Full text of DOI:10.1111/j.1467-842X.2001.tb00554.x

Measurement and Health
Application of the general health status questionnaire
SF36 to patients with gastrointestinal dysfunction:
initial validation and validation as a measure of change
A. Bensoussan
Abstract
Chinese Medicine Unit, School of Exercise & Health Science, University of
Western Sydney, New South Wales
Objective: To determine whether the Short
Form (SF36) Health Status Survey is a
valid measure of health status and health
change for patients with irritable bowel
syndrome (IBS).
S.W. Chang
School of Applied, Social & Human Sciences, University of Western Sydney,
New South Wales
Methods: The SF36 was self-administered
by 116 patients with IBS at the
commencement and end of a controlled
clinical trial. Patients were recruited through
two Sydney teaching hospitals and through
private gastroenterologists during 1997 and
treated with Chinese herbal medicine.
Results: The SF36 health concepts
demonstrated internal consistency,
construct validity and concurrent validity
when applied to patients with significant
bowel dysfunction. Patient scores on two
health scales of the SF36 (bodily pain,
general health) correlated significantly with
the bowel symptom scores recorded by
patients and gastroenterologists at the
beginning and end of the trial period.
Actively treated patients significantly
improved their scores in four out of eight of
the health scales of the SF36 and reported
overall improvement compared with
inactively treated patients.
R.G. Menzies
Department of Behavioral Sciences, University of Sydney, New South Wales
N.J.Talley
Department of Medicine, University of Sydney, Nepean Hospital, New South Wales
eneral health questionnaires are majority of sub-scales of physical and emo-
increasingly used as survey instru- tional health.5,12 The SF36 is now frequently
G
ments across population groups as recommended as the generic core in disease
a means of assessing health status. The prin- specific batteries.13
cipal assumption is that the questionnaires
The SF36 is a 36-item questionnaire that
are sufficiently sensitive to determine varia- consists of eight health concepts or scales.
tion in health status in the general popula- Each scale is made up of a number of dis-
tion.
tinct questionnaire items in the SF36. One
The SF36 is a self-reported measure of separately reported item also assesses change
health status that has been extensively vali- in a respondent’s health over the past year.
dated in overseas English-speaking countries The eight health scales are summarised in
and in Australia for adult age groups.1-7 It Table 1. Each of the scales is reported as a
Conclusions: The SF36 is a valid measure
of general health status in IBS patients, is
sensitive to the presence of IBS, and is
adequately sensitive to gastrointestinal
change in IBS patients.
has also been translated and validated in score from 0 (worst possible) to 100 (best
8
numerous other languages. Other com- possible).
monly used broader measures of health sta-
The SF36 has been applied widely to clini-
tus include the Sickness Impact Profile cal trials. Analyses of the physical and men-
Implications: While the SF36 general
health measure is used by the Australian
Bureau of Statistics and widely overseas,
until recently no data have been available
on the sensitivity of the SF36 to
9
(
(
SIP) and the Nottingham Health Profile tal health scales of the SF36 have
1
0
NHP). However, the SF36 has become an demonstrated that they are capable of dis-
increasingly popular choice among generic criminating between groups of persons:
health status measures and for the purposes healthy patients and those with moderate
of this study was chosen over the SIP and levels of psychiatric or physical illness.2,3,7,14
NHP for a number of reasons. The SIP takes The clinical validity of the instrument has
longer to complete, and both have been dem- been tested on samples with a range of
onstrated to be less sensitive to lower levels chronic disorders including low back pain,
of morbidity than the SF36.5,11 The NHP also menorrhagia, suspected peptic ulcer, vari-
appears less sensitive than the SF36 on the cose veins,3,15 asthma,16 migraine,17 HIV,18
gastrointestinal dysfunction or numerous
other disorders. The SF36 is not only
sensitive to the presence of IBS, it also
provides a useful adjunct to current
methods of evaluating treatment outcomes
for IBS, and potentially other disorders.
(Aust N Z J Public Health 2001; 25: 71-7)
Correspondence to:
Submitted: April 2000
Alan Bensoussan, School of Exercise & Health Science (Bankstown campus), University of
Western Sydney, Locked Bag 1797, Penrith South DC, New South Wales 1797.
