Clinical rheumatology, 1986, 5, N ~ 2 231-241

A preliminary evaluation of the dimensionality and clinical

importance of pain and disability in osteoarthritis of the hip
and knee

N. BELLAMY, W.W. BUCHANAN

Department of Rheumatology, University of Western Ontario, London, Canada;

D e p a r t m e n t o f R h e u m a t o l o g y , M c M a s t e r University, H a m i l t o n , O n t a r i o , C a n a d a .

SUMMARY Current methods o f clinical assessment in osteoarthritis show a high degree

o f variability. By contrast, patients with rheumatoid arthritis may be evaluated using a
number o f standardised and validated indices. One hundred patients with primary os-
teoarthritis o f the hip and knee were interviewed in order to determine the dimensionality o f
their discomfort and disability and to define the clinical importance o f each component
item. The symptomatology o f osteoarthritis was captured by five pain, one stiffness, twenty-
two physical, eight social and eleven emotional items. In spite o f a high degree o f variability
in the frequency o f involvement o f the individual items, their clinical importance was similar
both within as well as across dimensions. Further studies are indicated to determine the
reliability, validity and responsiveness o f each o f the items identified as a prelude to devel-
oping a standardized method o f assessing patients with osteoarthritis o f the hip and knee.

Key words: Osteoarthritis, Pain, Disability, Clinical Importance.

INTRODUCTION clinical trials, we have reported in a previous

edition of this journal, a review of 63 clini-
Discomfort (pain and stiffness) and dis- cal studies of nonsteroidal anti-inflammato-
ability (physical, social and emotional) are ry drugs reported between 1962 and 1982
the m a j o r symptoms of osteoarthritis. In and have observed a high degree of variabili-
spite of the greater prevalence of osteoar- ty in the outcome measures employed (4). In
thrifts, more attention has been directed to- addition to lacking any standardisation, cur-
wards the study of functional decrements rent measures presume a validity extrapolat-
and the quality of life of pmiems with rheu- ed from the rheumatoid arthritis literature.
matoid arthritis (1) than to patients with de- Thus, the majority of indices which have
generative forms of arthritis (2,3). In respect been developed for use in rheumatic diseases
to outcome measurement in osteoarthritis have been based on patients with
rheumatoid arthritis (5-29). However, fun-
damental differences exist between patients
Received 18 March 1985, with rheumatoid and osteoarthritis in re-
Revision - accepted 10 February 1986 spect of the age of onset, distribution of
Correspondence to: DR. N. BELLAMY joint involvement, natural history of the dis-
Suite 402 A, Victoria Hospital (Westminster Tower),
800 Commissioners Road East London, Ontario, ease and response to treatment. Only the
Canada, N6A 405 Doyle (7) and Lequesne (24) indices have
232 N . Bellamy, W.W. Buchanan

