Ann Surg Oncol DOI 10.1245/s10434-014-3656-z
ORIGINAL ARTICLE – HEAD AND NECK ONCOLOGY
Minimal Clinically Important Differences in Quality of Life Scores of Oral Cavity and Oropharynx Cancer Patients Yoav Binenbaum, MD1, Moran Amit, MD1,3, Salem Billan, MD2, Jacob T. Cohen, MD3, and Ziv Gil, MD, PhD1,3 Department of Otolaryngology Head and Neck Surgery, The Technion, Israel Institute of Technology, Haifa, Israel; 2The Radiology Institute, The Technion, Israel Institute of Technology, Haifa, Israel; 3The Clinical Research Institute at Rambam, Rambam Medical Center and the Rappaport Faculty of Medicine, The Technion Institute of Technology, Haifa, Israel 1
ABSTRACT Background. The minimal clinically important difference (MCID) is defined as the smallest difference in quality of life (QOL) that patients perceive as beneficial and that mandates a change in management. We aimed to determine the MCID among patients with oral cavity and oropharyngeal cancer and to identify domains that are significantly affected during treatment. Methods. The cohort consisted of 1,011 patients analyzed by a metaanalysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement. MCID values for the University of Washington Quality of Life Questionnaire (UW-QOLQ) and European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC C-30) and Head and Neck-35 questionnaires were calculated by using the distribution-based method. Results. The mean MCID for Global QOL was 13.07 points for the UW-QOLQ and 9.43 in the EORTC C-30 questionnaire. High consistency in the MCID values was found between the two questionnaires examined. Heat map analysis indicated a clinically significant improvement in head and neck-associated domains and in domains associated with general cancer treatment 1 year or more after treatment relative to 3 months after treatment (p \ 0.001
Electronic supplementary material The online version of this article (doi:10.1245/s10434-014-3656-z) contains supplementary material, which is available to authorized users. Ó Society of Surgical Oncology 2014 First Received: 31 October 2013 Z. Gil, MD, PhD e-mail: [email protected]
and p = 0.016, respectively). In contrast, improvement in general and functional domains was not evident 1 year or more after treatment (p = 0.69). Conclusions. This study suggests benchmark values for MCID and variation in QOL scores of oral and oropharyngeal cancer patients after treatment. Improvement in head and neck- and general cancer-associated domains may not be translated into a general and functional improvement during the first year of recovery.
With an estimated 442,760 new cases and 241,40000 deaths per year worldwide, squamous cell carcinoma of the oral cavity and the oropharynx (OSCC) is among the most common malignant tumors and a significant source of morbidity.1 Although its incidence rate has decreased in most developed countries over the past decades, OSCC remains a common cancer for both males and females in South Central Asia and in Central and Eastern Europe. Management of OSCC is largely surgical, whereas adjuvant treatment, including radiotherapy with or without chemotherapy, is administered for advanced-stage tumors.2 The severe implications of surgery in this area and the toxic effects associated with adjuvant therapy spurred the search for a reliable method to assess the physiologic and psychosocial effect of therapy.3 The number of publications on health-related quality of life (HRQOL) after head and neck cancer has increased over the years, since the term quality of life (QOL) was coined by Heckscher in 1960.3,4 HRQOL is a multidimensional measure that encompasses the physical, emotional, and social components associated with an illness. Patient self-completed questionnaires have become the mainstay of HRQOL evaluation. The questionnaires usually assess 4 main areas: global health measures,
Y. Binenbaum et al.
general cancer measures, specific head and neck cancer measures, and performance measures. Other aspects may also include satisfaction and financial, mental, and spiritual measures.3 The validated questionnaires most commonly used to assess HRQOL in oral cancer are the University of Washington Quality of Life Questionnaire (UW-QOLQ) and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQC30), which is frequently used in conjunction with the Head and Neck-35 (H&N-35) questionnaire.5–7 Studies of oral and oropharyngeal cancers that have incorporated these questionnaires have repeatedly demonstrated decreases in QOL scores immediately after treatment and a return to preoperative levels by approximately 1 year.3,8,9 However, whether these statistically significant changes in QOL measures are also meaningful for the patient is unknown. The minimal clinically important difference (MCID) is defined as ‘‘the smallest difference in score which patients perceive as beneficial and which would, therefore, mandate a change in the patient’s management.’’10 MCID may assist clinicians in interpreting QOL measures.10–13 In this study we performed a metaanalysis to determine the minimal change in QOL scores that has a clinically significance effect on patients with oral cavity and oropharyngeal cancer. Our goal was to reveal the MCID score for the most commonly used head and neck cancer questionnaires, EORTC QLQ-C30/H&N-35 and UW-QOLQ scales. These results may facilitate the utility and implementation of QOL instruments in clinical practice.
METHODS Meta-analysis Search Strategy and Selection Criteria During January 2013, we conducted a systematic electronic literature database search of PubMed, the Cochrane Central Register of Clinical Trials, the Cochrane Database of Systematic Reviews, ClinicalTrials.gov, and Google from 1975 to 2012. The searches were conducted by using the Medical Subject Heading terms (Oral) AND (Cancer) AND (quality of life OR QOL OR HRQOL) or (Oral) AND (UW-QOLQ or EORTC QOL), limited to ‘‘Human’’ and ‘‘English’’ language. Reference lists of retrieved manuscripts were hand-searched for additional publications. Two reviewers (Y. B. and M. A.) independently screened all articles and abstract titles generated by the electronic search strategies. Articles were rejected at the initial screening if their titles or abstracts indicated that they were irrelevant. The full texts of potentially relevant articles were reviewed to assess their suitability for inclusion in this meta-analysis. Inclusion criteria according to study design were randomized controlled trials, prospective and
retrospective cohorts, and case–control study designs. Criteria for study population inclusion were a histopathologic diagnosis of oral or oropharyngeal cancer, evaluation of QOL using the UW-QOLQ or the EORTC QLQ-C30/ QLQ-H&N-35 questionnaire, availability of QOL data specific for oral cancer, and availability of data from which the SD of the different QOL domains could be extracted.
HRQOL Questionnaires The UW-QOLQ, the EORTC QLQ-C30/QLQ-H&N-35, and the M. D. Anderson Dysphagia Inventory are patient based, self-administered questionnaires that have been commonly used and extensively validated in oral cancer patients. The main fields assessed by these questionnaires are global well-being measures, general cancer measures, head and neck-specific measures, and head and neck performance measures.3,14 The UW-QOLQ consists of 3 general items measuring global HRQOL and 12 diseasespecific items: pain, appearance, activity, recreation, swallowing, chewing, speech, shoulder problems, taste, saliva, mood, and anxiety.5 In all but one of the articles covered by this study, version 4, the current version of the questionnaire, was used. The EORTC QLQ-C30 contains five functional scales (physical, role, emotional, cognitive, and social), three symptom scales (fatigue, pain, and nausea), one global measure of health status, and six singleitem scales that assess symptoms (dyspnea, appetite loss, sleep disturbance, constipation, diarrhea, and the perceived financial impact of the disease treatment). The H&N-35 module measures symptoms specific for head and neck cancer patients and includes 7 domains (pain, swallowing, senses, speech, social eating, social contact, and sexuality), as well as 11 single items regarding problems with restricted mouth opening (including sticky saliva, cough, feeling ill, analgesic use, nutritional supplement use, feeding tube, and loss or gain of weight).7,14,15 Unfortunately, we could not find sufficient data in the literature to calculate the MCID for the M. D. Anderson Dysphagia Inventory questionnaire, and therefore it is not covered in this article. For the purpose of this study, we found it useful to group the items from both questionnaires into three basic fields of QOL: general and functional status (encompasses the UW-QOLQ’s appearance, activity, recreation, mood, and anxiety items), the five EORTC QLQ-C30/QLQ-H&N35 functional scales, and the financial difficulties domain. Head and neck-specific measures were defined as the UWQOLQ’s swallowing, chewing, speech, shoulder, taste, and saliva measures and the EORTC QLQ-C30/QLQ-H&N35’s pain, swallowing, senses, speech, social eating, teeth, restricted mouth opening, dry mouth, sticky saliva, and cough. General cancer measures included the EORTC
MCID for Quality of Life in Oral Cavity Cancer