Fax: (02) 9773 0998; e-mail: a.bensoussan@uws.edu.au
Revision requested: November 2000
Accepted: January 2001
2001 VOL. 25 NO. 1
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
71

Bensoussan et al.
Article
Table 1: Health concepts, number of items and levels, and summary of content for eight SF36 sub-scales and the
health transition item (summarised from Medical Outcomes Trust, 1994).30
Concepts
No. of items Summary of content
Physical functioning (PF)
Role limits – physical (RP)
Bodily pain (BP)
10
4
Extent to which health limits physical activities
Extent to which physical health interferes with work or other daily activities
Intensity and effect of pain on normal work
Personal evaluation of health
2
General health (GH)
Vitality (VT)
5
4
How energetic or tired
Social functioning (SF)
Role limits – emotional (RE)
Mental health (MH)
2
Extent to which physical health or emotional problems interfere with normal social activities
Extent to which emotional problems interfere with work or other daily activities
General mental health
3
5
Reported health transition (HT)
1
Evaluation of health compared with one year ago
post-operative functioning19,20, and more recently, functional
digestive disorders.21,22 This paper presents further validation of
the SF36 in IBS patients and new evidence of its capacity to meas-
ure health change following intervention.
interested in participating in the clinical trial had to undergo
referral to a specialist gastroenterologist and have a recent bowel
examination (generally a colonoscopy). Patients were subsequently
referred to receive CHM treatment at specified centres for 16
weeks continuously. At the CHM centres, patients were randomly
allocated into active (both standard and individualised herbal)
treatment or a placebo and were reviewed regularly by a tradi-
tional Chinese herbalist and by a gastroenterologist. Patients,
gastroenterologists and herbalists were all blinded to treatment
group. The trial protocol was approved by the ethics committees
of the University of Western Sydney Macarthur and the two par-
ticipating hospitals. The clinical outcomes of the trial interven-
tion have been reported elsewhere.33
IBS is a common bowel disorder that affects 10-20% of the
general population.23-26 A recent UK survey found that approxi-
mately one in three patients with gastrointestinal complaints who
attended a general practitioner were judged to have IBS.27
Approximately 30% of patients with IBS take sick leave as a con-
sequence of the symptoms, and it accounts for up to 50% of re-
ferrals to gastroenterologists.28,29 IBS represents significant
morbidity and economic cost to the community. If the SF36 is to
be broadly applicable and reliable as a general health status meas-
ure it ought to be able to discriminate patients with any of a wide
range of health disorders from healthy patients. IBS, with both
affective and physiological aspects, is ideally suited to testing the
SF36 health measure.
Patients and gastroenterologists completed bowel symptom
scales at the start and end of the 16-week treatment period.
The bowel scales consist of four 100 mm visual analogue scales
related to each symptom of IBS (pain/discomfort, bloating,
Objectives
Table 2: Inclusion criteria for IBS.
The present research sought to determine whether the Short
Form (SF36) Health Status Survey is a valid measure of health
status in patients with irritable bowel syndrome (IBS), a common
gastrointestinal disorder. In addition, the study sought to deter-
mine whether the SF36 is a valid measure of health change fol-
lowing intervention for IBS.
Age 18-75 years inclusive
Colonic evaluation (colonoscopy or barium enema) within the
previous five years (for 18-60 year olds) or within the previous three
years (for 61-75 year olds)
IBS by Rome criteria:
At least three months of continuous or recurrent symptoms of:
Abdominal pain or discomfort with at least some discomfort present
within the last two weeks
Methods
and
two or more of the following on at least one-quarter of occasions or
days:
The SF36 was self-administered by patients who were present-
ing for treatment of their IBS with Chinese herbal medicine
i. abdominal distension that is visible or felt by tight clothing
ii. pain relief with bowel action
(
CHM). Patients completed the SF36 at the beginning and end of
a 16-week intervention period. It took approximately five min-
utes to complete the questionnaire.
iii. more frequent stools with onset of pain
iv. looser stools with onset of pain
v. mucous in stools
All subjects were recruited and screened in two public hospital
gastroenterology units and by private gastroenterologists in
Sydney during 1997. Patients between the ages of 18 and 75 years
vi. feeling of incomplete evacuation
Bowel symptom scale: At least one marking on the visual analogue
scales for IBS symptoms to be at least 20 mm from the ‘not present’
end of the scale.
