a japanese version of the questionnaire for quality of …q-life.org/acrt/acrt03p045_053.pdfa...
Post on 10-Mar-2018
231 Views
Preview:
TRANSCRIPT
A Japanese version of the questionnaire
for quality of life measurement
ORIGINAL ARTICLE Annals of Cancer Research and Therapy
Kazue Yamaoka*1), Fumi Hayashi*2), Chikio Hayashi*3), Jifuji Misumi*4),
Shinichi Takezawa*5), Masuyo Maeda*6), Hiroshi Sato*7), Seizo Oe*7), Kyoji Ogoshi*8) and QOL Research Group
A self-administered, short Japanese version of a quality of life (QOL) questionnaire was devised. This questionnaire measured non-disease-specific QOL. It was devised on the basis of the working hypothesis that "attitude towards disease" (F) affects "state of disease" (D)
, and items related to both F and D were considered. The items were divided into two questionnaires, each of which included certain common key questions.
This artice presents the details of our strategy for developing a short Japanese questionnaire to measure non-disease-specific QOL.
Ann Cancer Res Ther 3 (1): 45-53, 1994/Received 14 Mar 1994, Accepted 22 Mar 1994
Key words: quality of life, snort puestionnaire hon-disease-specific, patient structure, Hayashi's quantification method III
Life expectancy has recently increased dramatically
not only in developed countries, but also in Japan.
Increased attention is now being given to the importance
of quality of life (QOL), and methods of QOL measure-
ment have accordingly been developed1).
The increase of chronic disease makes it difficult to
evaluate the effect of treatment over the short term and
necessitates the provision of long-term care. Considering
these changes in the nature of disease, QOL measures have become increasingly important in assessing the
effectiveness of medical care1~4). Recently, not only for
chronic disease patients5) and elderly patients6), but also
for cancer patients, QOL has achieved a more important
role in the medical (especially oncological) commu-
nity7~12). In clinical cancer trials, QOL measurement has
also become important as a key efficient parameter along
with survival 13,14).
Although a large number of studies have been perfor-
med on QOL questionnaires, little attention has been
given developing a non-disease-specific QOL question-naire for Japanese. In the present study, we developed a
short self-administered questionnaire for Japanese
patients to measure non-disease-specific QOL.
Quality of Life Research Groop
Quality of Life Research Groop has a multidisciplinary-membership drawn from the fields of medicine, psy-
chology, sociology, philosophy, nursing, economics, and
statistics. The study group was formed in 1985 and the
members are listed in the Appendix. The principal aim of
this group is to consider health-related QOL and the
members share a common interest in QOL measures. The
current study was conducted in order to develop a simple
QOL questionnaire.
Background, definition, and development strategy
We compiled over 300 questions to assess QOL from
1985 to 1990. At the same time, we tried to translate the
concept of QOL into Japanese and decided on the term "IKIGAI" . The scheme of the QOL questionnaire is
shown in Table 1; items are classified into physiological,
psychological, and environmental factors. In developing a QOL questionnaire, we felt that non-disease-specific
questions should form the core, so that the questionnaire could be modified with disease-specific questions as necessary. In the development process, we referred to
many other QOL studies5,7~9,13~17).
By taking the differences of national character between
Japan and English-speaking countries, we have made a
Japanese version of a non-disease-specific QOL question-
naire, and then have reduced the number of items in the
present study. This QOL questionnaire can also be used to measure QOL in healthy persons.
In this study, we adopted the following two working
hypotheses:
*1) School of Jurisprudence, Teikyo University *2) School of Humanity, Toyo Eiwa Womens' University
*3) Institute of Statistical Mathematics *4) Chikushi-Jyogakuen University *5) Rikkyo University
*6) Nursing Department, Tokai University *7) the deceased *8) Second Department of Surgery, School of Medicine, Tokai Univer-
sity Hiroshi Sato, Seizo Oe; deceased
Correspondence to: Kazue Yamaoka, Scool of Jurisprudence, Teikyo University, 359 Otuka, Hachioji, Tokyo 192-03, Japan
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994 45
Table 1 The factors related to QOL
(1) The QOL of a patient is similar to that of a healthy
person.