QLQ-C30’s fatigue, nausea, dyspnea, appetite loss, sleep disturbance, constipation, and diarrhea domains. Supplementary Table 1 lists domain grouping. Data Abstraction The two reviewers independently abstracted data and recorded scores of the UW-QOLQ and the EORTC C-30/ H&N-35 for the studies included. Discrepancies between the two reviewers were resolved by discussion and consensus. The reviewers were not blinded to publication details but used identical abstraction forms to record the collected data. Full details regarding all assessed outcome measures were not always available. Validity Assessment Two independent reviewers assessed both the quality and risk of bias of the retrieved publications. They were not blinded to publication details, but measures were taken to obscure outcome assessment, avoid the description of incomplete outcome data, and assess the risk of selective outcome reporting. All included studies were rated according to their study design by use of the suggested levels of evidence proposed by the Oxford Centre for Evidence-based Medicine (Supplemental Table 2).16 Calculating MCID A distribution-based method, rather than an anchorbased method, was used to determine MCID. Distributionbased methods are based on statistical significance, sample variability, and measurement precision. In this meta-analysis, the MCID was calculated as 0.5 SD, as was first proposed by Miller and later developed by Norman and others.17,18 MCID was calculated as follows: the SDs for the different QOL domains were extracted for three points of time relative to treatment: before treatment, 3 months after treatment, and C1 year after treatment. At each point of time, the average SD was calculated per domain, and the MCID was determined to be half that value. Heat Map Analysis It was previously shown that a change in QOL score of 0.5 SD represents the MCID, whereas changes of 1.00–1.50 SD or[1.5 SD are likely to represent a moderate or large change, respectively.18–20 To illustrate clinically significant changes in QOL scores over time, we performed a heat map analysis (Table 4). The table presents the change in QOL scores retrieved at the early (3 months) and late (12 months) postoperative period, relative to the corresponding MCID. The
relative change was calculated as follows: for the early postoperative period, [3-month QOL score - pretreatment QOL score]/MCID at 3 months, and for the late postoperative period, [12-month QOL score - pretreatment QOL score]/MCID at 12 months. On this scale, a ratio ±1 represents a statistically significant minimal clinically important change, a ratio ±2, a significant moderate change, and a ratio ±3, a significant large change. A ratio between -1 and 1 indicates a nonsignificant clinical change. Statistical Analysis Statistical analyses were performed with the SAS software package (SAS Institute, Cary, NC) by a certified statistician. The meta-analysis was performed in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.21 RESULTS Results of the study selection process are depicted in Fig. 1. Of the 4,259 articles reviewed, 2,132 were excluded because of duplications. An additional 2,054 articles were excluded because of irrelevance to the purpose of this meta-analysis. Of the remaining 73 full-text articles, 57 were excluded because of insufficient data or because they did not meet the inclusion criteria (e.g., lack of essential data or no distinction between oral and other head and neck cancers). Finally, 16 studies, comprising 1,011 patients, were included in the meta-analysis (Supplemental Table 2).22–38 Demographic and disease characteristics of the patients are presented in Table 1. The final study group consisted of 1,011 patients between the ages of 23 and 91 years, with a mean age of 57 years. Males represented 67 % of the cohort. The tumor was located in the oral cavity in 82.1 % and was oropharyngeal in 14.14 % of cases. A high T stage (T3–4) presented in 48 % and lymph node metastases in 37.8 %. A total of 99.5 % underwent surgery, and 53.5 % had radiotherapy. Reconstruction was required during surgery for 67 % of patients and included free flaps, regional flaps, and skin grafts. In this cohort, 32.5 % (n = 329) answered the EORTC C-30/H&N-35 questionnaire, 62.4 % (n = 631) answered the UW-QOLQ questionnaire, and 5.1 % answered both questionnaires. MCID for the UW-QOLQ Table 2 shows mean MCID values for the UW-QOLQ domains, retrieved [3 months after treatment. The mean MCID for global QOL was 13.07 points, ranging from 7.3 to 27.1. In the head and neck-associated domains, such as
Y. Binenbaum et al. FIG. 1 Flow chart of the study selection process21
taste, saliva, and chewing, relatively high mean MCID scores were observed, with a large range of values. The mean MCID score for the speech domain was relatively low. The highest range of MCID scores among the studies was in the anxiety domain, with a range of 22.3 points. MCID for the EORTC QLQ-C30/H&N-35 Questionnaire Mean MCID values for the EORTC QLQ-C30/H&N-35 domains retrieved [3 months after treatment are presented in Table 3. The mean MCID for global QOL was 9.43 (range, 5.35–13.75). To validate our results, we compared the MCID scores calculated in our cohort to reference MCID scores extracted from the EORTC group reference data (which are combined for all posttreatment intervals).15 Only minor discrepancies were apparent between the two data sets. Changes in MCID Over Time Supplemental Table 3 shows mean MCID values before treatment, 3 months after treatment, and C1 year after treatment. For the UW-QOLQ, the greatest pretreatment and posttreatment differences were in the speech and anxiety domains, with a 3.2-fold increase in the speech domain and a 1.8-fold increase in the anxiety domain. Similarly, in the EORTC H&N-35 questionnaire, the
speech domain also showed a 2.5-fold increase. Other domains with MCID scores that differed markedly C1 year after treatment relative to pretreatment values were finances, social contact, dry mouth, sticky saliva, and weight gain. Unfortunately, we could not find sufficient data in the literature for the calculation of the MCID values for the UW-QOLQ 3 months after treatment. QOL Domains Which Show a Clinically Significant Change Over the Course of Recovery We next sought to identify domains for which the change in QOL score over the trajectory of recovery reflected an actual clinically significant change as perceived by the patient. The heat map in Table 4 presents the ratio between the within-domain changes in QOL over the course of recovery (3 months and C1 year after treatment relative to pretreatment values) and the corresponding MCID values. From an overview perspective of the heat map of the EORTC questionnaire, 3 months after treatment there was a clinically significant decline in most of the QOL domains. However, 1 year after treatment, the QOL scores tended to return to the preoperative level. When we analyzed by QOL categories, most general cancer measures were perceived to have worsened at 3 months after treatment, but at 1 year ceased to make a clinically important difference on patients’ QOL relative to baseline (p = 0.016 for mean ratio at 3 months vs. [1 year). In the head and
MCID for Quality of Life in Oral Cavity Cancer TABLE 1 Characteristics of patients included in the meta-analysis (N = 1011) Topic
Parameter
Number
Demographics
Age (range)
23–91
Age (mean)
Tumor location
T stage N stage
Treatment
Reconstruction
Questionnaire
(%)
57.51
Male
643
(67.33)
Female
312
(32.67)
Oral cavity
830
(82.10)
Oropharynx
143
(14.14)
NS T1 ? 2
38 459
(3.76) (51.92)
T3 ? 4
425
(48.08)
N0
372
(62.21)
N1
123
(20.57)
N2 ? N3
103
(17.22)
Surgery only
455
(46.33)
S?RT
417
(42.46)
S?Chem?RT
105
(10.69)
RT only
3
(0.31)
Chem only
2
(0.20)
Free flap
174
(17.21)
Skin graft
34
(3.36)
Regional flap
58
(7.37)
NS
398
(39.37)
EORTC UW-QOLQ
329 631
(32.54) (62.41)
Both
51
(5.05)
S surgery, RT radiotherapy, Chem chemotherapy, EORTC European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, UW-QOLQ University of Washington Quality of Life Questionnaire, NS not specified
TABLE 2 UW-QOL questionnaire, [3 month post-treatment mean and range MCID scores Domain
Mean MCID scores
Ranges
Global
13.07
7.30
27.11
Activity
12.33
3.85
16.61
Recreation
11.87
0.50
16.30
Mood Anxiety
13.98 14.60
4.15 4.80
23.24 27.11 12.75
Appearance
9.61
4.10
10.86
3.20
19.35
9.72
0.60
17.50
Swallowing
13.12
0.70
19.35
Chewing
14.20
0.55
23.24
Taste
15.95
4.20
23.24
Saliva
15.00
4.80
26.11
Shoulder
13.97
13.35
14.59
Pain Speech
