ORIGINAL ARTICLE

Quality of Life in Patients

With Head and Neck Cancer
Lessons Learned From 549 Prospectively Evaluated Patients
Ernest A. Weymuller, Jr, MD; Bevan Yueh, MD, MPH; Frederic W. B. Deleyiannis, MD, MPhil, MPH;
Alice L. Kuntz, MD; Ramsey Alsarraf, MD, MPH; Marc D. Coltrera, MD

Objectives: To summarize our quality-of-life (QOL) re- obtained for 260 of these patients, of whom 210 pre-
search findings for patients with head and neck cancer, sented with an untreated first primary tumor (index cases)
to suggest areas for future productive QOL research, and to the University of Washington, Seattle.
to discuss how to undertake QOL studies in a cost-
effective manner. Intervention: Pretreatment QOL was assessed with an
interviewer-supervised self-administered question-
Design: Review of previously published analyses of ad- naire. Subsequent self-administered tests were com-
vanced larynx cancer, advanced oropharynx cancer, and pleted at 3, 6, 12, 24, and 36 months. Other data col-
neck-dissection cases and current data from the com- lected on each patient included cancer site, stage,
plete set of patients. treatment, histologic findings, type of surgical recon-
struction, and current disease and vital status.
Patients: From January 1, 1993, through December 31,
1998, data on 549 patients were entered in our head and Results/Conclusions: It is difficult to achieve “statis-
neck database. Of these patients, 364 met additional cri- tically significant” results in a single-instution setting. The
teria for histologic findings (squamous cell carcinoma) “composite” QOL score may not be a sufficiently sensi-
and the restriction of their cancer to 4 major anatomical tive tool. Analysis of separate domains may be more ef-
sites (oral, oropharynx, hypopharynx, or larynx). Of these, fective.
339 patients were more than 1 year beyond initial treat-
ment. Complete baseline TNM staging and QOL data were Arch Otolaryngol Head Neck Surg. 2000;126:329-335

O
UR QUALITY-OF-LIFE suggestions about specific areas that may
(QOL) project was initi- be fruitful for QOL research and con-
ated in the late 1980s be- sider how researchers may undertake
cause we realized that QOL studies in a cost-efficient manner.
while therapy was evolv-
ing in many ways, multi-institution trials
had failed to demonstrate meaningful sur- RESULTS
vival differences among various forms of
treatment. This led to a consideration of From January 1, 1993, through Decem-
measuring QOL as a way to compare treat- ber 30, 1998, data obtained on 549
ment modalities with similar cure rates. patients were entered in our head and
To undertake a QOL analysis of neck database. From this group, 364 met
From the Department of head and neck cancer, a disease-specific additional criteria for histologic findings
Otolaryngology–Head and QOL tool, the University of Washington (squamous cell carcinoma) and restric-
Neck Surgery, University of Quality-of-Life (UW-QOL) question- tion of their cancer to 4 major sites (oral,
Washington School of Medicine naire, was created and validated.1 Since oropharynx, hypopharynx, or larynx).
(Drs Weymuller, Yueh, 1993, every new cancer patient present- Of these 364 patients, 339 were more
Deleyiannis, Kuntz, Alsarraf,
ing to the University of Washington than 1 year beyond initial treatment. We
and Coltrera), and the Health
Services Research and Medical Center, Seattle, has been asked obtained complete baseline TNM and
Development Service and to participate in a prospective study. This QOL data on 260 of these patients, 50 of
Surgery Service, Veterans article summarizes our experience and whom had recurrent or persistent disease
Affairs Puget Sound Health discusses what we have learned from at the time of presentation and were
Care System (Dr Yueh), Seattle. studies of data within this set. We make eliminated from this analysis. Two hun-