© (2) QOL includes two main factors, i. e., "attitude to disease" (F) and "state of disease" (D), and with the changes of these factors the QOL of a patient also
changes.
Using these hypotheses, the Japanese questionnaire was constructed employing items related to both F and
D. Although QOL changes are systematic in patients,
they may be random in healthy persons. Because we
assessed QOL subjectively, the evaluation needed to be
done in a range of patient and normal groups.
Analysis
Construction of the questionnaire
There are generally three types of QOL measures. The
first type is a performance score, e. g. an activity score.
This type of measure is evaluated by a third party. The
second type of evaluation employs objective data such as
clinical findings. The third type is subjective evaluation
by the patient. The present QOL questionnaire belongs
Table 2 The QOL questionnaires
Table 3 Number of subjects.
46 Annals of Cancer Research and Therapy Vol. 3 NO. 1 1994
Fig. 1 Structure of disease state in the patients: Questionnaire A
(PDA, 17 items), scattergram of the first vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
Fig. 2 Example of a uni-dimensional structure (Guttman scale)
in QIII
Table 4 Similarity of QOL questionnaire structure between the patients and healthy persons.
Correlation coefficients were calculated for PDA vs. NDA and
for PDB vs.
NDB using 14 items in bothe cases.
Fig. 3 Structure of disease state in the patients: Questionnaire B (PDB, 17 items), scattergram of the first vector (horizonal
axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
to this third type. The subjects of our pretest were
patients with various diseases, so this limited the number of items in the questionnaire. Therefore, we divided the
items into two questionnaires (both including items
related to F and D), each of which contained the same
key questions.
Hayashi18) proposed that the Japanese have two uni-
que national characteristics. One is a unique view of interpersonal relations and the other concerns the num-
ber of categories in an item. In considering the Japanese
national character, a 3-point scale is appropriate , because many respondents select the neutral middle category even
if a 5-point scale is applied.
Subjects and methods
The subjects were inpatients and outpatients with
cancer and other diseases at Tokai University Hospital
as well as healthy persons (group N) such as business-
men, nurses, and doctors. The questionnaire for group N
included the same items as that for the patients , except
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994 47
Fig. 4 Structure of disease state in the patients: selected items (PD, 14 items), scattergram of the first
vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
Fig. 5 Structure of disease state in the healthy persons: selected items (ND, 10 items), scattergram of the first
vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
Table 5 Similarity between the score calculated
by QIII and the response score
Correlation coefficients were calculated for items common to
PDA vs. NDA and PDB vs. NDB in the patient or healthy person
groups.
for items related to disease and medical care. Because of
the limitation that a long questionnaire was not suitable
for the patients, we divided the items into two question-
naires (Questionnaire A and Questionnaire B). Both
questionnaires included the same 18 key questions. The items were classified into two groups related to D and F.
There were 30 items in Questionnaire A for inpatients
(IA), 30 items in Questionnaire B for inpatients (IB), 32 items in Questionnaire A for outpatients (OA), 28 items
in Questionnaire B for outpatients (OB), 22 items in
Questionnaire A for healthy persons (NA), and 20 items is Questionnaire B for healthy persons (NB).
For analysis, we combined inpatients and outpatients
together as a patient group (group P). For the items
related to D, we assessed 17 items of Questionnaire A for
group P and 14 items for group N, as well as 17 items of
Questionnaire B for group P and 14 items for group N. As the items relarted to F, we used 8 items of Question-
naire A for group P and 4 items for group N, as well as
10 items of Questionnaire B for group P and 5 items for
group N. (Table 2). The number of subjects in each group is summarized
in Table 3. The questionnaire was a self-administered
and the responses were written by the patient.
For statistical analysis, Hayashi's quantification
method III19,20) (QIII) was used to examine the structure
of the questionnaire and Spearman/Pearson correlation
coefficients were determined to examine the strength of
associations.