neck category, a return to baseline is evident in most domains at the 1 year point (p \ 0.001 for mean ratio at 3 months vs. [1 year), and a clinically significant positive change is notable in the swallowing and social eating domain. In contrast to the head and neck-associated domains and the general cancer-associated domains, QOL for most of the functional domains was not perceived to have improved at 1 year after treatment (p = 0.69 for the mean ratio at 3 months vs. [1 year). For the UW-QOLQ, a clinically positive change was perceived by patients in all general and functional domains but activity and recreation domains. Although some head and neck-associated domains, such as taste and saliva, improved markedly at 1 year, speech continued to be impaired. DISCUSSION HRQOL assessment, which was originally a researchoriented tool, is now recognized as an important treatment outcome. Its use has become common practice in clinical trials in the last two decades. Moreover, clinicians are increasingly turning to HRQOL measures for decisionmaking in daily practice, promoting the use of QOL questionnaires to improve patient-doctor interaction and to monitor patient experience with treatment.12,39 Although validated QOL questionnaires have become an integral part of goal assessment in clinical trials, the interpretation of QOL data is not intuitive. For this reason, the clinical significance of changes in QOL scores through the trajectory of recovery after treatment is not conclusive.40 In this metaanalysis, we analyzed the clinical and QOL data of 1,011 patients with OSCC. The male predominance of the study group (67 %) reflects the reported incidence of the disease.1 The half-SD method, used herein for the calculation of MCID, has been shown to reliably represent the MCID in a wide variety of chronic diseases, including cancer.18,41–43 We calculated mean MCID scores for the domains assessed in two widely used QOL questionnaires, UW-QOLQ and EORTC C-30/H&N-35, and provide firsttime benchmark values for oral/oropharyngeal cancer. We found a high consistency between the mean MCID scores reported by the EORTC C-30 and H&N-35 and the UWQOLQ. We aimed to identify QOL domains which patients perceive to be most affected during the course of recovery, by identifying which differential scores were significantly above or below the MCID. The heat map shown in Table 4 demonstrates patient-perceived clinically significant deterioration in most domains at 3 months after treatment, according to the EORTC questionnaires. The most affected domains were those that are directly associated with general cancer treatment, including nausea and vomiting, diarrhea, and constipation. The head and neck–associated
Y. Binenbaum et al. TABLE 3 EORTC QLQ-H&N35/QLQ-C30 Questionnaire, [3 month post-treatment mean and range MCID scores, and reference values Questionnaire
EORTC
EORTC QLQ-C30
Global QOL Physical function
Mean MCID
a
Role function
9.43
5.35
13.75
9.81
5.85
12.55
11.80 Preva
11.60
8.23
14.00
9.60
6.00
16.45
10.55
Emotional function
10.07
5.00
17.35
11.80
Social function
12.83
8.50
17.25
12.10
Financial Fatigue
14.26 11.62
9.50 6.50
18.60 15.75
13.20 13.55
14.50
6.51
3.80
10.50
6.85
Pain
11.68
7.00
16.20
13.60
Dyspnea
12.64
9.80
15.05
11.10
Insomnia
14.43
8.20
18.30
16.65
Appetite loss
12.64
7.15
16.40
14.45
Constipation
11.09
3.55
20.50
10.85
Diarrhea
7.83
0.00
14.45
8.90
Pain
8.67
5.10
11.45
Swallowing
8.04
3.75
10.85
Senses
8.27
6.15
11.85
Speech
9.87
7.85
12.15
10.32
7.60
14.70
Social eating Social contact
a
EORTC reference values
Cognitive function
Emesis
EORTC H&N35
Range
9.22
4.45
13.15
Sexuality Teeth
16.06 12.25
14.48 8.15
17.25 18.05
Opening mouth
13.79
9.66
17.55
Dry mouth
15.73
12.30
17.75
Sticky saliva
15.18
10.85
17.90
Coughing
10.08
4.90
13.90
Feeling ill
10.48
4.88
14.15
Pain killers
17.93
5.95
26.10
Nutritional support
13.04
0.00
25.35
Feeding tube
12.66
4.88
26.10
Weight loss
11.03
3.28
25.35
Weight gain
10.28
0.00
26.10
The physical function domain is not a component of the newest version of the EORTC questionnaire
domains were also perceived worse than at baseline, but, surprisingly, the swallowing, dry mouth, and sticky saliva domains did not seem to deteriorate. The wide-spectrum impairment in general cancer measures and head and neckspecific cancer measures elicit the need for very early intervention by speech therapists, nutritionists, dentists, physical therapists, and social workers. When adjusted by the MCID 1 year after treatment, most general cancer measures and head and neck-specific measures improved, and the scores in most of these domains reflected no clinically important difference relative to baseline. We were pleased to notice that the swallowing domain, presumably
the function that treatment aims to preserve mostly, clinically improved relative to baseline. Unlike the cancer treatment and the head and neck-associated domains, the general and functional domains did not improve at 1 year and continued to be perceived as worse than before treatment (p = 0.69 for change of general and functional domains over the trajectory recovery). This may indicate that the improvement in cancer-associated measures and head and neck-associated measures does not translate into an clinically significant improvement in the general and functional measures. Further study is needed to assess whether these domains can recover over a longer period of
MCID for Quality of Life in Oral Cavity Cancer TABLE 4 Heat map, representing QOL change over time to MCID ratio Questioner
UW-QOLQ
EORTC C-30
EORTC H&N35
Domain Global Pain Appearance Activity Recreation Speech Shoulder Swallowing Chewing Taste Saliva Mood Anxiety Physical function Role function Cognitive function Emotional function Social function Global health status Fatigue Emesis Pain Dyspnea Insomnia Appetite loss Constipation Diarrhea Financial Pain Swallowing Senses Speech Social eating Social contact Sexuality Teeth Opening mouth Dry mouth Sticky saliva Coughing Feeling ill Pain killers Nutritional support Feeding tube Weight loss Weight gain
Early QOL Change / MCID*
-1.63 -1.74 -0.16 -0.36 -1.41 -0.60 -0.37 -7.15 -1.61 -2.18 -0.95 -1.67 -4.61 -8.74 -1.41 -0.65 -0.91 -2.13 -2.10 -2.22 -1.87 -2.34 -1.15 0.15 0.34 -2.91 -0.89 0.34 -0.87 2.26 1.68 0.63
Late QOL Change / MCID 3.19 0.83 7.65 0.07 -0.44 -2.15 -1.69 1.19 -0.98 4.15 4.79 5.32 5.25 -1.66 -1.25 -1.20 -0.11 -1.50 -0.28 0.08 -0.44 0.15 -0.17 0.32 0.25 -0.23 -0.94 -0.35 0.89 1.85 0.84 0.46 1.38 1.97 -0.89 0.36 0.38 0.20 0.73 -0.15 0.97 0.70 -1.85 2.79 2.73 2.91
The table presents ratios between the changes in QOL scores retrieved 3 and 12 months after treatment, relative to the corresponding MCID (3 month QOL score - pretreatment QOL score/MCID at 3 month; 12 month QOL score - pretreatment QOL score/MCID at 12 month). Ratios between -1 and 1, which represent no clinically significant change, are colored grey. Blue spectrum blocks represent a positive clinically significant change (from bright—minimal change, to dark—large clinically significant change). Red spectrum blocks represent negative clinical change (from bright—minimal change, to dark—large clinically significant change). The ratios were adjusted to show positive values for improvements and negative for worsening function *MCID values for 3 month post-treatment could not be calculated for the UW-QOLQ (no sufficient data in the literature for calculation)
Y. Binenbaum et al.
follow-up and to develop comprehensive strategies that may help translate symptomatic improvement into a clinically significant functional one. The heat map for the UW-QOLQ depicts an improvement in both general and functional measures and in the head and neck-associated domains. Notable exceptions are the speech and shoulder domains, which significantly deteriorated clinically. These findings concur with data reported by So et al.,44 who suggested that some treatmentrelated morbidity domains do not recover within 12 months. The heat maps for the two questionnaires show only minor differences in directions of change. However, differences in the magnitude of change as patients deteriorate or improve differ markedly between the questionnaires. Whereas in the UW-QOLQ, the magnitude of QOL change was higher as patients improved (maximum positive change of ?7.65 vs. a negative change of -2.15), in the EORTC questionnaire the magnitude of change was higher as patients deteriorated (-8.74 vs. ?2.91). This may stem from fundamental differences between the two questionnaires: the UW-QOLQ is considered broad based and assesses disease-specific functional status while including a minimal number of global questions. The EORTC C-30 and H&N-35 questionnaire is longer and covers multiple domains and single items. Moreover, the magnitude of change in QOL scores differs if the patient’s condition is worsening or improving.11 For example, Ware and colleges found that the 1-year change in general health perception was ?13 for those who rated themselves as ‘‘much better’’ versus -34 for those who rated themselves as ‘‘much worse.’’45 Juniper et al. reported a similar effect in asthma patients.19 Our heat map probably demonstrates this phenomenon in the EORTC questioner, with a minimum ratio of -8.75 when patients are deteriorating, but a ratio of only ?2.79 as their condition improves. Interestingly, the proportion of the positive to negative change in Ware’s study approximated 3, roughly as in our study. Selection of a questionnaire for a future study should take these inherent differences into account, along with considerations regarding the purpose of the investigation, the time available for responding, and the resources available.14 In conclusion, our findings provide benchmark values for MCID, which may serve as a reference for future studies of QOL of oral cavity and oropharyngeal cancer patients using the UW-QOLQ and the EORTC QLQ-C30/ H&N-35 questionnaires. We suggest the utility of a heatmap analysis to facilitate the understanding of the contribution of each domain to patients’ perceived HRQOL. This analysis of clinically significant QOL changes in the general oral cavity and oropharyngeal cancer population demonstrates a normalization or improvement in general cancer measures and head and neck-associated measures at
1 year after treatment, whereas general and functional status may continue to be impaired. Analysis of HRQOL data has inherent biases that should be considered in the interpretation of data. ACKNOWLEDGMENT This research was supported by the Weizmann Institute—TASMC Joint Grant, the ICRF Barbara S. Goodman endowed research career development award (2011-601BGPC), and a Grant from the US–Israel Binational Science Foundation CONFLICT OF INTEREST
None for any authors.
FINANCIAL DISCLOSURE the authors are to be made.