been expressly developed for evaluating PT, P JR) and two clinical epidemiologists
patients with osteoarthritis. However, the experienced in clinical measurement in the
Doyle Index is unidimensional and is a mod- rheumatic diseases (CG, LC). Initial
ification of the Ritchie Index while the Le- questions were open-ended and probed the
quesne Index is oligodimensional and uti- clinical importance and characteristics of
lizes a restricted number of response alterna- any pain, stiffness, physical, social or emo-
tives. tional dysfunction. Once spontaneous re-
In view of these deficiencies in outcome sponses to these questions were exhausted, a
measurement in osteoarthritis clinical trials, battery of closed-ended questions derived
we are currently undertaking a series of from six existing questionnaires (10-13, 25,
studies to rationalize the measurement pro- 27, 31, 32) was used to complete the assess-
cess pertaining to patients with primary os- ment of each dimension and quantitative
teoarthritis of the hip and/or knee. In the any sources of discomfort or disability de-
present study the extent of each of five con- tected.
tent domains was assessed and component The following data were recorded : 1) The
items ranked according to their prevalence presence or absence of each of several types
and clinical importance. The objective was of discomfort or disability (Table I-IV); 2)
to define the dimensionality of pain and dis- The frequency with which each type of dis-
ability and identify those component items comfort or disability occurred (daily,
having the greatest clinical importance in a weekly, fortnightly, monthly or less) and 3)
group of potential drug-study patients. The importance of the discomfort or disabil-
ity to the patient (0 = none, 1 = slight, 2 =
MATERIALS A N D M E T H O D S moderate, 3 -- very, 4 = extremely). It
should be noted that patients were specifi-
One hundred out-patients with osteoar- caUy asked to record the perceived impor-
thritis of the hip a n d / o r knee were selected tance of each type of discomfort or disabili-
for study. To be eligible patients had to ful- ty reported in order to assess its clinical rele-
fill the following criteria: 1) Attend a vance. Furthermore, the discomfort and dis-
rheumatological clinic at either the Univer- ability sought was specified as having been
sity of Western Ontario, London, or recently experienced and directly related to
McMaster University Medical Centre, H a m - osteoarthritis of the hip a n d / o r knee. Thus,
ilton ; 2) Be ambulatory; 3) Have symptom- each patient was asked to report only those
atic primary osteoarthritis affecting at least symptoms which they felt were the direct re-
one hip or knee and requiring treatment with sult of their articular disease.
a nonsteroidal anti-inflammatory analgesic During questionnaire construction, items
medication; 4) Have minimal or no spinal directed specifically at patients o f one or
symptoms; 5) Be unrestricted (in their other sex (e.g. ironing) were avoided and the
functional capacity) by any co-morbid con- questions rephrased in more general terms
dition and; 6) Not have had prior hip or (e.g. light domestic duties). Patients were
knee replacement surgery or an osteotomy. not asked about sexual function in order to
The patients selected for study would a l l avoid embarrassment and because this has
have been eligible for a clinical trial of n6n- been previously noted to inhibit responses
steroidal anti-inflammatory drug therapy even to subsequent non-sexual questions.
since they were all typical of patients com- Before being formally applied the
monly used in such trials. questionnaire was pre-tested in 15 osteoar-
The survey questionnaire was developed thritic patients in order to assess its compre-
by a peer review process utilizing the opin- hensibility and feasibility. Thereafter, the
ions of four rheumatologists (WWB, NB, questionnaire was administered to 90
 A preliminary evaluation o f the dimensionality 233

patients by face-to-face interview (using Table III Social function rank ordered by prevalence
(P)
t r a i n e d i n t e r v i e w e r s ) a n d t o a f u r t h e r 10
patients by telephone. Telephone interviews
ITEM P MIS [DW%]
were permitted in order to be able to survey
patients, at either end of the severity spec- Restricted leisure activ-
trum, who though ambulatory did not wish ities .54 2.56 87
Attendance at commu-
to make a non-essential journey. Patients nity events .27 2.15 77
who were confined to a bed or wheelchair Attendance at church .23 2.52 74
w e r e e x c l u d e d f r o m t h e s u r v e y as t h e y w o u l d Relations with spouse .20 2.65 90
not normally have been admitted to a drug Relations with family .18 2.67 89
Relations with friends .14 2.64 79
Table I Pain rank ordered by prevalence (P) Relations with others .11 2.55 91
Dancing .03 2.33 33

ITEM P MIS [DW%]

