose of item reduction as well as to provide preliminary
data on the prevalence and severity of symptom burden in this cohort of patients. Qualitative outcomes included ease of use, clarity, and content validity. Study data were collected and managed using Research Electronic Data Capture (REDCap) electronic data capture tools hosted at Vanderbilt University.11 The REDCap is a secure, Web-based application designed to support data capture for research studies, providing (1) an intuitive interface for validated data entry, (2) audit trails for tracking data manipulation and export procedures, (3) automated export procedures for seamless data downloads to common statistical packages, and (4) procedures for importing data from external sources. Approval for both phases was obtained from the Vanderbilt institutional review board before study recruitment. Patient eligibility included noncurable recurrent, or metastatic patients with HNC, age >18 years, English speaking, and able to provide written informed consent. All the participants signed informed consent. Patients were instructed to report any concerns or questions regarding the tool and/or its administration to the study staff. Patients participating in phase 2 completed a demographic questionnaire. Data analysis. Demographic and clinical nominal or ordinal characteristics of the sample were summarized using frequency distributions. Median and 25th to 75th interquartile range (IQR) summarized continuous variables. Frequency distributions of the responses to each of the VHNSS-RM items were generated for clinically derived groupings of severity: none (0), mild (1-4), moderate (5-6), and severe (7-10). Conceptually derived domains of items and single-item symptom severity values were summarized using median and IQR.
Results Phase I: Tool development From July 2012 to March 2013, a literature review, a focus group, one-on-one patient interviews, and expert panel discussion identified 104 physical symptoms and 63 psychosocial issues. Items with similar constructs were combined. Subsequently, the expert panel reviewed and scored each item based on clinical importance for inclusion in the tool. Forty-six physical symptoms and 13 psychosocial issues were retained and advanced to the next phase of development. Twenty patients participated in the 2-step card sort. Items endorsed by 35% of patients and those rated as high impact was retained yielding 35 physical symptoms and 12 psychosocial issues. Comparison with existing tools6-9 highlighted 12 unique physical symptoms and 7 unique psychosocial issues. Each item was converted to a question that was phrased to yield information with clinical impact, to capture patients’ self-reported symptom burden as well as enrich providers’ understanding of the symptom burden, and to allow for better clinical decision making. Unique physical symptoms included diet change due to mouth swelling, tongue movement affecting speech or eating/swallowing, swelling of face or neck, cramping in neck or jaw, bad breath,
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
3
drooling, wound complications: drainage/pain/odor; nasal congestion or sinus drainage, eyes watering, numbness of face/tongue/ear/scalp, headaches, and confusion. Unique psychosocial issues included perception of burden to family/friends, lost independence, fear, embarrassment, mood swings, stress, and boredom.
Phase II: Preliminary Testing The primary outcomes were assessment of patient acceptability, feasibility of administration, and completion time. Secondary goals included determination of response frequency and range for each item with the intent of deleting items with low severity or response frequency. Demographic and treatment characteristics of the study sample are included in Table 1. There were 32 (64.0%) males, most were white (n ¼ 42, 84.0%), with a median age of 59.0 years. Patients were accrued between April 2013 and March 2014. Sixty-three percent reported a smoking history. The most common primary tumor sites were oropharynx (n ¼ 19, 38.0%), followed by oral cavity and salivary gland, each with 8 patients, 16%. There were no barriers to completion of the tool via computer interface. Patients found the tool acceptable, with high feasibility and readability. Completion time was a median of 9 minutes (IQR: 6-11.5 minutes, n ¼ 49). A full range of possible responses (0-10) was noted for all questions. Five items had a ‘‘not applicable (N/A): percutaneous endoscopic gastrostomy tube only’’ option. The maximum frequency with which N/A was chosen was n ¼ 4 for questions pertaining to taste, swallowing, and diet changes. The frequency and severity of physical symptoms and psychosocial issues are shown in Table 2. As noted, symptom burden was high. Based on prior work with the VHNSS version 2.0, as well as expert opinion, groups of items for hypothesized domains of symptoms were combined into a single score and evaluated for internal consistency using the Cronbach a statistic (see Table 3 for domain component items). Three of the domains, nutrition/ swallowing, systemic, and psychosocial, had excellent internal consistency (a 0.9) and 5 had good internal consistency (0.7 a < 0.9), speech/communication, musculoskeletal, secretions, respiratory/nasal, and neurologic. Of the 9 symptom domain scores, 8 demonstrated acceptable levels of internal consistency (0.75). The wound domain was slightly lower at 0.69 however considered acceptable for this preliminary small sample study. Pain, headache, xerostomia, and lacrimation were individual items not thought fit into any other domains.