No other financial disclosures from
REFERENCES 1. GLOBOCAN 2012: Country Fast Stat. http://globocan.iarc.fr/old/ summary_table_site-html.asp?selection=13010&selection=21030 &title=Lip%2C?oral?cavity%2C?Other?pharynx&sex=0&type= 0&window=1&africa=1&america=2&asia=3&europe=4&oceania= 5&build=6&sort=0&submit=%C2%A0Execute. Accessed Apr 2014. 2. Bessell A, Glenny AM, Furness S, et al. Interventions for the treatment of oral and oropharyngeal cancers: surgical treatment (review). Cochrane Database Syst Rev. 2011;9:CD006205. 3. Chandu A, Smith ACH, Rogers SN. Health-related quality of life in oral cancer: a review. J Oral Maxillofac Surg. 2006;64: 495–502. 4. Khafif A, Posen J, Yagil Y, et al. Quality of life in patients older than 75 years following major head and neck surgery. Head Neck. 2007;29:932–9. 5. Lowe D, Rogers SN. University of Washington quality of life questionnaire. Seattle: University of Washington; 2012, pp. 1–14. 6. Bjordal K, Hammerlid E, Ahlner-Elmqvist M, et al. Quality of life in head and neck cancer patients: validation of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-H&N35. J Clin Oncol. 1999;17:1008–19. 7. Torres-Carranza E, Infante-Cossı´o P, Herna´ndez-Guisado JM, Hens-Aumente E, Gutierrez-Pe´rez JL. Assessment of quality of life in oral cancer. Med Oral Patol Oral Cir Bucal 2008;13:E735–41. 8. Ronis DL, Duffy SA, Fowler KE, Khan MJ, Terrell JE. Changes in quality of life over 1 year in patients with head and neck cancer. Arch Otolaryngol Head Neck Surg. 2008;134:241–8. 9. Gritz ER, Carmack CL, de Moor C. et al. First year after head and neck cancer: quality of life. J Clin Oncol. 1999;17:352–60. 10. Jaeschke R, Singer J, Guyatt GH. Measurement of health status: ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10:407–15. 11. Cella D, Hahn EA, Dineen K. Meaningful change in cancerspecific quality of life scores: differences between improvement and worsening. Qual Life Res. 2002;11:207–21. 12. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol. 2003;56:395–407. 13. Speer DC. Clinically significant change: Jacobson and Truax (1991) revisited. J Consult Clin Psychol. 1992;60:402–8. 14. Ringash J, Bezjak A. A structured review of quality of life instruments for head and neck cancer patients. Head Neck. 2001;23:201–13. 15. Scott NW et al. EORTC QLQ-C30 Sprangers on behalf of the EORTC Quality of Life Group. 2008.
MCID for Quality of Life in Oral Cavity Cancer 16. The 2011 Oxford CEBM Levels of Evidence: introductory document. 2011:1–3. 17. Miller GA. The magic number seven, plus or minus two: some limits on our capacity for processing information. Psychol Rev. 1956;63:81–97. 18. Norman GR. Point/counterpoint: interpretation of changes in health-related quality of life—the remarkable universality of half a standard deviation. Med Care. 2003;41:582–92. 19. Juniper EF, Guyatt GH, Willan A, Griffith LE. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. 1994;47:81–7. 20. Sloan J, Symonds T, Vargas-Chanes D, Fridley B. Practical guidelines for assessing the clinical significance of health-related quality of life changes within clinical trials. Drug Inf J. 2003;37: 23–31. 21. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097. doi:10. 1371/journal.pmed1000097. 22. Chandu A, Sun KCV, DeSilva RN, Smith ACH. The assessment of quality of life in patients who have undergone surgery for oral cancer: a preliminary report. J Oral Maxillofac Surg. 2005;63:1606–12. 23. Oates J, Clark JR, Read J, Reeves N, Gao K, O’Brien CJ. Integration of prospective quality of life and nutritional assessment as routine components of multidisciplinary care of patients with head and neck cancer. ANZ J Surg. 2008;78:34–41. 24. Bekiroglu F, Ghazali N, Laycock R, Katre C, Lowe D, Rogers SN. Adjuvant radiotherapy and health-related quality of life of patients at intermediate risk of recurrence following primary surgery for oral squamous cell carcinoma. Oral Oncol. 2011;47:967–73. 25. Rogers SN, Lowe D, Brown JS, Vaughan ED. A comparison between the University of Washington Head and neck diseasespecific measure and the medical short form 36, EORTC QOQ-C33 and EORTC head and neck 35. Oral Oncol. 1998;34: 361–72. 26. Zwahlen RA, Dannemann C, Gra¨tz KW, et al. Quality of life and psychiatric morbidity in patients successfully treated for oral cavity squamous cell cancer and their wives. J Oral Maxillofac Surg. 2008;66:1125–32. 27. Kessler PA, Bloch-Birkholz A, Leher A, Neukam FW, Wiltfang J. Evaluation of quality of life of patients with oral squamous cell carcinoma. Comparison of two treatment protocols in a prospective study. Radiother Oncol. 2004;70:275–82. 28. Airoldi M, Garzaro M, Raimondo L, et al. Functional and psychological evaluation after flap reconstruction plus radiotherapy in oral cancer. Head Neck. 2011;33:458–68. 29. Kazi R, Johnson C, Prasad V, et al. Quality of life outcome measures following partial glossectomy: assessment using the UW-QOL scale. J Cancer Res Ther. 2012;4:116–20. 30. Hsing CY, Wong YK, Wang CP, et al. Comparison between free flap and pectoralis major pedicled flap for reconstruction in oral cavity cancer patients: a quality of life analysis. Oral Oncol. 2011;47:522–7.
31. Girod DA, Sykes K, Jorgensen J, Tawfik O, Tsue T. Acellular dermis compared to skin grafts in oral cavity reconstruction. Laryngoscope. 2009;119:2141–9. 32. Feng FY, Kim HM, Lyden TH, et al. Intensity-modulated chemoradiotherapy aiming to reduce dysphagia in patients with oropharyngeal cancer: clinical and functional results. J Clin Oncol. 2010;28:2732–8. 33. Li W, Yang Y, Xu Z, et al. Assessment of quality of life of patients with oral cavity cancer who have had defects reconstructed with free anterolateral thigh perforator flaps. Br J Oral Maxillofac Surg. 2013;51:10–4. 34. Iwase M, Takemi T, Manabe M, Nagumo M. Hypercalcemic complication in patients with oral squamous cell carcinoma. Int J Oral Maxillofac Surg. 2003;32:174–80. 35. Klug C, Neuburg J, Glaser C, Schwarz B, Kermer C, Millesi W. Quality of life 2–10 years after combined treatment for advanced oral and oropharyngeal cancer. Int J Oral Maxillofac Surg. 2002; 31:664–9. 36. Rogers SN, Miller RD, Ali K, Minhas AB, Williams HF, Lowe D. Patients’ perceived health status following primary surgery for oral and oropharyngeal cancer. Int J Oral Maxillofac Surg. 2006; 35:913–9. 37. Borggreven PA, Verdonck-de Leeuw IM, Muller MJ, et al. Quality of life and functional status in patients with cancer of the oral cavity and oropharynx: pretreatment values of a prospective study. Eur Arch Otorhinolaryngol. 2007;264:651–7. 38. Devine JC, Rogers SN, McNally D, Brown JS, Vaughan ED. A comparison of aesthetic, functional and patient subjective outcomes following lip-split mandibulotomy and mandibular lingual releasing access procedures. Int J Oral Maxillofac Surg. 2001;30: 199–204. 39. Schu¨nemann HJ, Akl EA, Guyatt GH. Interpreting the results of patient reported outcome measures in clinical trials: the clinician’s perspective. Health Qual Life Outcomes. 2006;4:62. 40. Kanatas AN, Mehanna HM, Lowe D, Rogers SN. A second national survey of health-related quality of life questionnaires in head and neck oncology. Ann R Coll Surg Engl. 2009;91:420–5. 41. Bonnetain F, Dahan L, Maillard E, et al. Time until definitive quality of life score deterioration as a means of longitudinal analysis for treatment trials in patients with metastatic pancreatic adenocarcinoma. Eur J Cancer. 2010;46:2753–62. 42. Maringwa J, Quinten C, King M, et al. Minimal clinically meaningful differences for the EORTC QLQ-C30 and EORTC QLQ-BN20 scales in brain cancer patients. Ann Oncol. 2011;22: 2107–12. 43. Amit M, Abergel A, Fliss DM, Gil Z. The clinical importance of quality-of-life scores in patients with skull base tumors: a meta-analysis and review of the literature. Curr Oncol Rep. 2012;14:175–81. 44. So WK, Chan RJ, Chan DN, et al. Quality-of-life among head and neck cancer survivors at one year after treatment: a systematic review. Eur J Cancer. 2012;48:2391–408. 45. Ware JE, Snow K, K. M. SF-36 Health Survey: manual and interpretation guide. 1993.