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
329

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 PATIENTS, MATERIALS, Composite QOL Score
AND METHODS The composite QOL score refers to the mean or numerical
change, with scores ranging from 0 (worst QOL) to 100
(best QOL).
The project began January 1, 1993, after approval by the
Human Subjects Committee of the University of Washing- Global QOL Score
ton. All new patients presenting to the University of Wash-
ington Medical Center with a diagnosis of head and neck The global QOL score is a direct overall assessment of QOL.
cancer were asked to participate in a prospective analysis We asked patients about both health-related QOL and QOL
of QOL changes during and after treatment. Patients com- related to all issues. For example, beginning in March 1995,
pleted an interviewer-supervised self-administered pre- all patients were asked the following: “Considering every-
treatment QOL questionnaire on the day of their initial thing in your life that contributes to your personal well-
workup. Subsequent self-administered questionnaires were being, rate your overall quality of life during the past 7 days.”
completed at 3, 6, 12, 24, and 36 months. Other data ob- Patient responses were scored as either outstanding, very
tained for each patient included site of treatment, cancer good, good, fair, poor, or very poor.
stage, type of treatment, histologic findings, surgical re-
Incremental QOL Change Score
construction, and current health status. A QOL registrar
was responsible for patient follow-up and data collection The incremental QOL change score is an assessment by the
and collation. All data were entered in the departmental patient of how his or her QOL has changed over a given pe-
relational database that was developed by one of the au- riod. For example, we asked, compared with 1 year prior to
thors (M.D.C.). the diagnosis of a patient’s illness, if the patient would rate
his or her health in general as much worse, somewhat worse,
DEFINITION OF TERMS AND DESCRIPTION about the same, somewhat better, or much better.
OF THE UW-QOL QUESTIONNAIRE
Importance Score
Domain Score
The importance score (after 1995) was determined by ask-
ing each patient to indicate how important he or she con-
The domain score reflects the QOL associated with differ- sidered each domain. For example, patients were asked to
ent aspects of head and neck cancer treatment. Twelve do- rank the importance of pain in their overall QOL as not
mains were probed: (1) activity, (2) appearance, (3) chew- important, a little bit important, somewhat important, quite
ing, (4) dry mouth, (5) employment, (6) pain, (7) recreation, important, or extremely important.
(8) saliva, (9) shoulder function, (10) speech, (11) swal-
lowing, and (12) taste. Responses ranging from 0 (worst) SUMMARY OF PREVIOUSLY PUBLISHED RESULTS
to 100 (best) were obtained with Likert scales. For ex-
ample, the options for pain were scored as follows: The results of analyses of 3 separate subsets of patients have
• I have no pain (score, 100). been previously published.2-4 These studies were undertaken
• There is mild pain not requiring medication (score, 75). during the evolution of the project and therefore included
• I have moderate pain, requires regular medication (co- a smaller group of patients.
deine or nonnarcotic) (score, 50). The QOL of Disease-Free Survivors of Advanced (Stage
• I have severe pain controlled only by narcotics (score, III or IV) Oropharyngeal Cancer
25).
• I have severe pain not controlled by narcotics (score, Our first study, analyzing 13 consecutive patients with ad-
0). vanced oropharyngeal cancer who had been treated with

dred ten patients remained who presented with an statistically significant fashion (P,.001 by analysis of vari-
untreated first primary tumor. ance) (Figure 1). A similar distribution was noted when
No. of Patients
the index cases were stratified by T1 through T4
Entered in study 549 (Figure 2). Site of tumor (oral cavity, oropharynx, hy-
Squamous cell carcinoma at 4 major sites 364 popharynx, or larynx) also affected composite QOL scores
.1 year beyond initial treatment 339 (Figure 3). Although the composite QOL scores were
Completed baseline QOL data 260 similar at baseline, the QOL curves had significantly dif-
Primary tumors at time of presentation 210 ferent patterns by 3 months (P,.05 by analysis of vari-
The distribution of these 210 patients by cancer stage ance), with laryngeal cases faring best and patients with
demonstrates that our institutional experience is domi- oropharyngeal and hypopharyngeal tumors faring worst.
nated by advanced-stage disease (stage III and IV, 76%). We were interested to determine whether differ-
Cancer Stage No. of Patients (%) ences in QOL emerged between treatment arms (sur-
I 17 (8) gery vs organ preservation) with this larger data set.
II 33 (16) Among advanced primary tumors (TNM stage III or IV)
III 52 (25) of the oropharynx and larynx, the composite scores did
IV 108 (51) not appear to demonstrate clinically or statistically sig-
An analysis of these 210 index cases revealed that nificant differences (Figure 4 and Figure 5). Simi-
TNM stage (I-IV) influenced composite QOL scores in a larly, we did not detect differences when global QOL or