Table IV Emotional function rank ordered by preva-
Walking .77 2.58 96 lence (P)
Stairs .75 2.62 94
In bed .67 2.63 96 ITEM P MIS [DW%]
Weight bearing .57 2.51 94
Sitting/lying .56 2.57 95 Frustration .56 2.44 86
Bending .01 3.00 100 Anxiety .55 2.64 74
Strenuous exercise .01 3.00 100 Irritability .53 2.59 81
Depression .45 2.49 65
Difficulty relaxing .44 2.39 73
Table II Physical disability rank ordered by preva- Difficulty sleeping .36 2.58 98
lence (P) Boredom .26 2.62 88
Loneliness .23 2.26 78
ITEM P MIS [DW%] Difficulty coping with
stress .17 2.19 75
Disturbed sense of well-
Rising from sitting .70 2.32 99 being .14 2.62 54
Descending stairs .69 2.60 94 Poor self-control .09 2.11 78
Ascending stairs .69 2.54 89
Standing .57 2.64 96 trial. The response rate amongst those invit-
Walking on flat sur-
faces .56 2.40 96 e d t o p a r t i c i p a t e w a s 97~
Getting in/out of car .56 2.26 91 F o l l o w i n g c o m p l e t i o n o f 100 i n t e r v i e w s ,
Bending to floor .55 2.51 95 the data were summarized to provide the fol-
Going shopping .49 2.40 94 l o w i n g v a l u e s : 1) P r e v a l e n c e o f e a c h t y p e o f
Going shopping .49 2.40 94
Putting on socks .46 2.38 96 d i s c o m f o r t o r d i s a b i l i t y (P) ; 2) M e a n i m p o r -
Rising from bed .45 2.37 100 tance score (MIS = the sum of the indivi-
Taking off socks .43 2.37 95 dual importance scores given by n affected
Getting in/out of bath .40 2.30 98 i n d i v i d u a l s d i v i d e d b y n ; a n d 3) T h e p e r c e n -
Lying in bed .39 2.36 95
tage of symptomatic patients experiencing
Heavy domestic duties .36 2.43 71
Light domestic duties .36 2.26 88 d a l l y o r w e e k l y s y m p t o m s (DW070 = H i g h
Sitting .35 2.54 100 Frequency). The individual items were then
Getting on/off toilet .33 2.67 85 ranked within each dimension in order of
Getting on/off bus .30 2.38 66
their prevalence. In this study prevalence
Getting in/out shower .16 2.31 94
Driving a car .16 2.25 75 w a s d e f i n e d as t h e p r o p o r t i o n o f p a t i e n t s i n
Going from bed to the "at risk" population who were con-
chair .15 2.21 85 cerned by ongoing symptomatology on a given
Running on fiat sur-
variable (Tables I-IV). Subsequent analyses
faces .13 2.42 75
e x a m i n e d t h e e f f e c t s o f age, sex a n d d i s e a s e
234 N. Bellamy, W.W. Buchanan

duration on the key symptoms within each complained of pain while bending from the
dimension (Tables V and VI, Fig. 1-3). waist and a second of pain during strenuous
exercise. The prevalence of different sources
RESULTS of pain varied from 56 to 77%, walking
being the most frequent cause of pain.
Sixty-three female and 37 male patients Amongst the five principle items, pain on
were interviewed. The mean age was 61.07 sitting or lying was the least frequent (56O7o).
years (range = 27-93) and mean disease du- Mean importance (MI) scores for pain
ration (i.e. symptomatic) 10.07 years (range varied from 2.509 to 2.627, and pain occur-
= 0.25 - 51). Eleven patients had hip disease red with high frequency, being present at
alone, 57 knee disease alone, and in 32 the least daily or weekly in 94-96% of affected
disease affected both hip and knee. Eight patients.
patients had previously undergone menisec- Pain prevalance for pain in bed and while
tomy some years earlier but none had been negotiating stairs showed little variation
subject to arthroplastic surgery or osteoto- with age but there was a tendency for pain
my. All patients were symptomatic at the experienced while sitting and while walking
time of assessment. to increase prevalence with advancing age
(Fig. 1). Overall MI scores (Fig. 1) failed to
Pain show any correlation (r=0.00) with age
(Table V) and there was no significant dif-
Pain was disaggregated into pain occur- ference (p=0.88) in MI scores for pain
ring during five types of activity (Table I). between males and females (Table VI).
Only two additional components of pain There was a low level of correlation
were identified by open-ended questions and (r=0.26) between MI scores for pain and
it can therefore be assumed that these five disease duration and there were modest but
principle items adequately represent the di- significant correlations between MI scores
mension of pain. Only one patient for pain and those for stiffness, physical

100%-

80-

o
~ 60- .- . . . . . . . . . -.... \ . . r o= 3 - J

"-,.. 9 .... ..-i

a. 4 0 - W -- -..I. 91499
g

213- 1-

]
0 40~49 50~59 60~69 70~79 over'80 40~49 50~59 60~69 70~79 over 80
Age Age
Fig. l:Prev~enceand mean importancescor~ for ~lectedsourcesofpainasafunction of age (- = W~king,
o = Negofiafingstairs, x = In bed~ mght, 1 = sitting orlying).
 A preliminary evaluation of the dimensionality 235