(
inclusive) were assessed by routine clinical workup for IBS
and according to the established Rome criteria for IBS (see Table
Normal liver function tests and full blood count, urea and creatinine
(within the last two weeks).
2
).31,32 Symptom presentations were more than trivial – patients
7
2
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 1

Measurement and Health
The SF36 in gastrointestinal dysfunction
Table 3: Mean scores for SF36 health scales and internal consistency measures (average intra-scale correlations) at
commencement and end of intervention period (all patients).
At start
n
At end
n
Health scale
Mean score (SE)
Cronbach’s alpha
Mean score (SE)
Cronbach’s alpha
Bodily pain
50.9 (2.3)
56.8 (2.3)
60.7 (1.9)
80.2 (2.1)
58.4 (4.1)
53.9 (3.9)
62.2 (2.5)
45.9 (2.0)
114
106
112
102
113
108
115
112
0.87
0.84
0.82
0.87
0.87
0.85
0.87
0.82
59.7 (2.4)
61.0 (2.7)
63.1 (2.3)
80.4 (2.2)
64.2 (4.3)
59.6 (4.1)
66.8 (2.7)
45.9 (2.4)
95
85
93
90
95
96
95
92
0.89
0.89
0.89
0.89
0.86
0.83
0.81
0.88
General health
Mental health
Physical functioning
Role limits – emotion
Role limits – physical
Social functioning
Vitality
constipation, diarrhoea). Validation of the bowel symptom scale,
used as a means of assessing response to treatment intervention,
identify the degree of correlation between the patients’SF36 scores
and their bowel symptom scores.
3
3
was previously reported. Further evaluation items administered
at the end of the trial included patient and gastroenterologist rat-
ings of overall improvement.
The sensitivity of the SF36 to change in health status of IBS
patients was tested by examining changes in SF36 scores from
beginning to end of the treatment period (paired sample t-tests).
Changes in bowel symptom scores from beginning to end of treat-
ment (recorded by gastroenterologists and patients) were also
compared against item 2 responses of the SF36 which asks ‘Com-
pared to one year ago, how would you rate your health in general
now?’ Similarly, responses to this item were compared against
categorical responses to a bowel symptom questionnaire item ask-
ing both gastroenterologists and patients about the extent of over-
all improvement (Kendall’s rank correlation coefficients).
Chi-square was used to examine the differences in categorical
responses to the SF36 item 2 between active and inactive treat-
ment groups at the end of treatment. p values were all 2-tailed
unless otherwise indicated, a level of significance was set at 0.05.
Missing scale and item scores were not replaced.
Statistical analysis
The SF36 was tested for validity and reliability in its applica-
tion to IBS sufferers. Internal consistency was assessed using
inter-item correlations (and Cronbach alphas calculated) for the
individual health scales. To test construct validity the student
t-test was used to compare SF36 scores for IBS patients and nor-
mative data for the Australian and UK populations. Factor analy-
sis was also applied to determine the construct validity of the
individual health scales of the SF36. Tests for concurrent validity
were used to determine which health scales in the SF36 most
clearly reflect the reported gastrointestinal dysfunction. Pearson
product moment correlation was employed in this analysis to
Figure 1: Comparison of age and
sex standardised SF36 scores
for the IBS cohort against age
and sex standardised normative
population data for Australia1
and unstandardised normative
1
1
20
00
8
6
4
2
0
0
0
0
0
6
population data for the UK at
the commencement of the trial.
Pain
Gen
hlth
Mntl
hlth
Phys fn Limits- Limits- Soc fn Vitality
emot
phys
IBS patients
Normative Aust
Normative UK
SF36 health subscales
2001 VOL. 25 NO. 1
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
73

Bensoussan et al.
Article
Table 4: Principal components factor analysis on SF36
scales at pre- and post-treatment with percentage
variance accounted for by principal factors (all patients).
Table 5: Pearson-product moment correlation
coefficients between patient responses on SF36 and
gastroenterologist and patient interpretations of severity
of bowel symptoms (all patients).