Preliminary Analysis of Symptom Frequency and Severity Based on Domains Nutrition/swallowing domain. Severe symptoms (VHNSS score 7-10) were noted in 20% to 36.2% of patients for all questions excluding difficulty swallowing thin liquids (n ¼ 9, 19.1%) and presence of nausea (n ¼ 4, 8%). However, 8 (16%) patients had moderate nausea (see Table 2).
Table 1. Demographics and Clinical characteristics.a Age, median (IQR) Gender, N (%) Male Female Race, N (%) White African American or Black Ethnicity (N ¼ 48), N (%) Hispanic or Latino Not Hispanic or Latino Marital status (N ¼ 49), N (%) Single Married Widowed Living with partner Other Highest grade of education completed (N ¼ 47), median (IQR) Number of people living in home (N ¼ 47), median (IQR) Employment (N ¼ 47), N (%) Employed full time Employed part time Unemployed Homemaker Retired On disability Other Yearly household income, N (%) US$10 000 or less US$10 001 to US$20 000 US$20 001 to US$30 000 US$40 001 to US$50 000 US$50 001 to US$60 000 Over US$60 000 Do not care to respond Primary financial support for household, N (%) No Yes Health insurance (N ¼ 49), N (%) No Yes Dental insurance (N ¼ 46), N (%) No Yes Currently drink alcohol, N (%) No, never drank No, previous drinker Yes Ever smoked (N ¼ 49), N (%) No Yes Years smoked (N ¼ 30), mean (SD) Currently smoke (N ¼ 48), N (%) No Yes Time since diagnosis (N ¼ 43), mean (SD), months Primary site, N (%) Oropharynx Oral cavity
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
59.0 (55-69) 32 (64.0) 18 (36.0) 42 (84.0) 8 (16.0) 1 (2.1) 47 (97.9) 10 (20.0) 29 (59.2) 6 (12.2) 2 (4.1) 2 (4.1) 12.0 (12-14) 2.0 (2-2)
6 (12.8) 3 (6.4) 3 (6.4) 2 (4.3) 14 (29.8) 16 (34.0) 3 (6.4) 11 (22.9) 4 (8.3) 3 (6.3) 2 (4.2) 2 (4.2) 12 (25.0) 14 (29.2) 21 (44.7) 26 (55.3) 7 (14.3) 42 (85.7) 29 (63.0) 17 (37.0) 15 (30.0) 26 (52.0) 9 (18.0) 18 (36.0) 31 (63.3) 27.4 (14.2) 41 (85.4) 7 (14.6) 21.2 (12-93) 19 (38.0) 8 (16.0) (continued)
American Journal of Hospice & Palliative Medicine®
4 Table 1. (continued) Salivary gland Larynx Nasal cavity and paranasal sinuses Hypopharynx Nasopharynx Other Initial chemotherapy (N ¼ 48), N (%) None Induction Concurrent Induction plus concurrent Chemotherapy for recurrent disease (N ¼ 41), N (%) None Induction Concurrent Induction plus concurrent Initial radiation (N ¼ 47), N (%) No Yes Radiation for recurrent disease (N ¼ 41), N (%) No Yes
8 5 4 2 1 3
(16.0) (10.0) (8.0) (4.0) (2.0) (6.0)
14 7 20 7
(29.2) (14.6) (41.7) (14.6)
10 27 3 1
(24.4) (65.9) (7.3) (2.4)
Headache. Severe headaches were experienced by 8 (16%) of the patients. Neurologic. Severe numbness was seen most commonly in the face (n ¼ 13, 26%) and tongue (n ¼ 12, 24%), followed by the scalp (n ¼ 7, 14.3%) and ear (n ¼ 6, 12.2%). Nine (18.8%) patients experienced hearing loss as a result of their cancer. Systemic. Between 30% and 36.7% of patients had severe generalized weakness, decreased energy, and difficulty with sleep. Severe confusion was found in 8 (16%) patients, severe impairment in thinking/concentration in 9 (18%) patients, and memory deficiencies in 11 (22%) patients.