ORIGINAL ARTICLE – HEAD AND NECK ONCOLOGY
Minimal Clinically Important Differences in Quality of Life Scores of Oral Cavity and Oropharynx Cancer Patients Yoav Binenbaum, MD1, Moran Amit, MD1,3, Salem Billan, MD2, Jacob T. Cohen, MD3, and Ziv Gil, MD, PhD1,3 Department of Otolaryngology Head and Neck Surgery, The Technion, Israel Institute of Technology, Haifa, Israel; 2The Radiology Institute, The Technion, Israel Institute of Technology, Haifa, Israel; 3The Clinical Research Institute at Rambam, Rambam Medical Center and the Rappaport Faculty of Medicine, The Technion Institute of Technology, Haifa, Israel 1
ABSTRACT Background. The minimal clinically important difference (MCID) is defined as the smallest difference in quality of life (QOL) that patients perceive as beneficial and that mandates a change in management. We aimed to determine the MCID among patients with oral cavity and oropharyngeal cancer and to identify domains that are significantly affected during treatment. Methods. The cohort consisted of 1,011 patients analyzed by a metaanalysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement. MCID values for the University of Washington Quality of Life Questionnaire (UW-QOLQ) and European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC C-30) and Head and Neck-35 questionnaires were calculated by using the distribution-based method. Results. The mean MCID for Global QOL was 13.07 points for the UW-QOLQ and 9.43 in the EORTC C-30 questionnaire. High consistency in the MCID values was found between the two questionnaires examined. Heat map analysis indicated a clinically significant improvement in head and neck-associated domains and in domains associated with general cancer treatment 1 year or more after treatment relative to 3 months after treatment (p \ 0.001
Electronic supplementary material The online version of this article (doi:10.1245/s10434-014-3656-z) contains supplementary material, which is available to authorized users. Ó Society of Surgical Oncology 2014 First Received: 31 October 2013 Z. Gil, MD, PhD e-mail: [email protected]
and p = 0.016, respectively). In contrast, improvement in general and functional domains was not evident 1 year or more after treatment (p = 0.69). Conclusions. This study suggests benchmark values for MCID and variation in QOL scores of oral and oropharyngeal cancer patients after treatment. Improvement in head and neck- and general cancer-associated domains may not be translated into a general and functional improvement during the first year of recovery.
With an estimated 442,760 new cases and 241,40000 deaths per year worldwide, squamous cell carcinoma of the oral cavity and the oropharynx (OSCC) is among the most common malignant tumors and a significant source of morbidity.1 Although its incidence rate has decreased in most developed countries over the past decades, OSCC remains a common cancer for both males and females in South Central Asia and in Central and Eastern Europe. Management of OSCC is largely surgical, whereas adjuvant treatment, including radiotherapy with or without chemotherapy, is administered for advanced-stage tumors.2 The severe implications of surgery in this area and the toxic effects associated with adjuvant therapy spurred the search for a reliable method to assess the physiologic and psychosocial effect of therapy.3 The number of publications on health-related quality of life (HRQOL) after head and neck cancer has increased over the years, since the term quality of life (QOL) was coined by Heckscher in 1960.3,4 HRQOL is a multidimensional measure that encompasses the physical, emotional, and social components associated with an illness. Patient self-completed questionnaires have become the mainstay of HRQOL evaluation. The questionnaires usually assess 4 main areas: global health measures,
Y. Binenbaum et al.
general cancer measures, specific head and neck cancer measures, and performance measures. Other aspects may also include satisfaction and financial, mental, and spiritual measures.3 The validated questionnaires most commonly used to assess HRQOL in oral cancer are the University of Washington Quality of Life Questionnaire (UW-QOLQ) and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQC30), which is frequently used in conjunction with the Head and Neck-35 (H&N-35) questionnaire.5–7 Studies of oral and oropharyngeal cancers that have incorporated these questionnaires have repeatedly demonstrated decreases in QOL scores immediately after treatment and a return to preoperative levels by approximately 1 year.3,8,9 However, whether these statistically significant changes in QOL measures are also meaningful for the patient is unknown. The minimal clinically important difference (MCID) is defined as ‘‘the smallest difference in score which patients perceive as beneficial and which would, therefore, mandate a change in the patient’s management.’’10 MCID may assist clinicians in interpreting QOL measures.10–13 In this study we performed a metaanalysis to determine the minimal change in QOL scores that has a clinically significance effect on patients with oral cavity and oropharyngeal cancer. Our goal was to reveal the MCID score for the most commonly used head and neck cancer questionnaires, EORTC QLQ-C30/H&N-35 and UW-QOLQ scales. These results may facilitate the utility and implementation of QOL instruments in clinical practice.
METHODS Meta-analysis Search Strategy and Selection Criteria During January 2013, we conducted a systematic electronic literature database search of PubMed, the Cochrane Central Register of Clinical Trials, the Cochrane Database of Systematic Reviews, ClinicalTrials.gov, and Google from 1975 to 2012. The searches were conducted by using the Medical Subject Heading terms (Oral) AND (Cancer) AND (quality of life OR QOL OR HRQOL) or (Oral) AND (UW-QOLQ or EORTC QOL), limited to ‘‘Human’’ and ‘‘English’’ language. Reference lists of retrieved manuscripts were hand-searched for additional publications. Two reviewers (Y. B. and M. A.) independently screened all articles and abstract titles generated by the electronic search strategies. Articles were rejected at the initial screening if their titles or abstracts indicated that they were irrelevant. The full texts of potentially relevant articles were reviewed to assess their suitability for inclusion in this meta-analysis. Inclusion criteria according to study design were randomized controlled trials, prospective and
retrospective cohorts, and case–control study designs. Criteria for study population inclusion were a histopathologic diagnosis of oral or oropharyngeal cancer, evaluation of QOL using the UW-QOLQ or the EORTC QLQ-C30/ QLQ-H&N-35 questionnaire, availability of QOL data specific for oral cancer, and availability of data from which the SD of the different QOL domains could be extracted.
HRQOL Questionnaires The UW-QOLQ, the EORTC QLQ-C30/QLQ-H&N-35, and the M. D. Anderson Dysphagia Inventory are patient based, self-administered questionnaires that have been commonly used and extensively validated in oral cancer patients. The main fields assessed by these questionnaires are global well-being measures, general cancer measures, head and neck-specific measures, and head and neck performance measures.3,14 The UW-QOLQ consists of 3 general items measuring global HRQOL and 12 diseasespecific items: pain, appearance, activity, recreation, swallowing, chewing, speech, shoulder problems, taste, saliva, mood, and anxiety.5 In all but one of the articles covered by this study, version 4, the current version of the questionnaire, was used. The EORTC QLQ-C30 contains five functional scales (physical, role, emotional, cognitive, and social), three symptom scales (fatigue, pain, and nausea), one global measure of health status, and six singleitem scales that assess symptoms (dyspnea, appetite loss, sleep disturbance, constipation, diarrhea, and the perceived financial impact of the disease treatment). The H&N-35 module measures symptoms specific for head and neck cancer patients and includes 7 domains (pain, swallowing, senses, speech, social eating, social contact, and sexuality), as well as 11 single items regarding problems with restricted mouth opening (including sticky saliva, cough, feeling ill, analgesic use, nutritional supplement use, feeding tube, and loss or gain of weight).7,14,15 Unfortunately, we could not find sufficient data in the literature to calculate the MCID for the M. D. Anderson Dysphagia Inventory questionnaire, and therefore it is not covered in this article. For the purpose of this study, we found it useful to group the items from both questionnaires into three basic fields of QOL: general and functional status (encompasses the UW-QOLQ’s appearance, activity, recreation, mood, and anxiety items), the five EORTC QLQ-C30/QLQ-H&N35 functional scales, and the financial difficulties domain. Head and neck-specific measures were defined as the UWQOLQ’s swallowing, chewing, speech, shoulder, taste, and saliva measures and the EORTC QLQ-C30/QLQ-H&N35’s pain, swallowing, senses, speech, social eating, teeth, restricted mouth opening, dry mouth, sticky saliva, and cough. General cancer measures included the EORTC
MCID for Quality of Life in Oral Cavity Cancer