Analysis of D-related items,
(1) Structure Structure of Questionnaire A: QIII was used to confirm the structure of Questionnaire A and the follow-
ing results were obtained in the patient group (PDA: 17
items): the maximum latent root was 0.33, the second
was 0.20, and the third was 0.09. QIII analysis of these
data revealed that two dimensions were sufficient for
expressing the order of the items. A scattergram of the
48 Annals of Cancer Research and Therapy Vol. 3 No. 1 1994
Fig. 6 Structure of the attitude towards disease in the patients:
Questionnaire A (PFA, 8 items), scattergram of the first vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir- cles: better. triangles: neutral. crosses: worse)
Table 6 The response scores of the patients and healthy persons.
Values indicate the mean (standard deviation) of 66 patients and
57 healthy persons.
category values corresponding to the maximum latent
root (the first vector; the horizontal axis) and the
category values corresponding to the second maximum
latent root (the second vector; the vertical axis) is shown
in Fig. 1 The items related to a "worse physical state"
were located around the origin while those related to a "worse psychological state" were located a way from the
origin in the negative direction. For instance, responses
such as "never forget the illness", "pain", "irritation", "cannot do any thing"
, and "melancholy" were located in this order from the origin in the negative direction. In
the positive direction, the items revealed a symmetrical
pattern with those in the negative direction. Thus, the items were distributed along a cup type curve.
Assuming a dichotomous set of responses like that
shown in Fig. 2, the structure revealed by QIII indicates
that the items eliciting positive or negative responses
Fig. 7 Structure of the attitude towards disease in the patients: Questionnaire B (PFA, 8 items), scattergram of the first
vector (horizonal axis) and the second vector (vertical axis) Numbers in the figure correspond to those in Table 2 (cir- cles: better. triangles: neutral. crosses: worse)
Fig. 8 Structure of the attitude towards disease in the patients: selected items (PF, 8 items): scattergram of the first
vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994 49
Fig. 9 Structure of the final questionnaire items for the patients:
selected items (19, items): scattergram of the first
vector (horizonal axis) and the second vector (vertical axis)
Numbers in the figure correspond to those in Table 2 (cir-
cles: better. triangles: neutral. crosses: worse)
were respectively located in a positive or negative direc-
tion from the origin, while the items eliciting both
positive and negative responses were located around the origin as in a Guttman's Scalogram21). When the questi-
onnaire structure is of this type, we can express the grade
of QOL as an uni-dimensional scale using the first vector.
Furthermore, the selection of items becomes easy,
because the employment of items equally spaced along
the curve will maintain the original structure (or at least
maintain the same order among items). On this basis,
reduction of the questionnaire items was performed.
In order to express the original structure with fewer
items, we selected items based on the results of QIII
analysis of PDA, choosing items at equal intervals on the
curve. For Questionnaire A, we selected the items fromthe "worse direction" (negative direction; marked as x in
the figure) and corresponding items from the "better
direction" (positive direction; marked as o in the figure).
Thus, if two items were at a similar location in the worse
direction but had different locations in the better direc-
tion, we selected both items. Selection was continued
until the number of items was reduced to approximately
2/3 of the original. As shown in Fig. 1, items were
selected from each cluster located alnog the curve. In
total, 12 items were selected for PDA (Table 4).
In the healthy person group (NDA: 14 items), the
maximum latent root was 0.41 and the second was 0.28.
As was the case for PDA, we used the first and second
Fig. 10 Structure of the final questionnaire items for the healthy
persons: selected items (17, items): scattergram of the first vector (horizonal axis) and the second vector (vertical
axis)
Numbers in the figure correspond to those in Table
2 (circles: better. triangles: neutral. crosses: worse)
Fig. 11 Examination of the influence of wording: PDA: scatte-
gram of the old version (horizonal axis) and the new version (vertical axis)
When a category is located around the regression, it
denotes the agreement of each category of an item
between the old and new versions.
vectors for the interpretation of the questionnaire struc-
ture. The structure was also a clear cup type curve. In the
healthy person group, the items related to physical condi-
tion were limited to "nausea", "pain", "edema", and
"fatigue". The items which corresponded to a worse
physical state were located around the origin, while those
for a worse psychological state were located in the
50 Annals of Cancer Research and Therapy Vol. 3 No. 1 1994
negative direction from the origin. The items "dissatisfac-
tion", "anxity", "state of health", "melancholy", "annoy-
ance", and "worse sentiment" were located in order from
the origin in the negative direction. Eight items were
selected.