Pre-treatment
Post-treatment
Eigen- % of
value variance
accounted
variance
Eigen-
value
% of
variance
accounted
for by
Pre-treatment correlations Post-treatment correlations
Gastro-
Patient
Gastro-
Patient
enterologist ratings
ratings
enterologist ratings
ratings
principal
factor
principal
Bodily pain
-0.31**
-0.28**
-0.18
-0.43**
-0.24*
-0.14
-0.21*
-0.14
-0.13
-0.17
-0.21*
-0.28*
-0.30*
-0.12
-0.02
-0.31**
-0.25*
-0.15
-0.18
-0.44**
-0.38**
-0.20
Health scales
factor
General health
Mental health
Bodily pain
1.7
3.1
3.0
4.8
2.4
2.7
1.8
2.6
85.8
62.0
59.0
48.5
79.3
68.5
88.4
65.4
1.8
3.5
3.5
5.2
2.4
2.6
1.7
2.9
88.4
69.7
70.5
51.9
78.5
65.8
83.9
73.4
General health
Mental health
Physical function -0.15
-0.12
Role emotion
Role physical
Social function
Vitality
-0.18
-0.21*
-0.25**
-0.16
-0.24*
-0.40**
-0.30**
-0.32**
Physical functioning
Role limits – emotion
Role limits – physical
Social functioning
Vitality
Note:
*
p<0.05, ** p<0.01
Results
A principal components factor analysis was also applied to each
health scale for both the first and second administrations of the
SF36. With the exception of the ‘physical functioning’ scale, on
each occasion the scales revealed only one factor with an
eigenvalue greater than 1.That is, all items within each scale (with
the exception of the ‘physical functioning’ scale) had significant
correlation with the first factor, confirming that items within each
scale are measuring the same construct. For the ‘physical func-
tioning’ scale on both occasions the items ‘walking one block’
and ‘bathing or dressing oneself’ were both more weakly corre-
lated with the remainder of items in the scale. This contributed to
a second factor in the ‘physical functioning’ scale. When com-
pared with ‘vigorous activity’ also within the same scale, this
would be expected given the vast difference in physical function-
ing these items signify.
All 116 patients who entered the trial completed the SF36 on
commencement (2.9% missing values), 99 patients completed it
on exit or completion of the trial (3.2% missing values).The mean
age of patients surveyed was 47.8 years (SD 14.1). Seventy-five
participants were female (64.7%) and 41 male (35.3%). All sub-
jects had a gastroenterologist-confirmed diagnosis of IBS.
Internal consistency
The internal consistency of the SF36 health scales were
assessed using inter-item correlations within each of the eight
health scales. Cronbach’s coefficient alpha (representing average
inter-item correlations) were high for all health scales of the SF36
on both occasions (see Table 3). Within each scale there were
strong individual inter-item correlations, except for the items to
do with ‘bathing’or ‘dressing and walking one block’, which were
both poorly correlated with other items representing the ‘physi-
Table 4 summarises eigenvalues and the percentage of variance
accounted for by each principal factor.
cal functioning’ scale. This has been noted previously in other
patient groups.2
Concurrent validity
The health scales were tested for concurrent validity at the com-
mencement and end of treatment against both patients’ and gas-
troenterologists’ ratings of severity of bowel dysfunction. The
Pearson correlation coefficient was in the range of -0.44 to -0.12
for each health scale against the total bowel symptom score as
completed by patients and gastroenterologists (see Table 5). The
only exception to this was the ‘physical functioning’scale, which
correlated poorly with the gastroenterologist measure of severity
of bowel symptoms (r=-0.02).
Construct validity
Construct validity is reflected in the ability of the SF36 to dis-
tinguish a patient cohort with a significant health dysfunction
from healthy patients. The IBS patient responses for each health
scale of the SF36 were compared with responses in the general
population (see Figure 1).
Mean scores for each of the eight health scales of the SF36
were standardised for age and sex for our IBS cohort of patients.