QLQ-C30’s fatigue, nausea, dyspnea, appetite loss, sleep disturbance, constipation, and diarrhea domains. Supplementary Table 1 lists domain grouping. Data Abstraction The two reviewers independently abstracted data and recorded scores of the UW-QOLQ and the EORTC C-30/ H&N-35 for the studies included. Discrepancies between the two reviewers were resolved by discussion and consensus. The reviewers were not blinded to publication details but used identical abstraction forms to record the collected data. Full details regarding all assessed outcome measures were not always available. Validity Assessment Two independent reviewers assessed both the quality and risk of bias of the retrieved publications. They were not blinded to publication details, but measures were taken to obscure outcome assessment, avoid the description of incomplete outcome data, and assess the risk of selective outcome reporting. All included studies were rated according to their study design by use of the suggested levels of evidence proposed by the Oxford Centre for Evidence-based Medicine (Supplemental Table 2).16 Calculating MCID A distribution-based method, rather than an anchorbased method, was used to determine MCID. Distributionbased methods are based on statistical significance, sample variability, and measurement precision. In this meta-analysis, the MCID was calculated as 0.5 SD, as was first proposed by Miller and later developed by Norman and others.17,18 MCID was calculated as follows: the SDs for the different QOL domains were extracted for three points of time relative to treatment: before treatment, 3 months after treatment, and C1 year after treatment. At each point of time, the average SD was calculated per domain, and the MCID was determined to be half that value. Heat Map Analysis It was previously shown that a change in QOL score of 0.5 SD represents the MCID, whereas changes of 1.00–1.50 SD or[1.5 SD are likely to represent a moderate or large change, respectively.18–20 To illustrate clinically significant changes in QOL scores over time, we performed a heat map analysis (Table 4). The table presents the change in QOL scores retrieved at the early (3 months) and late (12 months) postoperative period, relative to the corresponding MCID. The
relative change was calculated as follows: for the early postoperative period, [3-month QOL score - pretreatment QOL score]/MCID at 3 months, and for the late postoperative period, [12-month QOL score - pretreatment QOL score]/MCID at 12 months. On this scale, a ratio ±1 represents a statistically significant minimal clinically important change, a ratio ±2, a significant moderate change, and a ratio ±3, a significant large change. A ratio between -1 and 1 indicates a nonsignificant clinical change. Statistical Analysis Statistical analyses were performed with the SAS software package (SAS Institute, Cary, NC) by a certified statistician. The meta-analysis was performed in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.21 RESULTS Results of the study selection process are depicted in Fig. 1. Of the 4,259 articles reviewed, 2,132 were excluded because of duplications. An additional 2,054 articles were excluded because of irrelevance to the purpose of this meta-analysis. Of the remaining 73 full-text articles, 57 were excluded because of insufficient data or because they did not meet the inclusion criteria (e.g., lack of essential data or no distinction between oral and other head and neck cancers). Finally, 16 studies, comprising 1,011 patients, were included in the meta-analysis (Supplemental Table 2).22–38 Demographic and disease characteristics of the patients are presented in Table 1. The final study group consisted of 1,011 patients between the ages of 23 and 91 years, with a mean age of 57 years. Males represented 67 % of the cohort. The tumor was located in the oral cavity in 82.1 % and was oropharyngeal in 14.14 % of cases. A high T stage (T3–4) presented in 48 % and lymph node metastases in 37.8 %. A total of 99.5 % underwent surgery, and 53.5 % had radiotherapy. Reconstruction was required during surgery for 67 % of patients and included free flaps, regional flaps, and skin grafts. In this cohort, 32.5 % (n = 329) answered the EORTC C-30/H&N-35 questionnaire, 62.4 % (n = 631) answered the UW-QOLQ questionnaire, and 5.1 % answered both questionnaires. MCID for the UW-QOLQ Table 2 shows mean MCID values for the UW-QOLQ domains, retrieved [3 months after treatment. The mean MCID for global QOL was 13.07 points, ranging from 7.3 to 27.1. In the head and neck-associated domains, such as
Y. Binenbaum et al. FIG. 1 Flow chart of the study selection process21
taste, saliva, and chewing, relatively high mean MCID scores were observed, with a large range of values. The mean MCID score for the speech domain was relatively low. The highest range of MCID scores among the studies was in the anxiety domain, with a range of 22.3 points. MCID for the EORTC QLQ-C30/H&N-35 Questionnaire Mean MCID values for the EORTC QLQ-C30/H&N-35 domains retrieved [3 months after treatment are presented in Table 3. The mean MCID for global QOL was 9.43 (range, 5.35–13.75). To validate our results, we compared the MCID scores calculated in our cohort to reference MCID scores extracted from the EORTC group reference data (which are combined for all posttreatment intervals).15 Only minor discrepancies were apparent between the two data sets. Changes in MCID Over Time Supplemental Table 3 shows mean MCID values before treatment, 3 months after treatment, and C1 year after treatment. For the UW-QOLQ, the greatest pretreatment and posttreatment differences were in the speech and anxiety domains, with a 3.2-fold increase in the speech domain and a 1.8-fold increase in the anxiety domain. Similarly, in the EORTC H&N-35 questionnaire, the
speech domain also showed a 2.5-fold increase. Other domains with MCID scores that differed markedly C1 year after treatment relative to pretreatment values were finances, social contact, dry mouth, sticky saliva, and weight gain. Unfortunately, we could not find sufficient data in the literature for the calculation of the MCID values for the UW-QOLQ 3 months after treatment. QOL Domains Which Show a Clinically Significant Change Over the Course of Recovery We next sought to identify domains for which the change in QOL score over the trajectory of recovery reflected an actual clinically significant change as perceived by the patient. The heat map in Table 4 presents the ratio between the within-domain changes in QOL over the course of recovery (3 months and C1 year after treatment relative to pretreatment values) and the corresponding MCID values. From an overview perspective of the heat map of the EORTC questionnaire, 3 months after treatment there was a clinically significant decline in most of the QOL domains. However, 1 year after treatment, the QOL scores tended to return to the preoperative level. When we analyzed by QOL categories, most general cancer measures were perceived to have worsened at 3 months after treatment, but at 1 year ceased to make a clinically important difference on patients’ QOL relative to baseline (p = 0.016 for mean ratio at 3 months vs. [1 year). In the head and
MCID for Quality of Life in Oral Cavity Cancer TABLE 1 Characteristics of patients included in the meta-analysis (N = 1011) Topic
Parameter
Number
Demographics
Age (range)
23–91
Age (mean)
Tumor location
T stage N stage
Treatment
Reconstruction
Questionnaire
(%)
57.51
Male
643
(67.33)
Female
312
(32.67)
Oral cavity
830
(82.10)
Oropharynx
143
(14.14)
NS T1 ? 2
38 459
(3.76) (51.92)
T3 ? 4
425
(48.08)
N0
372
(62.21)
N1
123
(20.57)
N2 ? N3
103
(17.22)
Surgery only
455
(46.33)
S?RT
417
(42.46)
S?Chem?RT
105
(10.69)
RT only
3
(0.31)
Chem only
2
(0.20)
Free flap
174
(17.21)
Skin graft
34
(3.36)
Regional flap
58
(7.37)
NS
398
(39.37)
EORTC UW-QOLQ
329 631
(32.54) (62.41)
Both
51
(5.05)
S surgery, RT radiotherapy, Chem chemotherapy, EORTC European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, UW-QOLQ University of Washington Quality of Life Questionnaire, NS not specified
TABLE 2 UW-QOL questionnaire, [3 month post-treatment mean and range MCID scores Domain
Mean MCID scores
Ranges
Global
13.07
7.30
27.11
Activity
12.33
3.85
16.61
Recreation
11.87
0.50
16.30
Mood Anxiety
13.98 14.60
4.15 4.80
23.24 27.11 12.75
Appearance
9.61
4.10
10.86
3.20
19.35
9.72
0.60
17.50
Swallowing
13.12
0.70
19.35
Chewing
14.20
0.55
23.24
Taste
15.95
4.20
23.24
Saliva
15.00
4.80
26.11
Shoulder
13.97
13.35
14.59
Pain Speech