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
330

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 either surgery and radiation or chemotherapy and radia- function at 20 years posttreatment. The majority of patients
tion, was completed in 1995, and the results were pub- reported a good to excellent overall QOL, and most patients
lished in 1997.2 All 13 patients were disease free 1 year post- reported that their general health was the same as or better
treatment. The composite (total) QOL pretreatment and than it was 1 year prior to the diagnosis of cancer. We also
posttreatment scores were similar for the 2 groups. A sub- found that although loss of speech was disabling and a lar-
set analysis of the QOL domains demonstrated a worsen- yngectomy disfiguring, only a minority of patients reported
ing of chewing and swallowing in both groups after treat- speech or appearance as being more than “somewhat impor-
ment. The surgical group showed a more frequent worsening tant” to their overall QOL. They indicated that “activity” (ie,
of speech and appearance. Sixty-seven percent of the pa- socialization) was the most important issue.
tients who received surgical treatment (n = 6) reported pain This adaptation effect has been noted by others.5 Given
relief vs 29% of the nonsurgical group (n = 7). this natural tendency among survivors, it may be difficult
The results of this study led us to observe that the to- to document a difference among treatments on the basis
tal score of the UW-QOL questionnaire is derived from a of QOL. A corollary to this conclusion is that, if a differ-
variety of “domains” and that the functional changes cre- ence in total QOL score is demonstrated, it is probably real
ated by various therapies demonstrate their effects in the and probably significant. The addition of a global QOL ques-
domain scores. For example, patients who present with pain tion provided a valuable contrast to the disease-specific do-
appear to have better relief by surgery than by radiation, mains of the UW-QOL questionnaire.
and this domain score improves with surgery. In contrast,
surgical treatment causes a decrease in the domain score The Impact of Neck Dissection on QOL
for appearance, while both radiation and surgery create a
transient worsening in chewing and swallowing. These vari- In a study by Kuntz and Weymuller,4 149 consecutive pa-
ous changes cause a “cancellation effect” (ie, while one score tients who underwent neck dissection between June 1993
is rising, another is falling; therefore, the composite score and March 1997 were evaluated. After patients with inad-
does not change much). Because of this, the composite QOL equate follow-up or compliance were excluded, 84 pa-
score may be less sensitive to therapeutically induced change tients were included. We analyzed the QOL changes with
than individual domain scores. respect to radical neck dissection, modified radical neck
dissection (accessory nerve preserved), and selective neck
Are Functional Disabilities Important in the dissection (levels II, III, and IV removed). Specific atten-
Determination of QOL After Laryngectomy? tion was directed at the domain that assesses shoulder func-
tion. Changes in this domain reflected the anticipated im-
In another study, 25 patients with carcinoma of the larynx pact of the various forms of neck dissection. Shoulder
and hypopharynx occurring from 1993 through 1995 were domain scores for the radical neck group were lower than
identified.3 All patients were treated with surgery (total lar- those of the selective neck dissection group at 6 and 12
yngectomy) and radiation therapy; some patients were also months (P = .004). In addition, there was a trend toward
treated with induction or concomitant chemotherapy. (The decreased pain after treatment in the selective neck dissec-
study also assessed 10 patients who were alive without dis- tion and modified radical neck dissection groups.
ease at 2-year follow-up.) We found that the postlaryngectomy The results of this study demonstrated that the shoul-
total QOL score was not significantly different from the prelar- der domain question of the UW-QOL questionnaire was
yngectomy QOL score and that 2 years after laryngectomy, able to discriminate among selective neck dissection, modi-
the domain scores reflected a variety of functional disabili- fied radical neck dissection, and radical neck dissection.
ties. However, when patients were asked to judge the impor- Similar analysis of other domains (speech, swallowing, etc)
tance of these disabilities, there was no significant correlation. will be much more complex, since the surgical interven-
In all QOL domains, including speech and appearance, 50% tions, such as composite resection, are not as well defined
or more of the patients reported having the same or better as the variations of neck dissection.

health-related QOL was used as an outcome (data not with an organ-sparing approach. Unfortunately,
shown). because of a small sample size (most laryngectomy
However, incremental QOL change scores suggest cases at this institution present with recurrent disease
that a difference was present. When patients with ad- and this analysis is limited to patients with previously
vanced oropharyngeal cancer were asked how their health untreated disease), statistical significance could not be
had changed since prior to tumor identification, those reached, although the clinical difference was readily
who were treated with surgery perceived that their health apparent.
was better, but those who were treated with chemora-
diation perceived that their health was worse (Figure 6). COMMENT
The clinical difference is small, but we believe it is sig-
nificant. The difference was found to be statistically sig- QOL INFORMATION IN OUR DATABASE
nificant at 1 year (P,.05, t test), but diminishing sample
size prevented the demonstration of statistical signifi- Figure 2 provides the composite QOL score values
cance beyond 1 year. over time for this cohort stratified by TNM stage. As
Patients with laryngeal cancer showed the opposite hypothesized, QOL was best for patients with early
effect ( Figure 7 ). It appeared that patients with disease and worst for those with advanced disease.
advanced laryngeal cancer who were treated with sur- Patients with advanced tumors also demonstrated the
gery perceived less improvement than those treated most precipitous decline in QOL, presumably associ-