This was achieved by adjusting mean scores for each age group
and sex in the IBS group to reflect the normal population distri-
The SF36 scales that correlate most closely with the severity of
IBS as perceived by gastroenterologists and patients include
1
bution figures provided by the Australian Bureau of Statistics. A
‘
bodily pain’ and ‘general health’. Limited health scales of the
student t-test comparison of the standardised means between the
known IBS group and normative data for the Australian popula-
SF36 therefore exhibit concurrent validity with patient and
gastroenterologist interpretation of the severity of the specific
bowel symptoms. Other health scales demonstrated some corre-
lation with severity of bowel symptoms, although these were
6
tion and an unstandardised UK population sample showed
significance at p<0.01 on all health scales.
7
4
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 1

Measurement and Health
The SF36 in gastrointestinal dysfunction
Table 6: Mean scores for health sub-scales of actively
treated patients at commencement and end of
intervention period.
Table 7: Difference in overall health improvement
reported by active and inactive treatment groups at end
of treatment (chi-squared). Due to small cell sizes, the
five categories of response to the SF36 item 2 have been
collapsed into three. Percentages are within active or
inactive groups.
Pre-treatment
Post-treatment
n
Sub-scale
Mean score (SE) Mean score (SE) pairs
Bodily pain
51.7 (2.8)
60.3 (2.7)
62.4 (2.1)
78.8 (2.6)
57.8 (4.9)
56.7 (4.6)
62.0 (3.0)
48.4 (2.5)
63.4 (2.6)**
64
54
59
57
62
60
64
59
67.2 (3.2) **
Post-treatment
SF36 item 2
Active
no. (%)
Inactive
no. (%)
p value
General health
Mental health
Physical functioning
Role limits – emotion
Role limits – physical
Social functioning
Vitality
65.9 (2.7)
80.9 (2.9)
68.7 (4.8)
66.5 (4.7) **
70.5 (3.3) **
49.9 (2.8)
Much or somewhat better
About the same
27 (40.3)
35 (52.2)
5 (7.5)
6 (18.8)
15 (49.9)
11 (34.4)
0.002
}
Much or somewhat worse
Note:
SF36 item 2 between active and inactive treatment groups on com-
mencement of the trial, patients in the active group reported sig-
nificant improvement over those in the inactive group by the end
*
p<0.05, ** p<0.01
2
of the trial (χ =12.9, df=2, p=0.002; see Table 7). This is consist-
ent with the principal outcome measures used in this trial.33
neither sufficiently strong nor consistent. This is to be expected
as the bowel symptom scales do not focus on activities specified
in these items. The SF36 clearly assesses additional material to
the bowel scales that focus on four key gastrointestinal symp-
toms.
Conclusions
The SF36 Health Status Survey has been demonstrated to have
adequate internal consistency and both construct and concurrent
validity when applied to patients with irritable bowel syndrome.
It is capable of distinguishing between known groups (healthy
and dysfunctional GIT) effectively. Given the variety of presenta-
tions of IBS (from constipation to diarrhoea predominant, and
from presence to absence of discomfort and bloating), it appears
the SF36 could be sensitive to most bowel dysfunction, despite it
comprising no direct questions related to bowel function.
Of interest, IBS impacted significantly on all health scales of
the SF36, in comparison with healthy populations (see Figure 1).
This supports other findings in the US and Europe.21,22,36 Impor-
tantly, these results indicate that IBS has a significant impact not
only on the measures of physical health in the SF36 but also on
other health measures to do with energy levels, and mental and
social well-being. These findings demonstrate the real and di-
verse impact of IBS on patients’ lives, an impact not restricted to
the physical expressions of pain, bloating, constipation and diar-
rhoea.
The significant correlation with the ‘bodily pain’ and ‘general
health’scales also indicates the SF36 may be a valid, albeit weaker,
measure of the severity of IBS presentations – with more severe
bowel dysfunction correlating with lower scores on these two
health scales of the SF36.
Sensitivity to change
As reported elsewhere, IBS patients receiving CHM improved
significantly over those patients receiving placebo.33 Hence, it
was of interest to see if the SF36 was sufficiently sensitive to
detect these changes in the active treatment group. The SF36 has
not been previously tested as a measure of change in health status
following intervention. It has, however, been reported as suffi-
ciently sensitive over a 12-month period to note health changes
in older patients with debilitating disease,34 and over 12 to 18
3
5
months in patients with psoriatic arthritis. The CHM interven-
tion period was 16 weeks.