7 (14.9) 40 (85.1)
Psychosocial. In all, 18% to 26.5% of patients experienced severe burden for 10 of the 12 questions. The least troubling psychosocial issues were severe embarrassment (n ¼ 7, 14%) and severe problems relating to family/friends/partner (n ¼ 5, 10%).
31 (75.6) 10 (24.4)
Discussion
Abbreviations: IQR, interquartile range; SD, standard deviation. a N ¼ 50.
Speech/communication domain. Severe symptoms were found in 30% to 36.7% of patients for all questions excluding decreased tongue movement making it hard to speak for which 12 (24%) of patients had severe difficulty and 8 (16%) had moderate difficulty. Musculoskeletal. Swelling in the face/neck and decreased movement of the neck, shoulders, and jaw were severe in 30.6% to 34% of patients, while severe cramps in the neck were seen in 8 (16%) patients and 7 (14%) patients had severe cramps in the jaw. Wound. Our sample contained 8 of the 50 patients who answered in the affirmative to at least 1 wound question. One (2%) had a wound that had severe drainage/bleeding and 3 (6%) had wounds that were severely painful or malodorous. Eleven (22%) of the patients reported severe bad breath. Secretions. Of the patients, 11 (22%) had severe drooling and 17 (34.7%) had increased mucus/saliva. Xerostomia. In all, 19 (38%) had severe dry lips/mouth. Respiratory/nasal. Of the patients, 10% to 14% had cough, shortness of breath, nasal stuffiness/congestion, or sinus drainage. Mild symptoms were reported for 26% to 34% of patients for this domain of questions. Lacrimation. Seven (14%) of the patients had severe eye watering. Pain. Severe pain was experienced by 15 (30%) of the patients
We identified a need for a tool to capture symptom burden and functional impairment in patients with RMHNC. The tool was developed for use by researchers and providers treating and managing patients with metastatic or noncurable recurrent HNC. One-on-one patient interviews and a focus group with providers experienced in the care of this specific patients with HNC population helped ensure relevant content selection. The card sort method helped ensure that all items were either highfrequency or high-impact symptoms. All questions responses are scored on a Likert-type scale from 0 (none) to 10 (severe) making it easy for patients to complete and for health care providers to interpret. The tool was completed in an online format in less than 11.5 minutes. Thus, our results support that the VHNSS-RM is a feasible, brief, and easy-to-use tool to identify symptom burden in this cohort of patients. The number of patient in the pilot test was small but sufficient to allow a preliminary domain analysis to provide a conceptual understanding of the symptom domains in this cohort of patients. The results of this analysis demonstrated 9 domains and 3 single items. The domains demonstrated high internal consistency. The exception was the wound domain, which was composed of 2 items: ‘‘I have bad breath’’ and ‘‘I have a wound that drains or bleeds/is painful/smells bad.’’ Originally, it was hypothesized that bad breath in this cohort of patients was secondary to necrotic or superinfected recurrent cancer involving the mucosa and soft tissues. The preliminary results of the domain analysis bring up the possibility that bad breath and wound care issues address distinct issues and may need to be considered single items. That being said there were only 8 patients who answered affirmative to the wound questions. Thus a larger sample size is needed to better understand how these items relate. Further testing is planned to confirm the results of the conceptual domain analysis. We conducted a content analysis comparing commonly used PROs (FACT, EORTC, VHNSS, MDASI, and UWQOL) in the
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
5
Table 2. Vanderbilt Head and Neck Symptom Survey Recurrent Metastatic: Frequency and Severity.a Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) question
N
None: 0, n (%)
Mild: 1-4, n (%)
Moderate: 5-6, n (%)
Severe: 7-10, n (%)
50 46 46 48 48 47 47 50 50 49 49 50 49
17 (34.0) 26 (56.6) 28 (60.9) 21 (43.8) 21 (43.8) 14 (29.8) 25 (53.2) 12 (24.0) 26 (52.0) 17 (34.7) 24 (49.0) 27 (54.0) 20 (40.8)
10 (20.0) 4 (8.7) 3 (6.5) 12 (25.0) 4 (8.3) 8 (17.0) 10 (21.3) 14 (28.0) 12 (24.0) 8 (16.3) 7 (14.3) 6 (12.0) 10 (20.4)
10 (20.0) 4 (8.7) 5 (10.9) 1 (2.1) 8 (16.7) 8 (17.0) 3 (6.4) 8 (16.0) 8 (16.0) 6 (12.2) 2 (4.1) 2 (4.0) 4 (8.2)
13 (26.0) 12 (26.1) 10 (21.7) 14 (29.2) 15 (31.3) 17 (36.2) 9 (19.1) 16 (32.0) 4 (8.0) 18 (36.7) 16 (32.7) 15 (30.0) 15 (30.6)
50
22 (44.0)
7 (14.0)
7 (14.0)
14 (28.0)
50 49 50
23 (46.0) 18 (36.7) 15 (30.0)
7 (14.0) 10 (20.4) 11 (22.0)
8 (16.0) 6 (12.2) 8 (16.0)
12 (24.0) 15 (30.6) 16 (32.0)
50
16 (32.0)
11 (22.0)
6 (12.0)
17 (34.0)
50 50 50 50 49 50 50 50 50 50 50 50 50 50 50 48 50 50 49 49 49 50 50 50 50 50 50 50
30 (60.0) 28 (56.0) 30 (60.0) 22 (44.0) 17 (34.7) 12 (24.0) 45 (90.0) 42 (84.0) 45 (90.0) 22 (44.0) 26 (52.0) 19 (38.0) 18 (36.0) 27 (54.0) 27 (54.0) 31 (64.6) 24 (48.0) 32 (64.0) 33 (67.3) 35 (71.4) 20 (40.8) 28 (56.0) 26 (52.0) 23 (46.0) 14 (28.0) 9 (18.0) 17 (34.0) 11 (22.0)
9 (18.0) 13 (26.0) 5 (10.0) 10 (20.0) 8 (16.3) 10 (20.0) 2 (4.0) 3 (6.0) 1 (2.0) 13 (26.0) 16 (32.0) 17 (34.0) 17 (34.0) 12 (24.0) 10 (20.0) 8 (16.7) 8 (16.0) 3 (6.0) 7 (14.3) 6 (12.2) 7 (14.3) 8 (16.0) 9 (18.0) 15 (30.0) 16 (32.0) 17 (34.0) 11 (22.0) 17 (34.0)
3 (6.0) 2 (4.0) 4 (8.0) 7 (14.0) 7 (14.3) 9 (18.0) 2 (4.0) 2 (4.0) 1 (2.0) 9 (18.0) 3 (6.0) 7 (14.0) 8 (16.0) 4 (8.0) 5 (10.0) 0 (0.0) 5 (10.0) 3 (6.0) 3 (6.1) 1 (2.0) 4 (8.2) 5 (10.0) 4 (8.0) 4 (8.0) 5 (10.0) 8 (16.0) 7 (14.0) 4 (8.0)
8 (16.0) 7 (14.0) 11 (22.0) 11 (22.0) 17 (34.7) 19 (38.0) 1 (2.0) 3 (6.0) 3 (6.0) 6 (12.0) 5 (10.0) 7 (14.0) 7 (14.0) 7 (14.0) 8 (16.0) 9 (18.8) 13 (26.0) 12 (24.0) 6 (12.2) 7 (14.3) 18 (36.7) 9 (18.0) 11 (22.0) 8 (16.0) 15 (30.0) 16 (32.0) 15 (30.0) 18 (36.0)
50 50 50 50 50
28 (56.0) 30 (60.0) 27 (54.0) 23 (46.0) 20 (40.0)