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
331

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 TNM Stage I TNM Stage II TNM Stage III TNM Stage IV Larynx Oral Cavity Oropharynx Hypopharynx

100 100

90 90
Composite UW-QOL Scores

Composite UW-QOL Scores

80 80

70 70

60 60

50 50

40 40

30 30
Treatment 6 12 24 36 Treatment 6 12 24 36
Time After Treatment, mo Time After Treatment, mo

Figure 1. Pretreatment and posttreatment composite University of Figure 3. Pretreatment and posttreatment composite University of
Washington Quality-of-Life (UW-QOL) questionnaire scores by presenting Washington Quality-of-Life (UW-QOL) questionnaire QOL scores by
TNM stage for the inception cohort (n = 210). Numbers of patients who anatomical site of tumor for the inception cohort (n = 210). Numbers of
completed the QOL questionnaire pretreatment and 2 years posttreatment patients who completed the QOL questionnaire pretreatment and 2 years
were 17 and 8, respectively, for TNM stage I, 33 and 9 for TNM stage II, posttreatment were 34 and 12, respectively, for patients with larynx cancer,
52 and 11 for TNM stage III, and 108 and 22 for TNM stage IV. 85 and 18 for patients with oral cavity cancer, 77 and 16 for patients with
oropharynx cancer, and 14 and 4 for patients with hypopharynx cancer.

T1 T2 T3 T4 The importance scores, however, have not improved the

100 precision of our data.
In preliminary psychometric analyses, the concur-
90
rent validity of the UW-QOL questionnaire composite
Composite UW-QOL Scores

80 scores (compared with the global and incremental items)

is worse when importance ratings are used in the calcu-
70
lations. We suspect this may be in part a result of the way
60 importance ratings were obtained; our findings indicate
50
that there appeared to be a tendency of patients to rate
every item as “important,” which prevented us from iden-
40 tifying distinctions between the domains in terms of their
30 relative importance. A proposed modification for obtain-
Treatment 6 12 24 36 ing this information will be presented separately.
Time After Treatment, mo

Figure 2. Pretreatment and posttreatment composite University of INCREMENTAL QOL

Washington Quality-of-Life (UW-QOL) questionnaire scores by presenting
T stage for the inception cohort (n = 210). Numbers of patients who
completed the QOL questionnaire pretreatment and 2 years posttreatment Using the composite QOL score, we were unable to dis-
were 29 and 12, respectively, for T1, 63 and 15 for T2, 65 and 16 for T3, cern a difference between surgically and nonsurgically
and 53 and 7 for T4. treated patients at either anatomical site where these treat-
ment options exist (larynx or oropharynx). However, the
technique of measuring incremental change did iden-
ated with more aggressive therapy, and they did not tify a statistically significant separation between treat-
return to their baseline QOL. These patterns are con- ments.6
sistent with the clinical course we expect for various Since QOL is such a subjective construct, we feel
stages and treatments and are an indication of the that assessing incremental QOL change is more repre-
validity and responsiveness of the UW-QOL question- sentative of what we intend to measure when we
naire to clinical change. examine QOL issues in head and neck cancer therapy.
A large group of our patients is not addressed by these Quality of life is more (or less) than the sum of the
data. The decrease from 210 index patients at entry to domains of the UW-QOL questionnaire. As indicated
50 at 2 years posttreatment is the result of deaths (with by Calman:
or without disease), loss of contact, and exclusion of any The quality of life can only be described and measured in in-
patient with less than complete QOL data at any time. It dividual terms, and depends on present lifestyle, past experi-
is therefore important to note that we are reporting QOL ence, hopes for the future, dreams and ambitions. Quality of
scores, not survival. life must include all areas of life and experience and take into
account the impact of illness and treatment.7
IMPORTANCE SCORES
OPEN-ENDED TEXT
Since 1995, patients have been asked to rate the impor-
tance of each domain in the UW-QOL questionnaire on The availability of open-ended text provides yet another
a scale from “not important” to “extremely important.” avenue of information regarding individual patients.