100%-

80-
.a~--. ~ " q ' . \ ~, ... P

60-

~- ~. e.. ~-~.~r f
~- 4 0 -
e-

20-

40149 50159 70179 over'80 40149 50159 6 0 1 6 9 70~79 over180

60169
Age Age
Fig. 2: Prevalenceand mean importance~oresfor ~lectedphysicaldisab~tiesasafunction ofage(* = Stan~ng,
o = Ascen~ngstairs, x =Walking on aflatsurface, 9 =Desccndingstairs).

100%-

80-
o '-

60- " \ / m 3-
\~,

40- 2-

20- 1-

0
40-=49 501-59 60-=69 70-=79 over'80 40-=49 50-=59 60-=69 70-=79 over'80
Age Age
Fig. 3: P r e v a l e n c e
a n d m e a n i m p o r t a n c e scores for selected sociocmotional disabilities as a function o f age~(L =
Anxiety, x = F r u s t r a t i o n , 9 = Restricted leisure activities).

dysfunction, and key social and emotional morning and 73% after prolonged sitting or
items (Table V). lying at other times. Mean importance
scores were 2.524 and 2.303 respectively.
Stiffness The mean duration of morning stiffness was
14.5 minutes (range 1-120). Stiffness occur-
Forty-seven percent of patients complain- red on a daily basis in almost all affected
ed of joint stiffness after wakening in the individuals.
236 N. Bellamy, W.W. Buchanan

Table V Correlation matrix for mean importance scores for pain, stiffness, physical, social and emotional
function, age and disease duration

Pain Stiffness Physical Social Emotional function Age Disease

function function Anxiety Frus- Irritability Duration
tration
Pain 1.00
Stiffness 0.21 1.00
Physical 0.74 0.22 1.00
Social 0.39 0.23 0.44 1.00
Anxiety 0.44 0.23 0.52 0.49 1.00
Frustration 0.46 0.23 0.50 0.41 0.59 1.00
Irritability 0.40 0.16 0.40 0.21 0.57 0.55 1.00
Age 0.00 -0.03 0.06 -0.14 -0.14 -0.13 -0.20 1.00
Disease duration 0.26 -0.01 0.08 -0.03 0.14 -0.01 -0.02 0.14 1.00

Table VI Comparison o f the mean importance scores given by symptomatic male versus female patients for each
dimension

Pain Stiffness Physical Social Emotional

Male 2.56 2.28 2.16 2.38 2.44

Female 2.52 2.32 2.38 2.23 2.35

p value 0.88(NS) 0.84(NS) 0.22(NS) 0.43(NS) 0.5 t(NS)

(2-tailed)

N o significant correlation was n o t e d b y o p e n - e n d e d questions a n d t h e r e f o r e these

between M I scores for stiffness a n d either 22 questions are considered as a d e q u a t e l y re-
age (r = 4).03) or disease d u r a t i o n (r = -0.01) presenting this dimension. T h e p r e v a l e n c e o f
(Table V). Low levels o f correlation were de- physical disability v a r i e d f r o m 13 to 7 0 % .
tected between stiffness scores a n d those o f Negotiating stairs, rising f r o m the seated p o -
pain, physical function, restricted leisure ac- sition, standing, bending, w a l k i n g a n d get-
tivity, anxiety a n d frustration (Table V). ting in a n d o u t o f a car were the most fre-
A l t h o u g h m o r n i n g stiffness was m o r e p r o - quent forms o f disability. It m a y be interest-
longed in females ( m e a n = 1 3 . 5 5 minutes) ing to note t h a t r u n n i n g o n a flat surface,
than males (mean = 7.42 minutes) this differ- transferring f r o m bed to chair, driving a car
ence was not statistically significant a n d getting in and o u t o f the s h o w e r were
( p = 0 . 2 4 ) . Similarly, no significant differ- infrequent causes o f disability. M e a n i m p o r -
ences ( p = 0 . 8 4 ) were detected in M I scores tance scores varied f r o m 2.214 to 2.667.
for stiffness between males a n d females Physical disability o c c u r r e d with high fre-
(Table V). quency, being present d a i l y o r weekly in 80
to 100% o f affected patients in t h e m a j o r i t y
Physical Dysfunction o f instances.
T h e prevalence o f the f o u r p r i n c i p a l
Physical dysfunction was disaggregated forms o f physical d y s f u n c t i o n d i s p l a y e d a
into disability occurring during 22 types o f tendency t o increase with a d v a n c i n g age
activity (Table II). N o a d d i t i o n a l c o m p o - (Fig. 2). I n c o n t r a s t M I scores f o r these f o u r
nents o f physical disability were identified items showed little v a r i a t i o n with age (Fig. 2)
 A preliminary evaluation of the dimensionality 237