9 (18.0) 12 (24.0) 11 (22.0) 12 (24.0) 17 (34.0)
2 (4.0) 3 (6.0) 2 (4.0) 6 (12.0) 0 (0.0)
11 (22.0) 5 (10.0) 10 (20.0) 9 (18.0) 13 (26.0)
Symptoms 1 2a 2b 3 4 5 6 7 8 9 10 11a 11b 12a 12b 13 14a 14b 15a 15b 16 17 18 19 20a 20b 20c 21 22 23a 23b 24 25 26 27a 27b 27c 27d 28 29 30 31 32 33 34 35
My appetite is _____ My diet has changed because of PAIN in my mouthb My diet has changed because of SWELLING in my mouthb My taste has changedb My tumor makes it difficult to swallowb I have difficulty swallowing SOLID foodsb I have difficulty swallowing THIN liquidsb I have lost weight I have nausea My voice has changed It is difficult for me to communicate I cannot talk for a long time because of PAIN in my mouth or throat I cannot talk for a long time because of DRYNESS in my mouth or throat Decreased movement of my tongue makes it hard to EAT or SWALLOW Decreased movement of my tongue makes it hard to SPEAK I have swelling in my face or neck I have decreased movement or increased stiffness in the tissues of my NECK or SHOULDERS I have decreased movement or increased stiffness in the tissues of my JAW I get cramps in my NECK I get cramps in my JAW I have bad breath I have problems with drooling I have increased mucus or saliva My mouth or lips are dry I have a wound that drains or bleeds I have a wound that is painful I have a wound that smells bad I have a cough I feel short of breath I have nasal stuffiness or congestion I have sinus drainage My eyes water I have frequent headaches I have hearing loss from my cancer I have numbness of my FACE I have numbness of my TONGUE I have numbness of my EAR I have numbness of my SCALP I have difficulty sleeping or falling asleep I have problems with my thinking or concentration I have problems with my short or long term memory I sometimes get confused I feel tired all the time I have decreased energy My CURRENT pain level is: I feel weak in general Psychosocial issues: because of my disease . . .
1 2 3 4 5
I am a burden to family or friends I have problems relating to my partner/family/friends I have lost my independence I avoid social situations I feel depressed
(continued) Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
6 Table 2. (continued)
6 7 8 9 10 11 12
Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic (VHNSS-RM) question
N
None: 0, n (%)
Mild: 1-4, n (%)
Moderate: 5-6, n (%)
Severe: 7-10, n (%)
I feel anxious I feel sad I feel fearful I have mood swings I feel stressed I feel bored I feel embarrassed
50 48 50 50 50 49 50
22 (44.0) 19 (39.6) 25 (50.0) 18 (36.0) 18 (36.0) 22 (44.9) 30 (60.0)
12 (24.0) 15 (31.3) 11 (22.0) 13 (26.0) 14 (28.0) 8 (16.3) 6 (12.0)
6 (12.0) 5 (10.4) 5 (10.0) 8 (16.0) 7 (14.0) 6 (12.2) 7 (14.0)
10 (20.0) 9 (18.8) 9 (18.0) 11 (22.0) 11 (22.0) 13 (26.5) 7 (14.0)
Abbreviation: PEG, percutaneous endoscopic gastrostomy. a Bolded questions represent unique items. b Patients had the option to choose: not applicable: PEG tube only.