Structure of questionnaire B: In the patient group
(PDB: 17 items), the maximum latent root was 0.28, the second was 0.17, and the third was 0.14. We used the first
and second vectors and a scattergram is illustrated in Fig.
3. The questionnaire structure was very similar to that of
PDA, but the items in the negative direction varied.
Selection of items was carried out by basically the same
method as for PDA, except that the key questions already
selected for PDA were selected while ignoring the results
of PDB. As a result, 12 items were selected.
In the healthy person group (NDB: 14 items), the
maximum latent root was 0.30, the second was 0.16. and
the third was 0.11. We used the first and the second
vectors. The questionnaire structure was less regular than
that of NDA. However, the items related to physical condition, such as "fatigue", "melancholy", and "state of
health", were well ordered. In this case, 8 items were
selected.
In order to confirm the similarity of questionnaire
structure between the patient group and the healthy
person group, Spearman's rank correlation coefficients for the order of the first vector for the common 14 items
were calculated (PDA vs. NDA and PDB vs NDB,
respectively). This was done in both the positive and
negative directions for Questionnaire A and Question-
naire B, respectively. The results indicated that question-
naire structures were very similar, except for the negative
direction of Questionnaire B. This would be expected,
since there should be few responses in the worse direc-tion in the healthy person group.
(2) Structure obtained with the selected items Structure: In this analysis, subjects who answered
both Questionnaires A and B were used (P: 66, N: 57).
In the patient group (PD: 13 items), the maximum latent root was 0.37 and the second was 0.23. A scattergram of
the first and second vectors is shown in Fig. 4. Again, a
cup type curve is seen. The scattergram helps to define
the order of the items. In the positive direction, the least
important items were "better morbility" and "nausea".
More important items were "less discomfort (MUNE NO
TUKAE, in Japanese)" and "less loneliness". These were
followed in order by "do not think about illness", "more
satisfied", "better sentiment", "better physical condition", "less pain"
, "less irritation", and "able to do as one wishes". The most important items (indicating the best
condition) were "less anxiety" and "less tired".
On the other hand, the order of worse condition was
expressed by "cannot forget the illness", "pain", "irrita-
tion", and "tired", followed by "lack of mobility", "lone-
liness", "cannot do anything", "bad physical condition", "anxiety" , "discomfort", "nausea", and "unsatisfied". Finally, "worse sentiment" denoted the worst condition
(Fig. 4). In the healthy person group (ND: 10 items), the
maximum latent root was 0.28 and the second was 0.20.
However, the questionnaire structure was not a clear cup
type one and a definite order did not emerge (Fig. 5).
Scoring the responses: When the pattern of responses
forms a cup type curve, most of the information can be
explained uni-dimensionally, and the scores for each
item can be summed as a total score (the response score).
The response score was calculated as follows. A positive
response score was determined as the sum of the positive
responses and a negative response score as the sum of the
negative responses. A single response to a worse category
for an item added -1 to the negative response score, while a response to a better category for that item
added +1 to the positive response score. In the response
was neutral, nothing was added to either score. If this
questionnaire, the items had 3 categories (better, neutral, worse). The intervals between the categories in as item
could not always be equal, so the relative interval
between the category values determined by QIII might
differ for one item. Thus, there is no guarantee that the
categories were symmetrical in the positive and negative
directions. Since we cannot say that QOL is better expres-
sed by positive or negative items, we devised two scores.
For instance, we can use the negative response score for
the measurement of QOL in order to lessen the negative
factors impacting on a patient. In order to examine the response score correlations
with the sample scores calculated by QIII, Pearson corre-
lation coefficients were calculated (Table 5). The correla-
tion coefficients were comparatively high, except for the
positive response score of PF and the negative response score of NF.