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
332

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 Surgery Chemoradiation Surgery Chemoradiation

100 Advanced Primary Oropharyngeal Tumors Advanced Primary Oropharyngeal Tumors

Incremental Change in UW-QOL Scores

90 Much Better
Composite UW-QOL Scores

80 Better

70 Same

60 Worse

50 Much Worse

Treatment 6 12 24 36
40
Time After Treatment, mo
30
Treatment 6 12 24 36 Figure 6. Pretreatment and posttreatment incremental change in composite
Time After Treatment, mo University of Washington Quality-of-Life (UW-QOL) questionnaire scores for
advanced primary oropharyngeal tumors by treatment (n = 44). Numbers of
Figure 4. Pretreatment and posttreatment composite University of patients who completed the QOL questionnaire pretreatment and 2 years
Washington Quality-of-Life (UW-QOL) questionnaire scores for advanced posttreatment were 20 and 6, respectively, for patients who underwent
primary oropharyngeal tumors by treatment (n = 72). Numbers of patients surgical treatment and 24 and 9 for patients who received chemoradiation.
who completed the QOL questionnaire pretreatment and 2 years
posttreatment were 37 and 6, respectively, for patients who underwent
surgical treatment and 35 and 9 for patients who received chemoradiation.
Surgery Chemoradiation

Advanced Primary Laryngeal Tumors

Surgery Chemoradiation Incremental Change in UW-QOL Scores Much Better

100 Advanced Primary Laryngeal Tumors
Better

90 Same
Composite UW-QOL Scores

80 Worse

70 Much Worse

60 Treatment 6 12 24 36
Time After Treatment, mo
50
Figure 7. Pretreatment and posttreatment incremental change in University
40
of Washington Quality-of-Life (UW-QOL) questionnaire scores for advanced
30
primary laryngeal tumors by treatment (n = 12). Numbers of patients who
Treatment 6 12 24 36 completed the QOL questionnaire pretreatment and 2 years posttreatment
Time After Treatment, mo were 8 and 4, respectively, for patients who underwent surgical treatment
and 4 and 2 for patients who received chemoradiation.
Figure 5. Pretreatment and posttreatment composite University of
Washington Quality-of-Life (UW-QOL) questionnaire scores for advanced
primary laryngeal tumors by treatment (n = 22). Numbers of patients who Cost, $*
completed the QOL questionnaire pretreatment and 2 years posttreatment Senior faculty member (major programming effort 3 3 y) ...
were 15 and 4, respectively, for patients who underwent surgical treatment Data manager ($40 000/y 3 5 y) 200 000
and 7 and 2 for patients who received chemoradiation.
Computer programmer ($40 000/y 3 1 y) 40 000
Computer equipment 15 000
*Ellipses indicate no cost value available.
While collecting data with the UW-QOL questionnaire, Even when thorough data collection is accomplished,
we received a variety of written comments, some allow- other realities can compromise the usefulness of the data
ing us to recognize that there are health care problems in light of the unavoidable loss of patients as a result of
in need of attention that would have otherwise escaped cancer recurrence, death, and lack of sustained patient
our attention. We also received a variety of humorous participation in long-term studies. For this reason, we
and tragic remarks, cartoons, photographs, etc. This have identified some actions that may reduce the costs
material is more difficult to analyze statistically, but it of comparable studies that are conducted by single in-
can provide meaningful interaction with respect to an stitutions.
individual patient.
Eliminate Exhaustive Data Collection
RESOURCE CONSUMPTION It is our recommendation that prospective longitudinal
studies be limited to analysis of the most common types
Prospective collection of comprehensive QOL data is an and sites of tumors at a particular institution.
expensive and time-consuming enterprise, requiring a
full-time registrar to maintain data integrity. A conserva- Obtain Long-term Funding
tive estimate of our cost that does not begin to address
the sustained commitment by attending surgeons The separation of QOL outcome in our laryngeal cases did
appears below. not become apparent until the 2-year interval. When lon-