Table 3. Conceptual Domains: VHNSS-RM Domain Item Score Reliability and Summaries.a Domain Nutrition and swallowing Speech and communication Musculoskeletal Secretions Xerostomia Wound Respiratory and nasal Lacrimation Frequent headaches Neurologic Pain Systemic Psychosocial
Number items
Question numbers
Cronbach a
10 5 5 2 1 4 4 1 1 5 1 7 12
1-8, 12a 9-11, 12b 13-15 17, 18 19 16, 20 21-23 24 25 26, 27 34 28-33, 35
0.90 0.85 0.87 0.80 – 0.69 0.75 – – 0.88 – 0.94 0.94
Median [IQR] (min, max) 3.2 [1-6] 3.0 [0-7] 3.4 [0-6] 3.75 [0-7] 5.00 [0-9] 0.0 [0-3] 2.3 [0-4] 0.0 [0-4] 0.0 [0-5] 1.0 [0-4] 3.0 [0-7] 2.3 [1-6] 1.7 [0-5]
(0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 9) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10) (0, 10)
Abbreviations: IQR, interquartile range; min, minimum; max, maximum; VHNSS-RM, Vanderbilt Head and Neck Symptom Survey—Recurrent/Metastatic. a N ¼ 50.
HNC population with the VHNSS-RM to assess the utility and relevance of available tools for patients with RMHNC. Many of the symptoms that were graded as moderate to severe are related to residual effects of primary therapy, thus, these symptoms are captured by currently existing tools directed at the acute and late effects of HNC therapy. However, the content analysis demonstrated that the VHNSS-RM included 19 items that are unique to the patient population with RMHNC (see Table 2) and are not included in the other tools that were examined. Furthermore, items found on other tools were not incorporated into the VHNSS-RM, as they likely reflect acute effects of primary therapy and have therefore resolved. Because the VHNSS-RM items were generated through a rigorous process of patient and expert interview, the item base is felt to be more reflective of patient experience and provides a comprehensive picture of patient symptom burden. Through this study, we were able to develop a conceptual framework for symptom burden in patients with RMHNC. There are 3 critical sources of symptoms in the population with RMHNC, namely, residual symptoms from prior therapy, tumor-related symptoms, and side effects from ongoing therapy (see Figure 1). The vast majority of patients who develop RM disease were initially treated with single modality or
combined modality therapy for localized disease. Most patients have some degree of residual toxicities from primary treatment which fail to resolve over time and may potentially increase in severity. In the setting of RM disease, the residual toxicities of primary treatment are compounded by symptom burden secondary to recurrent cancer. One of the primary goals of therapy in patients with incurable disease is to decrease tumor-related symptoms. For example, systemic therapy may result in tumor shrinkage and subsequent healing of an open wound. Unfortunately, treatment itself is associated with symptoms burden due to side effects. When treatment is effective at ameliorating tumor-related symptoms and side effects are minimal, the benefit to patients is clear. It also clear that treatment is not of benefit when there is both disease progression and significant toxicity from treatment. The major quandary for patients and clinicians is when there is a mixed response. For example, if the cancer responds to therapy but there is substantial treatment toxicity, the risks of therapy may outweigh the benefits. Under such circumstances, the decision to continue therapy must be carefully considered. Valid and reliable measures of symptom burden from patients’ perspectives may enhance decision making in these difficult circumstances. Furthermore, such tools may alert the clinician to potentially treatable symptom control
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
Jackson et al
7
Figure 1. Conceptual framework: symptom burden in patients with recurrent–metastatic head and neck cancer.
issues. We hypothesize that the VHNSS-RM could play a role in patient education and early symptom control intervention. In addition, minimizing symptom burden might allow a patient to continue receiving therapy, possibly improving outcomes, including survival.12 Determining the prevalence of previously unrecognized symptoms such as watery eyes or nasal congestion could allow for better understanding of the cause of these symptoms and potentially improve symptom treatment strategies. For example, watery eyes and nasal congestion might be attributable to mass effect from tumor or to the presence of lymphedema.