Paired sample t-tests were used to compare mean patient SF36
scores for each health scale from commencement to end of the
trial for all actively treated patients (see Table 6). While patient
mean scores on all scales showed improvement, changes in scale
scores were statistically significant (p<0.05) for ‘bodily pain’,
The SF36 is a valid measure of general health status in IBS
patients in that it is sensitive to the presence of IBS. An important
caveat is that this cohort of patients on whom the SF36 was tested
was generally at the more severe end of clinical presentation. The
mean duration of symptoms was reported as between five and 10
years and all patients were required to have had a bowel examina-
tion within the previous five years. The ability of the SF36 to
reflect changes in the health status of mild to moderate cases of
IBS remains to be established.
‘
general health’, ‘role limits – physical’ and ‘social functioning’.
No significant differences were noted in any of the scales for pla-
cebo patients over the duration of the trial.
At the end of the treatment period, the changes in the bowel
symptom scores recorded by both gastroenterologists and patients
correlated significantly with patient responses to item 2 of the
SF36, which inquires about degree of overall health improvement
compared with one year ago (patients τ=-0.31, gastroenterolo-
gists τ=-0.37) (p<0.01 on both occasions).
The SF36 also appears adequately sensitive to gastrointestinal
change in IBS patients, albeit within the limitations of the four
health constructs: ‘bodily pain’, ‘general health’, ‘role limits –
physical’ and ‘social functioning’. This improvement had been
3
3
While there were no significant differences in response to the
separately confirmed by changes in bowel symptom scores. All
2001 VOL. 25 NO. 1
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 75

Bensoussan et al.
Article
four scales directly reflect important aspects of the IBS condition
Acknowledgements
–
bodily pain, general health, general physical well-being, and
The authors wish to acknowledge the financial assistance of
the University of Western Sydney Macarthur (A$14,000), Mei
Yu Imports for the contribution of all herbal materials, and Pan
Laboratories for assistance with the design and preparation of the
placebo. This project would not have been completed without the
ongoing support of gastroenterologists, nursing staff, research
assistants and Chinese herbalists. Particular thanks is given to
Kathryn Taylor, and Sue Huntley; to the Gastroenterology Units
of Nepean Hospital and Concord Hospital; to herbalistsYu Long
Yu, Henry Liang and Anna Guo; and to Drs Gavin Barr, Philip
Barnes, Chris Pokorny, John Garvey, Tom Borody and Laura
Pearce, Mosman Medical Centre; and, of course, to all patients
that contributed their time and more to this study.
interference in social life. Sensitivity of the SF36 was restricted
to these four scales. The scales insensitive to change include
‘
physical functioning’, ‘mental health’, ‘role limits-emotional’and
vitality’. ‘Physical functioning’ has already been identified as
‘
problematic because of its item span across a wide range of con-
trasting activities. ‘Mental health’and ‘role limits-emotional’were
understandably less sensitive to IBS symptom change, as they
measure non-physical symptoms not recorded by the tradition-
ally physical IBS symptom scales. Similarly, ‘vitality’ was nei-
ther directly measured by the IBS scale, nor does it correlate closely
with severity of symptoms.
Following active treatment the SF36 recorded significant
improvements, not only in ‘bodily pain’ and ‘role limits – physi-
cal’, but also for the ‘general health’ and ‘social functioning’ sub-
scales. These latter two scales are distinct to the more physical
changes commonly recorded by bowel symptom scores. Clearly,
improvement in the IBS condition resulted in a perceived improve-
ment in other aspects of patient lives. It can be argued that the
SF36 added a degree of sensitivity to the changes in patient well-
being not recorded by any physical measures of change in bowel
function, and again reflects the broader health impact of IBS.
References
1
.
Australian Bureau of Statistics. National Health Survey: SF36 Population
Norms Australia. Canberra: ABS; 1997.
2
.
McCallum J. The SF-36 in an Australian sample: validating a new, generic
health status measure. Aust J Public Health 1995;19(2):160-6.
3. Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT. The SF-36
health survey questionnaire: an outcome measure suitable for use within the
NHS? Br Med J 1993;306:1440-4.