neck category, a return to baseline is evident in most domains at the 1 year point (p \ 0.001 for mean ratio at 3 months vs. [1 year), and a clinically significant positive change is notable in the swallowing and social eating domain. In contrast to the head and neck-associated domains and the general cancer-associated domains, QOL for most of the functional domains was not perceived to have improved at 1 year after treatment (p = 0.69 for the mean ratio at 3 months vs. [1 year). For the UW-QOLQ, a clinically positive change was perceived by patients in all general and functional domains but activity and recreation domains. Although some head and neck-associated domains, such as taste and saliva, improved markedly at 1 year, speech continued to be impaired. DISCUSSION HRQOL assessment, which was originally a researchoriented tool, is now recognized as an important treatment outcome. Its use has become common practice in clinical trials in the last two decades. Moreover, clinicians are increasingly turning to HRQOL measures for decisionmaking in daily practice, promoting the use of QOL questionnaires to improve patient-doctor interaction and to monitor patient experience with treatment.12,39 Although validated QOL questionnaires have become an integral part of goal assessment in clinical trials, the interpretation of QOL data is not intuitive. For this reason, the clinical significance of changes in QOL scores through the trajectory of recovery after treatment is not conclusive.40 In this metaanalysis, we analyzed the clinical and QOL data of 1,011 patients with OSCC. The male predominance of the study group (67 %) reflects the reported incidence of the disease.1 The half-SD method, used herein for the calculation of MCID, has been shown to reliably represent the MCID in a wide variety of chronic diseases, including cancer.18,41–43 We calculated mean MCID scores for the domains assessed in two widely used QOL questionnaires, UW-QOLQ and EORTC C-30/H&N-35, and provide firsttime benchmark values for oral/oropharyngeal cancer. We found a high consistency between the mean MCID scores reported by the EORTC C-30 and H&N-35 and the UWQOLQ. We aimed to identify QOL domains which patients perceive to be most affected during the course of recovery, by identifying which differential scores were significantly above or below the MCID. The heat map shown in Table 4 demonstrates patient-perceived clinically significant deterioration in most domains at 3 months after treatment, according to the EORTC questionnaires. The most affected domains were those that are directly associated with general cancer treatment, including nausea and vomiting, diarrhea, and constipation. The head and neck–associated
Y. Binenbaum et al. TABLE 3 EORTC QLQ-H&N35/QLQ-C30 Questionnaire, [3 month post-treatment mean and range MCID scores, and reference values Questionnaire
EORTC
EORTC QLQ-C30
Global QOL Physical function
Mean MCID
a
Role function
9.43
5.35
13.75
9.81
5.85
12.55
11.80 Preva
11.60
8.23
14.00
9.60
6.00
16.45
10.55
Emotional function
10.07
5.00
17.35
11.80
Social function
12.83
8.50
17.25
12.10
Financial Fatigue
14.26 11.62
9.50 6.50
18.60 15.75
13.20 13.55
14.50
6.51
3.80
10.50
6.85
Pain
11.68
7.00
16.20
13.60
Dyspnea
12.64
9.80
15.05
11.10
Insomnia
14.43
8.20
18.30
16.65
Appetite loss
12.64
7.15
16.40
14.45
Constipation
11.09
3.55
20.50
10.85
Diarrhea
7.83
0.00
14.45
8.90
Pain
8.67
5.10
11.45
Swallowing
8.04
3.75
10.85
Senses
8.27
6.15
11.85
Speech
9.87
7.85
12.15
10.32
7.60
14.70
Social eating Social contact
a
EORTC reference values
Cognitive function
Emesis
EORTC H&N35
Range
9.22
4.45
13.15
Sexuality Teeth
16.06 12.25
14.48 8.15
17.25 18.05
Opening mouth
13.79
9.66
17.55
Dry mouth
15.73
12.30
17.75
Sticky saliva
15.18
10.85
17.90
Coughing
10.08
4.90
13.90
Feeling ill
10.48
4.88
14.15
Pain killers
17.93
5.95
26.10
Nutritional support
13.04
0.00
25.35
Feeding tube
12.66
4.88
26.10
Weight loss
11.03
3.28
25.35
Weight gain
10.28
0.00
26.10
The physical function domain is not a component of the newest version of the EORTC questionnaire
domains were also perceived worse than at baseline, but, surprisingly, the swallowing, dry mouth, and sticky saliva domains did not seem to deteriorate. The wide-spectrum impairment in general cancer measures and head and neckspecific cancer measures elicit the need for very early intervention by speech therapists, nutritionists, dentists, physical therapists, and social workers. When adjusted by the MCID 1 year after treatment, most general cancer measures and head and neck-specific measures improved, and the scores in most of these domains reflected no clinically important difference relative to baseline. We were pleased to notice that the swallowing domain, presumably
the function that treatment aims to preserve mostly, clinically improved relative to baseline. Unlike the cancer treatment and the head and neck-associated domains, the general and functional domains did not improve at 1 year and continued to be perceived as worse than before treatment (p = 0.69 for change of general and functional domains over the trajectory recovery). This may indicate that the improvement in cancer-associated measures and head and neck-associated measures does not translate into an clinically significant improvement in the general and functional measures. Further study is needed to assess whether these domains can recover over a longer period of
MCID for Quality of Life in Oral Cavity Cancer TABLE 4 Heat map, representing QOL change over time to MCID ratio Questioner
UW-QOLQ
EORTC C-30
EORTC H&N35
Domain Global Pain Appearance Activity Recreation Speech Shoulder Swallowing Chewing Taste Saliva Mood Anxiety Physical function Role function Cognitive function Emotional function Social function Global health status Fatigue Emesis Pain Dyspnea Insomnia Appetite loss Constipation Diarrhea Financial Pain Swallowing Senses Speech Social eating Social contact Sexuality Teeth Opening mouth Dry mouth Sticky saliva Coughing Feeling ill Pain killers Nutritional support Feeding tube Weight loss Weight gain
Early QOL Change / MCID*
-1.63 -1.74 -0.16 -0.36 -1.41 -0.60 -0.37 -7.15 -1.61 -2.18 -0.95 -1.67 -4.61 -8.74 -1.41 -0.65 -0.91 -2.13 -2.10 -2.22 -1.87 -2.34 -1.15 0.15 0.34 -2.91 -0.89 0.34 -0.87 2.26 1.68 0.63
Late QOL Change / MCID 3.19 0.83 7.65 0.07 -0.44 -2.15 -1.69 1.19 -0.98 4.15 4.79 5.32 5.25 -1.66 -1.25 -1.20 -0.11 -1.50 -0.28 0.08 -0.44 0.15 -0.17 0.32 0.25 -0.23 -0.94 -0.35 0.89 1.85 0.84 0.46 1.38 1.97 -0.89 0.36 0.38 0.20 0.73 -0.15 0.97 0.70 -1.85 2.79 2.73 2.91
The table presents ratios between the changes in QOL scores retrieved 3 and 12 months after treatment, relative to the corresponding MCID (3 month QOL score - pretreatment QOL score/MCID at 3 month; 12 month QOL score - pretreatment QOL score/MCID at 12 month). Ratios between -1 and 1, which represent no clinically significant change, are colored grey. Blue spectrum blocks represent a positive clinically significant change (from bright—minimal change, to dark—large clinically significant change). Red spectrum blocks represent negative clinical change (from bright—minimal change, to dark—large clinically significant change). The ratios were adjusted to show positive values for improvements and negative for worsening function *MCID values for 3 month post-treatment could not be calculated for the UW-QOLQ (no sufficient data in the literature for calculation)
Y. Binenbaum et al.
follow-up and to develop comprehensive strategies that may help translate symptomatic improvement into a clinically significant functional one. The heat map for the UW-QOLQ depicts an improvement in both general and functional measures and in the head and neck-associated domains. Notable exceptions are the speech and shoulder domains, which significantly deteriorated clinically. These findings concur with data reported by So et al.,44 who suggested that some treatmentrelated morbidity domains do not recover within 12 months. The heat maps for the two questionnaires show only minor differences in directions of change. However, differences in the magnitude of change as patients deteriorate or improve differ markedly between the questionnaires. Whereas in the UW-QOLQ, the magnitude of QOL change was higher as patients improved (maximum positive change of ?7.65 vs. a negative change of -2.15), in the EORTC questionnaire the magnitude of change was higher as patients deteriorated (-8.74 vs. ?2.91). This may stem from fundamental differences between the two questionnaires: the UW-QOLQ is considered broad based and assesses disease-specific functional status while including a minimal number of global questions. The EORTC C-30 and H&N-35 questionnaire is longer and covers multiple domains and single items. Moreover, the magnitude of change in QOL scores differs if the patient’s condition is worsening or improving.11 For example, Ware and colleges found that the 1-year change in general health perception was ?13 for those who rated themselves as ‘‘much better’’ versus -34 for those who rated themselves as ‘‘much worse.’’45 Juniper et al. reported a similar effect in asthma patients.19 Our heat map probably demonstrates this phenomenon in the EORTC questioner, with a minimum ratio of -8.75 when patients are deteriorating, but a ratio of only ?2.79 as their condition improves. Interestingly, the proportion of the positive to negative change in Ware’s study approximated 3, roughly as in our study. Selection of a questionnaire for a future study should take these inherent differences into account, along with considerations regarding the purpose of the investigation, the time available for responding, and the resources available.14 In conclusion, our findings provide benchmark values for MCID, which may serve as a reference for future studies of QOL of oral cavity and oropharyngeal cancer patients using the UW-QOLQ and the EORTC QLQ-C30/ H&N-35 questionnaires. We suggest the utility of a heatmap analysis to facilitate the understanding of the contribution of each domain to patients’ perceived HRQOL. This analysis of clinically significant QOL changes in the general oral cavity and oropharyngeal cancer population demonstrates a normalization or improvement in general cancer measures and head and neck-associated measures at
1 year after treatment, whereas general and functional status may continue to be impaired. Analysis of HRQOL data has inherent biases that should be considered in the interpretation of data. ACKNOWLEDGMENT This research was supported by the Weizmann Institute—TASMC Joint Grant, the ICRF Barbara S. Goodman endowed research career development award (2011-601BGPC), and a Grant from the US–Israel Binational Science Foundation CONFLICT OF INTEREST
None for any authors.
FINANCIAL DISCLOSURE the authors are to be made.