and MI scores for the entire dimension were mension. The prevalence of emotional dys-
poorly correlated with both age (r=0.06) function ranged from 9 to 56% : anxiety, ir-
and disease duration (r = 0.08) (Table V). MI ritability, and frustration being most com-
scores for physical dysfunction were highly mon. Difficulty coping with stress (17%),
correlated with those for pain (r=0.74), disturbed sense of well-being (14%) and
moderately correlated with those for the key poor self-control (9%) were infrequent
social and emotional items and poorly corre- problems. Mean importance scores varied
lated with that for stiffness (Table V). No from 2.111 to 2.636. With the exception of a
significant difference (p = 0.22) was detected disturbed sense of well-being and de-
in MI scores for physical dysfunction pression, emotional dysfunction occurred
between males and females (Table VI). with high frequency in the majority of af-
fected individuals (73-98%).
Social Dysfunction The prevalence of the key emotional disa-
bilities showed no consistent relationship to
Social function was disaggregated into age (Fig. 3). MI scores for these items
showed little variation (Fig. 3) and were
seven component items which essentially
captured the dimension (Table III). With the poorly correlated with both age (r=-0.13,
-0.14, -0.20) and disease duration (r= 0.14,
exception of restricted leisure activities
(54%) social dysfunction was relatively un- -0.01, -0.01) (Table V). In contrast modest
correlations were noted between key emo-
common (3 to 27%). In spite of this infre-
tional items and both pain and physical
quency, the mean importance scores
amongst affected individuals range from function and low levels of correlation were
demonstrated with stiffness (Table V). No
2.154 to 2.667, and were comparable with
scores for discomfort and disability on other significant differences (p=0.51) were noted
dimensions. With the exception of dancing, in MI scores for emotional items between
social dysfunction occurred with high fre- males and females (Table VI).
quency in affected individuals (74-91%).
The prevalence of the key social disability CONCLUSION
(restricted leisure activities) showed no con-
stant relationship to age (Fig. 3). MI scores The assessment of pain (34) and disability
for this item showed little variation (Fig. 3) (35) is a complex process which may be af-
and were poorly correlated with age fected by a multiplicity of interacting biolog-
(r=-0.14) and disease duration (r=-0.03) ical and environmental factors (36). Not
(Table V). In contrast modest correltions only is there significant day-to-day variabili-
were noted between social dysfunction and ty in an individual's pain sensitivity and phy-
pain, stiffness, physical and emotional dys- sical performance but there is also often sig-
function (Table V). No significant differ- nificant diurnal or circadian variation
ences (p = 0.43) were noted in MI scores for (37,38). Furthermore, patients with chronic
social items between males and females musculoskeletal disease frequently show a
(Table VI). high and unpredictable degree of individual
variability in their response to therapeutic
interventions (39). Thus, in attempting to
Emotional Dysfunction measure the dimensionality of discomfort
and disability in osteoarthritis, the clinical
Emotional function was disaggregated methodologist must address issues which re-
into 11 component items (Table IV). Since late to the reliability, validity and respon-
no additional items were added, these items siveness of the measuring instrument (4). A
are considered representative of this di- variety of validated scales are currently
 238 N. BeUamy, W.W. Buchanan