Further testing is planned in order for formally establishing the reliability, validity, and responsiveness to change over time. Authors’ Note The article was orally presented at the Multinational Association of Supportive Care in Cancer (MASCC), Miami, FL, June 2014.
Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
Limitations The number of patients included in the pilot testing was small, thus, we were unable to do a sufficiently powered cluster analysis. The analysis only allows for an initial conceptualization of the item domains. Formal analysis with larger numbers of patients is needed to confirm our preliminary results.
Conclusions The VHNSS-RM is a promising tool for use in patients with HNC having RM disease. It has potential to help improve our understanding of the frequency and severity of symptom burden and functional deficits in this population. It may also prove to be an effective screening and ongoing assessment tool to identify issues that require intervention and/or education.
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a fellows grant from Bristol-Myers Squibb Corporation (VUMC39969-R). REDCap is supported by Vanderbilt Institute for Clinical and Translational Research grant UL1 TR000445 from NCATS/NI.
References 1. Murphy BA. Advances in quality of life and symptom management for head and neck cancer patients. Curr Opin Oncol. 2009;21(3):242-247. 2. Kolnick L, Deng J, Epstein JB, et al. Associations of oral health items of the Vanderbilt Head and Neck Symptom Survey with a dental health assessment. Oral Oncol. 2014;50(2):135-140. 3. Murphy BA, Gilbert J, Cmelak A, Ridner SH. Symptom control issues and supportive care of patients with head and neck cancers. Clin Adv Hematol Oncol. 2007;5(10):807-822.
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
American Journal of Hospice & Palliative Medicine®
8 4. Murphy BA, Gilbert J, Ridner SH. Systemic and global toxicities of head and neck treatment. Expert Rev Anticancer Ther. 2007; 7(7):1043-1053. 5. Bentzen SM, Trotti A. Evaluation of early and late toxicities in chemoradiation trials. J Clin Oncol. 2007;25(26):4096-4103. 6. Cooperstein E, Gilbert J, Epstein JB, et al. Vanderbilt head and neck symptom survey version 2.0: report of the development and initial testing of a subscale for assessment of oral health. Head Neck. 2011;34(6):797-804. 7. List MA, D’Antonio LL, Cella DF, et al. The performance status scale for head and neck cancer patients and the functional assessment of cancer therapy-head and neck scale. A study of utility and validity. Cancer. 1996;77(11):2294-2301. 8. Rosenthal DI, Mendoza TR, Chambers MS, et al. Measuring head and neck cancer symptom burden: the development and validation of the M. D. Anderson symptom inventory, head and neck module. Head Neck. 2007;29(10):923-931.
9. Rogers SN, Gwanne S, Lowe D, Humphris G, Yueh B, Weymuller EA Jr. The addition of mood and anxiety domains to the University of Washington quality of life scale. Head Neck. 2002; 24(6):521-529. 10. Osthus AA, Aarstad AK, Olofsson J, Aarstad HJ. Prediction of survival by pretreatment health-related quality-of-life scores in a prospective cohort of patients with head and neck squamous cell carcinoma. JAMA Otolaryngol Head Neck Surg. 2013;139(1): 14-20. 11. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2): 377-381. 12. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med. 2010;363(8):733-742.
Downloaded from ajh.sagepub.com at UNIV OF MICHIGAN on February 25, 2015