Analysis of the attitude to disease
(1) Structure Structure of questionnaire A: For the patient
group (PFA: 8 items), the maximum latent root was 0.20 and the second was 0.18. A scattergram of the first and
second vectors is shown in Fig. 6. The structure was not
a cup type curve and the worse and better directions were
different, indicating that the structure of F could not be
expressed uni-dimensionally. Furthermore, some of the
responses were answered wrongly to what was originally
intended. Therefore, we removed these items and selected
3 items (Table 2). In the healthy person group, the same
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994 51
items were selected.
Structure of questionnaire B: For the patient
group (PFB: 10 items), the maximum latent root was 0.28 and the second was 0.24. Some of the items followed
a cup type curve and others were parallel to the first
vector (Fig. 7). The latter items had many missing
answers. Four items were selected by this analysis. In the
case of the healthy person group, the same items were
selected.
(2) Structure obtained with the selected items For the patient group (PF: 8 items), the maximum
latent root was 0.37 and the second was 0.32. The items
in the positive direction varied in order and were cluster-
ed around the origin. In the negative direction, some
items such as "financial apprehension", "not optimistic",
and "bad company" formed a cluster, while other items
such as "anxious about illness", "unexpected by the
family", and "stress" made a separate cluster (Fig. 8).
In the healthy person group (NF: 6 items), the
maximum latent root was 0.38 and the second was 0.25.
The structure was vague and the attitudes were less
defined than in the patient group.
Practical short QOL questionnaire
(1) Structure A total of 19 items for the patients and 17 items for the
healthy person group were analyzed. The scattergrams
are shown in Fig. 9 and Fig. 10, respectively. The patient
group showed a cup type structure and the items related to F were located on the curve. This provides support for
the hypothesis that D might be influenced by F. Assum-
ing this hypothesis, the positive direction of category
values for F is constant, but when D and F are analyzed
together the category values of F will very according to
the state of D. On the other hand, the healthy person
group did not show a clear cup type structure. One explanation for this is that the QOL of these subjects was
more varied than that of the patient.
(2) Improvement of wording As some of the items were considered to be poorly
expressed, we next tried to improve the wording on the
basis of a survey of the Japanese language. Subsequently,
analysis was carried out to examine the effect of differ-
ences in wording between the old and new versions. In
order to do this, the second surver was conducted using
the new version of the questionnaire. The subjects were
only patients (107 for Questionnaire A and 106 for
Questionnaire B with 55 common). The structure was then compared with respect to the order of the items in
the first vectors of each questionnaire. Figure. 11 illus-
trates the differences between the old version (horizontal
axis) and the new version (the vertical axis) of Question-
naire A. When an item is located around the regression
line, it denotes the agreement of each vectors. It was
found that 4 items were influenced by the wording. A
similar comparison was also carried out for Question-naire B and for both questionnaires combined.
From the results of these comparisons, the items in the
questionnaire were confirmed. This questionnaire was designated as the short QOL questionnaire (QL-20) (it is
presented in Japanese in the Japanese Abstract).
Discussion
Life expectancy has increased dramatically in Japan.
The concept on QOL was developed in Western coun-
tries and the majority of the available QOL measures
have been developed in English-speaking countries,
primarily in the United States. In recent years, the necessity of assessing of QOL in patients with cancer and
other chronic diseases, or even in healthy individuals, has been stressed.
The measurement of QOL, as well as the measurement
of mortality and morbidity, now has an important role in the evaluation of treatment in clinical trials. However,
very different meanings were assigned to the concept of
QOL22) by different investigators and the concept also varies with type of illness, the bodily functions involved,
and the physical and emotional problems. Although
most people in Japan have an intuitive understanding of
the concept of QOL, it is difficult to define it and solve
the translation problems related to adapting foreign
questionnaires. Accordingly, we designed a Japanese
QOL questionnaire. Even the phrase "QOL" could not be translated ade-
quately into Japanese. From 1985, we have discussed a suitable translation of QOL and the problems of assess-
ment in the "QOL Research Group", and we eventually
expressed QOL in Japanese as "IKIGAI" (matter related
to one's life). QOL is a concept that is subject to multiple
viewpoints. Good QOL is usually expressed in terms of
satisfaction, contentment, happiness, fulfillment, and the
ability to cope. Thus, physical well-being, mental well-
being, social well-being, and activity were considered as
the usual factors of QOL.