No other financial disclosures from
REFERENCES 1. GLOBOCAN 2012: Country Fast Stat. http://globocan.iarc.fr/old/ summary_table_site-html.asp?selection=13010&selection=21030 &title=Lip%2C?oral?cavity%2C?Other?pharynx&sex=0&type= 0&window=1&africa=1&america=2&asia=3&europe=4&oceania= 5&build=6&sort=0&submit=%C2%A0Execute. Accessed Apr 2014. 2. Bessell A, Glenny AM, Furness S, et al. Interventions for the treatment of oral and oropharyngeal cancers: surgical treatment (review). Cochrane Database Syst Rev. 2011;9:CD006205. 3. Chandu A, Smith ACH, Rogers SN. Health-related quality of life in oral cancer: a review. J Oral Maxillofac Surg. 2006;64: 495–502. 4. Khafif A, Posen J, Yagil Y, et al. Quality of life in patients older than 75 years following major head and neck surgery. Head Neck. 2007;29:932–9. 5. Lowe D, Rogers SN. University of Washington quality of life questionnaire. Seattle: University of Washington; 2012, pp. 1–14. 6. Bjordal K, Hammerlid E, Ahlner-Elmqvist M, et al. Quality of life in head and neck cancer patients: validation of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-H&N35. J Clin Oncol. 1999;17:1008–19. 7. Torres-Carranza E, Infante-Cossı´o P, Herna´ndez-Guisado JM, Hens-Aumente E, Gutierrez-Pe´rez JL. Assessment of quality of life in oral cancer. Med Oral Patol Oral Cir Bucal 2008;13:E735–41. 8. Ronis DL, Duffy SA, Fowler KE, Khan MJ, Terrell JE. Changes in quality of life over 1 year in patients with head and neck cancer. Arch Otolaryngol Head Neck Surg. 2008;134:241–8. 9. Gritz ER, Carmack CL, de Moor C. et al. First year after head and neck cancer: quality of life. J Clin Oncol. 1999;17:352–60. 10. Jaeschke R, Singer J, Guyatt GH. Measurement of health status: ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10:407–15. 11. Cella D, Hahn EA, Dineen K. Meaningful change in cancerspecific quality of life scores: differences between improvement and worsening. Qual Life Res. 2002;11:207–21. 12. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol. 2003;56:395–407. 13. Speer DC. Clinically significant change: Jacobson and Truax (1991) revisited. J Consult Clin Psychol. 1992;60:402–8. 14. Ringash J, Bezjak A. A structured review of quality of life instruments for head and neck cancer patients. Head Neck. 2001;23:201–13. 15. Scott NW et al. EORTC QLQ-C30 Sprangers on behalf of the EORTC Quality of Life Group. 2008.
MCID for Quality of Life in Oral Cavity Cancer 16. The 2011 Oxford CEBM Levels of Evidence: introductory document. 2011:1–3. 17. Miller GA. The magic number seven, plus or minus two: some limits on our capacity for processing information. Psychol Rev. 1956;63:81–97. 18. Norman GR. Point/counterpoint: interpretation of changes in health-related quality of life—the remarkable universality of half a standard deviation. Med Care. 2003;41:582–92. 19. Juniper EF, Guyatt GH, Willan A, Griffith LE. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. 1994;47:81–7. 20. Sloan J, Symonds T, Vargas-Chanes D, Fridley B. Practical guidelines for assessing the clinical significance of health-related quality of life changes within clinical trials. Drug Inf J. 2003;37: 23–31. 21. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097. doi:10. 1371/journal.pmed1000097. 22. Chandu A, Sun KCV, DeSilva RN, Smith ACH. The assessment of quality of life in patients who have undergone surgery for oral cancer: a preliminary report. J Oral Maxillofac Surg. 2005;63:1606–12. 23. Oates J, Clark JR, Read J, Reeves N, Gao K, O’Brien CJ. Integration of prospective quality of life and nutritional assessment as routine components of multidisciplinary care of patients with head and neck cancer. ANZ J Surg. 2008;78:34–41. 24. Bekiroglu F, Ghazali N, Laycock R, Katre C, Lowe D, Rogers SN. Adjuvant radiotherapy and health-related quality of life of patients at intermediate risk of recurrence following primary surgery for oral squamous cell carcinoma. Oral Oncol. 2011;47:967–73. 25. Rogers SN, Lowe D, Brown JS, Vaughan ED. A comparison between the University of Washington Head and neck diseasespecific measure and the medical short form 36, EORTC QOQ-C33 and EORTC head and neck 35. Oral Oncol. 1998;34: 361–72. 26. Zwahlen RA, Dannemann C, Gra¨tz KW, et al. Quality of life and psychiatric morbidity in patients successfully treated for oral cavity squamous cell cancer and their wives. J Oral Maxillofac Surg. 2008;66:1125–32. 27. Kessler PA, Bloch-Birkholz A, Leher A, Neukam FW, Wiltfang J. Evaluation of quality of life of patients with oral squamous cell carcinoma. Comparison of two treatment protocols in a prospective study. Radiother Oncol. 2004;70:275–82. 28. Airoldi M, Garzaro M, Raimondo L, et al. Functional and psychological evaluation after flap reconstruction plus radiotherapy in oral cancer. Head Neck. 2011;33:458–68. 29. Kazi R, Johnson C, Prasad V, et al. Quality of life outcome measures following partial glossectomy: assessment using the UW-QOL scale. J Cancer Res Ther. 2012;4:116–20. 30. Hsing CY, Wong YK, Wang CP, et al. Comparison between free flap and pectoralis major pedicled flap for reconstruction in oral cavity cancer patients: a quality of life analysis. Oral Oncol. 2011;47:522–7.
31. Girod DA, Sykes K, Jorgensen J, Tawfik O, Tsue T. Acellular dermis compared to skin grafts in oral cavity reconstruction. Laryngoscope. 2009;119:2141–9. 32. Feng FY, Kim HM, Lyden TH, et al. Intensity-modulated chemoradiotherapy aiming to reduce dysphagia in patients with oropharyngeal cancer: clinical and functional results. J Clin Oncol. 2010;28:2732–8. 33. Li W, Yang Y, Xu Z, et al. Assessment of quality of life of patients with oral cavity cancer who have had defects reconstructed with free anterolateral thigh perforator flaps. Br J Oral Maxillofac Surg. 2013;51:10–4. 34. Iwase M, Takemi T, Manabe M, Nagumo M. Hypercalcemic complication in patients with oral squamous cell carcinoma. Int J Oral Maxillofac Surg. 2003;32:174–80. 35. Klug C, Neuburg J, Glaser C, Schwarz B, Kermer C, Millesi W. Quality of life 2–10 years after combined treatment for advanced oral and oropharyngeal cancer. Int J Oral Maxillofac Surg. 2002; 31:664–9. 36. Rogers SN, Miller RD, Ali K, Minhas AB, Williams HF, Lowe D. Patients’ perceived health status following primary surgery for oral and oropharyngeal cancer. Int J Oral Maxillofac Surg. 2006; 35:913–9. 37. Borggreven PA, Verdonck-de Leeuw IM, Muller MJ, et al. Quality of life and functional status in patients with cancer of the oral cavity and oropharynx: pretreatment values of a prospective study. Eur Arch Otorhinolaryngol. 2007;264:651–7. 38. Devine JC, Rogers SN, McNally D, Brown JS, Vaughan ED. A comparison of aesthetic, functional and patient subjective outcomes following lip-split mandibulotomy and mandibular lingual releasing access procedures. Int J Oral Maxillofac Surg. 2001;30: 199–204. 39. Schu¨nemann HJ, Akl EA, Guyatt GH. Interpreting the results of patient reported outcome measures in clinical trials: the clinician’s perspective. Health Qual Life Outcomes. 2006;4:62. 40. Kanatas AN, Mehanna HM, Lowe D, Rogers SN. A second national survey of health-related quality of life questionnaires in head and neck oncology. Ann R Coll Surg Engl. 2009;91:420–5. 41. Bonnetain F, Dahan L, Maillard E, et al. Time until definitive quality of life score deterioration as a means of longitudinal analysis for treatment trials in patients with metastatic pancreatic adenocarcinoma. Eur J Cancer. 2010;46:2753–62. 42. Maringwa J, Quinten C, King M, et al. Minimal clinically meaningful differences for the EORTC QLQ-C30 and EORTC QLQ-BN20 scales in brain cancer patients. Ann Oncol. 2011;22: 2107–12. 43. Amit M, Abergel A, Fliss DM, Gil Z. The clinical importance of quality-of-life scores in patients with skull base tumors: a meta-analysis and review of the literature. Curr Oncol Rep. 2012;14:175–81. 44. So WK, Chan RJ, Chan DN, et al. Quality-of-life among head and neck cancer survivors at one year after treatment: a systematic review. Eur J Cancer. 2012;48:2391–408. 45. Ware JE, Snow K, K. M. SF-36 Health Survey: manual and interpretation guide. 1993.