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
333

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 gitudinal studies are planned, they will need long-term fund- the composite QOL score (sum of domain scores) is a
ing to achieve meaningful results. If sustained funding is relatively insensitive measure because patients tend to
not available, such studies should not be undertaken. adjust to their disabilities. Moreover, it is not likely
that the composite QOL score will identify a difference
Direct Data Entry between 2 forms of treatment. It is, however, true that
if a difference is identified, it is probably quite mean-
From 1993 through 1998, our QOL data were col- ingful.
lected on paper forms that were forwarded to the data The addition of global and incremental change
manager for computer entry. As this project has scores brings balance to the analysis of the QOL of
evolved, it is now possible for physicians to directly patients. A single global question is an appropriate
enter specific staging and treatment data. With this addition to a disease-specific questionnaire and allows
system in place, a part-time tumor registrar could be investigators to differentiate between disease-specific
employed to focus exclusively on patient follow-up issues and the patient’s overall life satisfaction, which
and QOL information. will be influenced by many psychological and social fac-
tors that may be totally unrelated to the disease-specific
PROSPECTIVE DATA COLLECTION DOES NOT interventions under analysis. Additionally, incremental
GUARANTEE MEANINGFUL INFORMATION change variables may be more sensitive to clinical
change. In our data, incremental variables suggested
Even in prospective studies with intense commitment that differences between treatment arms existed,
from faculty and staff, data may be incomplete because whereas the other QOL measures were unable to detect
patients (1) fail to keep appointments, (2) do not return these differences.
questionnaires, and (3) fail to complete forms accu- It is evident to us that domain scores should be
rately. Furthermore, those patients whose QOL is poor- analyzed separately and that treatment-specific effects
est are least likely to respond to questionnaires or re- may become apparent through contrasting specific
turn for follow-up. Clearly, these issues will inject bias responses within the domains, as was noted in our
into the data. analysis of neck dissection with respect to shoulder
function.
STATISTICAL POWER AND SINGLE-
INSTITUTION STUDIES AREAS FOR SINGLE-INSTITUTION STUDIES

Even though our data set contains over 500 prospec- It is our impression that single-institution studies need
tively evaluated patients, analyses of patient subsets did sharp focus and are probably best done as (1) cross-
not achieve statistical significance. There are 3 unavoid- sectional studies of survivors, (2) short-term studies in
able issues in single-institution studies that must be rec- which QOL is the dominant or perhaps only issue (eg,
ognized. First, randomization does not occur, and at the palliative care analysis, alternative pain management strat-
outset there is a selection bias for treatment decisions egies, and phase I and phase II analyses of the toxic ef-
that is unavoidable. Second, when stratifying patients for fects of different chemoradiation or gene therapy regi-
cancer site, stage, and treatment, small populations do mens), or (3) study cohorts of patients that constitute
not allow for statistical comparisons of QOL, especially your largest institutional experience.
in advanced-stage tumors, because 40% to 50% of
patients do not survive for 2-year comparisons. Third, if CONCLUSIONS
QOL is to be used as an end point for the comparison of
2 treatments, a multi-institution trial will generally be 1. Achieving statistically significant results in QOL stud-
necessary to achieve numeric sufficiency. For single- ies of patients with head and neck cancer is challenging,
institution projects, it is recommended that cross- especially in a single-institution setting while stratify-
sectional studies be considered. With the cross-sectional ing patients for cancer site, stage, and treatment and ac-
technique, a subset of surviving patients can be identi- counting for predictable attrition rates.
fied and studied using a single interaction based on a 2. When QOL is a secondary end point (as in multi-
questionnaire. Of course, this format has its own limita- institution studies comparing survival and local-
tions (ie, it does not address the QOL of those who fail regional control in radiation vs surgery), it will be diffi-
to survive treatment). cult to demonstrate statistically significant differences in
patient-oriented QOL, since survivors tend to accept and
SELECTION AND USE OF QOL INSTRUMENTS adjust to their disabilities. A corollary is that if a QOL
difference is demonstrated, it is probably real and sig-
In the design of new QOL studies, it is important to nificant.
construct a research question first and then analyze 3. The composite QOL score (sum of the domain
the various QOL instruments that are available. No scores) is subject to an “internal cancellation” effect. It
single QOL instrument currently exists that is appro- is therefore less sensitive to overall change when com-
priate for all studies. Each QOL instrument should be paring treatment options.
carefully considered to ensure that the various 4. Separate analysis of QOL domains provides a more
domains address issues that are pertinent to the par- accurate picture of the complex functional changes that
ticular study being developed. We have learned that occur during cancer therapy.