Our QOL questionnaire included items related to these
factors. From a clinical point of view, it is important to
recognize the complex nature of the doctor-patient rela-
tionship and to develop a pragmatic strategy for
introducing QOL measurenent into clinical medicine.
Even in European countries, the process of producing
high quality translations of QOL questionnaire is
extremely complex. From the comparison of national
characteristics between Japan and the United States
52 Annals of Cancer Research and Therapy Vol. 3 No. 1 1994
performed by Hayashi18), many differences were highligh-ted such as family relationships. Therefore, we thought
that a translated version of an English questionnaire
would not adequately measure QOL in Japan, and thus
we devised the present questionnaire.
In construction of the QOL questionnaire, we consid-
ered the working hypothesis that "F affects D and the
QOL alters as D cahnges". Thus, the process of improve-ment of QOL can be described as follows. F is improved
by the intervention of medical and other staff, while D is
improved through cognitive behavior therapy and other
measures along with the improvement of F. Subsequently
the patient's QOL will improve.
There are several other questionnaires for the measure-
ment of QOL23~25) but their validity has not been suffi-
ciently examined. Because most of them were devised on
the basis of factor analysis, the additivity of the scores
was not assured. In the case of the present questionnaire,
however, the response score was calculated validly
because our scale was based on the result of QIII, from which additive relations result. However, some problems
still remain in relation to the questionnaire structure. For
example, it is unclear whether the structure is maintained when the severity of illness changes. When an illness
becomes more severe, it is possible that the structure of
QOL will change and that items related to the physical state will vary more. Thus, further study of this issue is
necessary.
References
1) Bowling A: Measuring health: A review of quality of life
measurement scales. Open University Press, 1-55, 1991.
2) Calman KC: The quality of life in cancer patients-an
hypothesis. J Med Ethics, 10: 124-127, 1984.
3) Sato H: A concept of quality of life. Parma Medica, 7: 25-29,
1989.
4) Hamu S: Expectation of generalized health policy in Japan
supporting QOL for its aging society. Health Sciences, 4: 1-2,
1988.
5) Kayaba K, Nagashima N, Saito M, Omae T, Seki A, Arakawa
K, Ishii M, Kameyama M, Kokubu T, Fujii J.: The develop-
ment of quality of life scales for Japanese patients with
cardiovascular diseases. JACD, 25: 89-96, 1990.
6) Hinohara S: Quality of life of elderly patients. Chuoho-
kisyuppan, p 217 1988.
7) Spitzer WO, Dobson AJ, Hall J, Chesterman E, Levi J,
Shepherd R, Battista RN, Catchlove BR: Measuring the
quality of life of cancer patients. J Chron Dis, 34: 585-597, 1981.
8) Selby PJ, Chapman JAW, Amoli JE, Dalley D, Boyd NF:
The development of a method for assessing the quality of life
of cancer patients. Br J Cancer, 50: 13-22, 1984.
9) Schipper H, Clinch J, McMurry A, Levi M: Measuring the
quality of life of cancer patients: The functional living index-Cancer: Development and validation. J Clin Oncol, 2
:472-483, 1984.
10) Kurihara M. Izumi T. Denda T: Quality of life. Current
review. Cancer Chemothrapy, 1990-1991: 224-262, 1990.
11) Furue S: Quality of life of advanced cancer patients-prob- lems of the method for evaluation (in Japanese). Jpn J Cancer
Chemother, 14 (1): 1-10, 1987.
12) Takeda F: Cancer therapy and quality of life. Public Health,
55: 538-542, 1991.
13) Ganz PA, Haskell CM, Figlin RA, Soto NL, Siau J: Estimat-
ing the quality of life in a clinical trial of patients with
metastatic lung cancer using the Karnofsky performance
status and the functional living index-Cancer. Cancer, 61:
849-856, 1988.