4
.
Lyons RA, Perry HM, Littlepage BNC. Evidence for the validity of the short-
form 36 questionnaire (SF-36) in an elderly population. Age Aging
1
994;23:182-4.
5
.
Brazier JE, Harper R, Jones NMB, O’Cathain A, Thomas KJ, Usherwood T, et
al. Validating the SF-36 health survey questionnaire: New outcome measure
for primary care. Br Med J 1992;305:160-4.
Implications
6
7
.
.
Jenkinson C, Coulter A, Wright L. Short form (SF 36) health survey question-
naire: Normative data for adults of working age. Br Med J 1993;306:1437-40.
McHorney CA, Ware JE, Raczek AE. The MOS 36-item short-form health
survey (SF-36): II. Psychometric and clinical tests of validity in measuring
physical and mental health constructs. Med Care 1993;31(3):247-63.
Special issue, J Clin Epidemiol 1998; 51(11).
Deyo RA, Inui TS, Leininger JD, Overman SS. Measuring functional out-
comes in chronic disease: A comparison of traditional scales and a self
administered health status questionnaire in patients with rheumatoid arthri-
tis. Med Care 1983;21(2):180-92.
The SF36 general health measure is used widely internation-
ally as a component of general and disease-specific health meas-
ures, and for the collation of Australian health data by the
Australian Bureau of Statistics. For our cohort of patients, the
SF36 has proved a consistent and valid measure of health. De-
spite its widespread use and validation in varying clinical cir-
cumstances, the SF36 general health measure has remained largely
untested as a tool to evaluate treatment outcomes.
8
9
.
.
10. Hunt SM, McEwan J, McKenna SP. Measuring Health Status. Beckenham:
Croom Helm; 1986.
1
1. Hall J, Hall N, Fisher E, Killer D. Measurement of outcomes of general prac-
These findings indicate that specific health sub-scales of the
SF36 (bodily pain, general health, role limits – physical, social
functioning) registered improvement in patients consistent with
determinations made by gastroenterologists and patients using
symptom-specific bowel scores. Importantly, these findings also
lend some support to the use of the SF36 in the evaluation of
other clinical interventions. The limitations are that any particu-
lar clinical disorder may only register substantial changes in spe-
cific health sub-scales and not in all. It should be remembered
that in our study while all health sub-scales improved by the end
of treatment, only the four above registered significant improve-
ment. These could be used as additional measures of clinical out-
come in IBS.
tice: Comparison of three health status measures. Fam Practice 1987;4:117-
2
3.
1
1
1
2. Kind P, Carr-Hill R. The Nottingham health profile: a useful tool for epide-
miologists? Soc Sci Med 1987;25:905-10.
3. Bowling A. Measuring Disease. Buckingham: Open University Press; 1995.
p281.
4. McHorney CA, Ware JE, Lu JFR, Sherbourne CD. The MOS 36-item short-
form health survey (SF-36): III. Tests of data quality, scaling assumptions,
and reliability across diverse patient groups. Med Care 1994;32(1):44-66.
5. GarrattAM, Macdonald LM, Ruta DA, Russell IT, Buckingham JK, Krukowski
ZH. Towards measurement of outcome for patients with varicose veins. Qual
Health Care 1993;2:5-10.
6. Bousquet J, Knani J, Dhivert H, Richard A, Chicoye A, Ware JE, et al. Quality
of life in asthma: I Internal consistency and validity of the SF-36 question-
naire. Am J Respir Crit Care Med 1994;149:371-5.
1
1
1
1
7. Osterhaus JT, Townsend RJ, Gandek B, Ware JE. Measuring the functional
status and well-being of patients with migraine headache. Headache
1
994;34:337-43.
8. Wachtel T, Piette J, Mor V, Stein M, Fleishman J, Carpenter C. Quality of Life
in persons with human immunodeficiency virus infection: measurement by
the medical outcomes study instrument. Annu Intern Med 1992;116:129-37.
9. Phillips RC, Lansky DJ. Outcomes management in heart valve replacement
surgery: Early experience. J Heart Valve Dis 1992;1(1):42-50.