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
334

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 5. Global QOL and incremental change assess- Administration, Department of Veterans Affairs, Washing-
ments provide important information and balance to ton, DC (Dr Yuch).
QOL investigations. Compared with the composite Presented at the annual meeting of the American Head
QOL score, we feel that these measures more effectively and Neck Society, Palm Desert, Calif, April 25, 1999.
portray this subjective construct. Assessing incremental Corresponding author: Ernest A. Weymuller, Jr, MD,
change is more representative of what we intend to University of Washington, Department of Otolaryngology–
measure when we examine QOL issues in head and Head and Neck Surgery, Box 356515, Seattle, WA 98195
neck cancer therapy. (e-mail: [email protected]).
6. There are specific areas in which single-
institution QOL research for head and neck cancer therapy REFERENCES
is likely to be fruitful, especially in short-term studies in
which QOL is the dominant end point, such as the ex-
1. Hassan SJ, Weymuller EA Jr. Assessment of quality of life in head and neck can-
amination of changes associated with surgery or chemo-
cer patients. Head Neck. 1993;15:485-496.
radiation during the first 6 months of treatment, pain man- 2. Deleyiannis FW, Weymuller EA Jr, Coltrera MD. Quality of life of disease-free sur-
agement, and palliative care. vivors of advanced (stage III or IV) oropharyngeal cancer. Head Neck. 1997;19:
7. Comprehensive, prospective QOL data collec- 466-473.
tion is expensive. Cost reduction in QOL studies may be 3. Deleyiannis FW, Weymuller EA Jr, Coltrera MD, Futran NF. Quality of life after lar-
yngectomy: are functional disabilities important? Head Neck. 1999;21:319-324.
accomplished through the use of cross-sectional studies 4. Kuntz AL, Weymuller EA Jr. Impact of neck dissection on quality of life. Laryn-
and by addressing incremental change. Cost reduction goscope. 1999;109:1334-1338.
may also be achieved by more aggressive utilization of 5. Long SA, D’Antonio LL, Robinson EB, Zimmerman G, Petti G, Chonkich G. Fac-
computerized data collection. tors related to quality of life and functional status in 50 patients with head and
neck cancer. Laryngoscope. 1996;106:1084-1088.
6. Fischer D, Stewart AL, Bloch DA, Lorig K, Laurent D, Holman H. Capturing the pa-
Accepted for publication December 23, 1999. tient’s view of change as a clinical outcome measure. JAMA. 1999;282:1157-1162.
Supported in part by grant CD-98318-A from the Health 7. Calman KC. Quality of life in cancer patients: an hypothesis. J Med Ethics. 1984;
Services Research and Development Service, Veterans Health 10:124-127.

Editorial Footnote

I n this issue, Weymuller and coworkers1 describe their experience with the prospective evaluation of quality of life
(QOL) in a large group of patients with head and neck cancer. The authors focus on the value of the information gained
in attempting to measure QOL in all of their patients at the time of presentation, as well as at 5 intervals following
treatment. This information was prospectively gathered in a relational database, which also included demographic, site, stage,
treatment, histological, reconstructive, and current disease status data. Most of their patients (77%) presented with stages III
and IV disease. Ninety-eight (40%) of 245 patients had a complete data set at 2 years. The chance of achieving a complete
data set for an individual patient was strongly influenced by stage at the time of presentation, with 20 (95%) of 21 patients
with stage I cancer fully characterized at 2 years, while just 41 (32%)of 129 patients with stage IV cancer are fully charac-
terized at 2 years. The data reported for the fully characterized group of 98 patients includes mean QOL score over time by
TNM stage (I-IV), T stage,2-4 and site. The authors also give 3 examples of QOL outcomes studies done by mining this rich
database to evaluate QOL results achieved with specific treatment options.
The University of Washington (Seattle, Wa) Head and Neck QOL Questionnaire is a well-designed, validated, disease-
specific QOL assessment instrument. The investigators involved in this work have played a leadership role in promoting
high-quality clinical research techniques in head and neck ontology. The creation of this large relational database required
a major commitment by the leadership of the Department of Otolaryngology; the physicians, nurses, and registrars who worked
with the patients; and the patients themselves who completed the assessment instrument on multiple occasions. Has it been
worth it? Probably so, but there are many challenges and there is much yet to learn. The study reviewed here sheds more
light on these challenges and the current status of QOL research analysis than on any specific clinical question of interest to
physicians and patients.
The use of measured patient perceptions of outcome as data points in a clinical research study was initially viewed with
considerable skepticism by the academic medical community. At that time, physicians were trained to prefer more objective
data. Later, the concept was embraced with considerable enthusiasm as part of a broader interest in outcomes research. Pro-
ponents predicted widespread benefits, including better information for patients and physicians, valid guidelines for medi-
cal practice, and better decisions by health care purchasers. Arnold Relman, the highly respected past editor of the New En-
gland Journal of Medicine, called the growing interest in outcomes measurement “the third revolution in medical care.”2 More
recently, some skepticism seems to have resurfaced, related to concerns about the quality of the measurement, analysis, and
interpretation of this data. A thoughtful review of the challenges involved with using QOL data in the highly charged world
of new oncology drug approval was recently presented in The Cancer Letter3; that review focuses on the need to continuously
improve the rigor of QOL outcomes research.
Continued on next page