14) Cox DR, Fletcher AE, Gore SM, Spiegelhalter DJ, Jones DR
:Quality-of-life assesment: Can we keep it simple? JR Statist
Soc A, 155, part 3: 353-393, 1992.
15) Evans DR, Burrns JE, Robinson WE, Garrett OJ et al: The
quality of life questionnaire: A multidimentional measure.
Am J Comm Psychol, 13: 305, 1983.
16) The EuroQOL Group: EuroQOL-a new facility for the measurement of health-related quality of life. Health Policy,
16: 199-208, 1990. 17) Stewart AL, Ware JE, Brook RH: Advances in the measure-
ment of functional status: Construction of aggregate indexes. Medical Care, 473-488, 1981.
18) Hayashi C: Statistical study on Japanese national character. J Japan Statist Soc (special issue). 71-95, 1987.
19) Hayashi C: Response reliability and multidimensional data
analysis. Classification and related methods of data analysis.
In: Bock HH (Ed), Elsevier Science Publishers, Amsterdam,
North-Holl, 1988.
20) Hayashi C: Quantification method III or correspondence
analysis in medical science. Ann Cancer Res Ther, 1: 17-21,
1992.
21) Guttman L: The quantification of a class of attributes. A theory and method of scale construction. The prediction of personal adjustment. In: Horst P (Ed), Social Science
Research Council, New York, 1941. 22) Calman KC: Definitions and dimensions of quality of life.
The quality of life of cancer patients. Raven Press, New York, 1987.
23) Yokoe T, Ishida T, Tominaga S, Kuroishi T, Morimoto T, Tashiro H, Itoh S, Abe R, Ota J, Horino T, Taguchi T,
Honda K, Sasakawa M, Kashiki Y, Yamamoto S, Enomoto
K, Ogita M, Yoshida K, Kido C, Fukuda M, Watanabe H,
lzuo M: Effect of mass screening for breast cancer from the
aspect of psychosocial of the quality of life. Jpn J Cancer Res,
84 (3): 364-370, 1993.
24) Kobayashi K, Nomura K, Wakasawa S, Sudou Y, Takahashi
T, Nukariya N, Hisakatsu S, Hayashihara K, Yoshimori K,
Murata A, Niitani H: Quality of life (QOL) and nutrition.
Jpn J Cancer Chemother, 18 (6): 1031-1032, 1991.
25) Shimozuma K, Tominaga T, Hayashi K, Kosaki G: Evalua-
tion of quality of life in breast cancer patients. J Jpn Cancer
Ther, 26 (8): 1504-1510, 1991.
Appendix The other members of the "Quality of Life Research Group" Keizou Takemi, Mituko Satou, Kimiko Inao, Katsurou
Haruyama, Minoru Kurihara, Youko Ishihara, Kunihiko Kobayashi, Norio Satou, Takashi Takahashi, Jirou Mano Kunihiro Iwata.
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994 53
Japanese version of questionnaire for quality of life
mesurement p 45~53
Kazue Yamaoka
本研究では,自 記式簡易QOL調 査票の日本語版の作成
を行った.こ の調査票は,患 者の一般的な(疾 患特有で
ない)QOLを 測定するための調査票であ り,疾 患に関連
する態度(F)が 病気の状態(D)に 影響 し,QOLが 変
化するという作業仮説に基づ いて作成された ものである.
調査票の作成 に当っては,FとDの それぞれに関連す る
項 目をスキームに応 じて考 えだ した.こ の簡略化を,患
者を対象とした実際の調査から行った.す なわち,患 者
を対象とした調査が可能なように,こ れ らの項 目をそれ
ぞれ共通の質問項 目を含む短い二つの調査票にわけ,実
際の調査結果に基づ き,数 量化III類による構造分析 を利
用 して簡略化を行った.
本研究では,こ の ような方法による一般的なQOLを 測
定するための 日本語版簡易調査票の作成方法について述
べ る.
Annals of Cancer Research and Therapy Vol. 3 No. 1 1994
Ann Cancer Res Ther
top related