0. Kantz ME, Harris WJ, Levitsky K, Ware JE, Davies AR. Methods for assess-
ing condition-specific and generic functional status outcomes after total knee
replacement. Med Care 1992;30(5 Suppl): MS240-252.
Utilising the SF36 general health questionnaire offers the
advantage that it allows patients to communicate the impact of
their illness and assesses clinical outcomes across a broad range
of health aspects. Disease specific measures do not usually pro-
vide this feature. The SF36 is a simple, easily administered in-
strument. Its suitability as a method of evaluating treatment
outcomes should be explored for other clinical conditions.
1
2
2
1. Chassany O, Marquis P, Scherrer B, Read NW, Finger T, Bergmann JF, et al.
Validation of a specific quality of life questionnaire for functional digestive
disorders. Gut 1999;44(4):527-33.
7
6
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 1

Measurement and Health
The SF36 in gastrointestinal dysfunction
2
2
2
2. Hahn BA, Yan S, Strassels S. Impact of irritable bowel syndrome on quality
of life and resource use in the United States and the United Kingdom. Diges-
tion 1999;60(1):77-81.
3. Talley NJ, Zinsmeister AR, Van Dyke C, Melton LJ. Epidemiology of colonic
symptoms and the irritable bowel syndrome. Gastroenterology 1991;101:927-
30. Medical Outcomes Trust. SF-36 Health Survey Scoring Manual Language
Adaptations: Australia/New Zealand, Canada, United Kingdom. Boston:
Medical Outcomes Trust; 1994.
31. Thompson WG, Creed F, Drossman DA, Heaton KW, Mazzacca G. Func-
tional bowel disease and functional abdominal pain. Gastroenterol Int
1992;5:75-91.
32. Talley NJ, Nyren O, Drossman DA, Heaton KW, Veldhuyzen van Zanten SJO,
Koch MM, et al. The irritable bowel syndrome: Toward optimal design of
controlled treatment trials. Gastroenterol Int 1993;6:189-211.
33. Bensoussan A, Talley NJ, Hing M, Menzies R, Guo A, Ngu M. Treatment of
irritable bowel syndrome with Chinese herbal medicine: a randomized con-
trolled trial. J Am Med Assoc 1998;280(18):1585-9.
34. Wolinsky FD, Wan GJ, Tierney WM. Changes in the SF36 in 12 months in a
clinical sample of disadvantaged older adults. Med Care 1998;36(11):1589-
98.
3
4.
4. Heaton KW, O’Donnell LJD, Braddon FEM, Mountford RA, Hughes AO,
Cripps PJ. Irritable bowel syndrome in a British urban community: consult-
ers and non-consulters. Gastroenterology 1992;102:1962-7.
5. Jones R, Lydeard S. Irritable bowel syndrome in the general population. Br
Med J 1992;304:87-90.
6. Talley NJ, Boyce PM, Owen BK, Newman P, Paterson KJ. Initial validation of
a bowel symptom questionnaire and measurement of chronic gastrointestinal
symptoms in Australians. Aust N Z J Med 1995;25:302-7.
7. Thompson WG, Heaton KW, Smyth GT, Smyth C. Irritable bowel syndrome
in general practice: prevalence, characteristics and referral. Gut 2000;46:78-
2
2
2
2
2
35. Husted JA, Gladman DD, Cook RJ, Farewell VT. Responsiveness of health
status instruments to changes in articular status and perceived health in
patients with psoriatic arthritis. J Rheumatol 1998;25(11):2146-55.
36. Hahn BA, Kirchdoerfer LJ, Fullerton S, Mayer E. Patient-perceived severity
of irritable bowel syndrome in relation to symptoms, health resource utiliza-
tion and quality of life. Aliment Pharmacol Ther 1997;11(3):553-9.
8
2.
8. Drossman DA, Li Z, Andruzzi E, Temple R, Talley NJ, Thompson WG, et al.
US householder survey of functional gastrointestinal disorders: prevalence,
sociodemography and health impact. Dig Dis Sci 1993;38:1569-80.
9. Farthing MJG. Irritable bowel, irritable body, or irritable brain? Br Med J
1
995;310:171-5.
2001 VOL. 25 NO. 1
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
77