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
335

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015
 Editorial Footnote (cont)

Investigators who are interested in adding QOL outcomes to their clinical research face a number of challenges. First, there
has been little standardization of the way in which QOL data are analyzed and presented. There is, if you will, no Kaplan-Meier
standard for the presentation of QOL data as there is for survival data. Weymuller and colleagues present their data as mean total
QOL scores for a fully characterized group of patients over time. This gives a clear picture of the group as a whole, but the out-
look for individuals within the group is less clear, and many patients are excluded by the lack of a complete data set. Requiring
a complete 2-year data set for inclusion in the data analysis excludes those patients who die early of their disease, as well as those
who may be less motivated for recommended follow-up, and may introduce bias. In addition, Weymuller et al point out that
analysis of data using total QOL scores obtained with the University of Washington instrument may obscure the efficacy of a
treatment through cancellation, which occurs when the patient improves in 1 or more domains but also experiences an equal
decline in other domains. For example, laryngectomy may relieve pain and improve swallowing, but it may worsen communi-
cation. Analysis by individual domains (eg, pain, communication, eating, shoulder function) minimizes the chance of a cancel-
lation effect, but complicates the presentation of the data. The University of Washington Head and Neck QOL Questionnaire
includes information about 12 separate domains. Requesting a global QOL assessment from the patient and asking the patient to
rate the importance of each domain began later in the study, and both were helpful in avoiding the cancellation effect. Finally,
there are a number of different validated instruments to chose from, and investigators are only now gaining experience with the
strengths and weaknesses of each. As discussed by the authors, no single instrument is appropriate for all studies.
Another challenge experienced by investigators using QOL outcomes is in the actual conduct of a study. Despite a great
deal of commitment and a high-quality team effort, Weymuller and colleagues were able to obtain a complete data set in only
a minority of their patients. In addition, this type of work currently consumes a great deal of time and resources, especially
if one attempts to maintain a complete relational database on all patients within the institution. Automation of data entry
may minimize this issue in time, but for now the expense is prohibitive for many groups. The ability to implement these
studies in a cost-effective manner is an important issue.
As today’s multidisciplinary treatment teams investigate the value of specific therapeutic options, it is important to have
effective methods for measuring the impact of treatment on QOL, in addition to more effectively measuring objective out-
comes, such as the chance of survival and the incidence of complications. Survival remains the most important single issue for
most patients with head and neck cancer, but QOL issues are next in significance.4 The lessons learned from the large experi-
ence of the group at the University of Washington are a welcome addition to the growing body of knowledge in this field.

W. Jarrard Goodwin, MD
Miami, Fla
(e-mail: [email protected])

1. Weymuller EA Jr, Yueh B, Deleyiannis FWB, Kuntz AL, Alsarraf R, Coltrera MD. Quality of life in head and neck cancer: lessons learned from 549 prospec-
tively evaluated patients. Arch Otolaryngol Head Neck Surg. 2000;126:329-335.
2. Relman AS. Assessment and accountability: the third revolution in medical care. N Engl J Med. 1988;319:1220-1222.
3. Moinpour C, Ganz P, Gritz E, et al. The value of quality of life data in judging patient benefit: experts respond to ODAC. Cancer Lett. 1999;25:1-5.
4. List MA, Butler P, Vokes EE, et al. Head and neck cancer patients: how do patients prioritize potential treatment outcomes [abstract]? American Society of
Clinical Oncology. 1998;17:Abstract 0737.

ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, MAR 2000 WWW.ARCHOTO.COM
336

©2000 American Medical Association. All rights reserved.

Downloaded From: http://archotol.jamanetwork.com/ by a Georgian Court University User on 05/12/2015