available to measure qualitative and quanti- might at first be thought. Thus, amongst the
tative aspects of pain (40) and various forms patients who reported no difficulty with a
of physical, social and emotional disability given activity some were able to perform the
(41). The majority of such scales are not activity without difficulty and other avoided
based on musculoskeletal populations and it and therefore did not encounter any actual
of those which are, few have been designed disability. The question regarding transfer
to assess multidimensional symptoms in os- from bed to chair proved to be a poor
teoarthritic patients (43). question since the majority o f ambulant
This evaluation was specifically based on patients did not require undertaking this ac-
100 patients who could fulfill musculoskele- tivity. These different conditions clearly dis-
tal criteria for entry into a clinical trial of tort the true prevalence of various forms of
nonsteroidal anti-inflammatory drug thera- physical disability, nevertheless, the study
py. It is evident from the data that the ma- provides a reasonable estimate o f the pro-
jority of patients surveyed experienced some portion of patients who were concerned with
form of discomfort or disability on each of any ongoing functional restriction. Since
the five dimensions. Pain was the most com- this survey was conducted for the purpose of
mon symptom, particularly while walking, identifying the dimensionality o f pain and
on negotiating stairs or in bed at night. Stat- disability and not as an epidemiologic survey
ic pain whereas rated somewhat higher in to determine the exact prevalence of each
importance nevertheless occurred less often. item, the principal objective was not com-
This observation is in keeping with the fact promised by this nuance. Restricted leisure
that pain at rest occurs with more severe dis- activities were the only frequent source of
ease and is usually preceded by pain with social disability. The relative infrequency of
those activities which place joints under other sources of social disability, clearly re-
greater mechanical stress. It is commonly flect not only avoidance o f these activities
taught that joint stiffness is mild and of due to disease but also the restriction in
short duration in osteoarthritis (33). Our ob- social activity which attends the process of
servations are consistent with this doctrine, ageing. Thus ageing osteoarthritics are more
at least in respect of duration, although a likely to be widowed, geographically dis-
minority of individuals had prolonged placed from their offspring or financially
morning stiffness. The mean importance constrained.
score indicates that while often reported as Anxiety, irritability and frustration are
"mild", for the affected individuals it is common emotional responses even in heal-
nevertheless an important source of discom- thy individuals. It is not surprising therefore
fort. Difficulty negotiating stairs, standing that patients with osteoarthritis experience
up from a sitting position, standing still, similar, albeit more intense, symptoms in
bending, walking and getting in and out of this area. It is o f interest that less than half
the car occurred in the majority of patients. the patients complained of depression and
Theselocomotor disabilities were more com- that while 67% of patients experienced pain
mon than those associated with less dynamic in bed only 36% had difficulty sleeping.
activities, e.g. putting on and taking o f f Furthermore, in spite o f nocturnal and diur-
socks, lying in bed and sitting. However, nal pain, only 17% had difficulty coping
there is one qualitative aspect of t h e e data with stress and only 14% a disturbed sense
which should be noted. While tasks such as of well-being. These data can likely be
getting on and o f f the toilet cannot be explained by the recurrent observation in
avoided, others, such as heavy domestic clinical practice that patients with chronic
duties and driving a car are generally avoid- disease accommodate to their illness. Thus,
able and thus assume less importance than just as the expectation o f cure or successful
 A preliminary evaluation of the dimensionality 239

treatment changes with time, so does the only of the pain and of physical disability
perception of pain. In addition patients find produced by this disorder but also recog-
ways o f minimizing the discomfort and disa- nition of its social and emotional conse-
bility which the disease causes. quences. Furthermore, while pain and physi-
We consider the questionnaire used in this cal disability are regarded as the most im-
survey as having face and content validity portant symptoms of the disease, these data
(44) by virtue of the development strategy indicate that in affected individuals each
employed, i.e. the utilization of both open- symptom is regarded with similar clinical
ended and closed-ended questions. Further- importance. Although certain physical and
more, the questionnaire was pre-tested in 15 social activities can be avoided these data
patients with osteoarthritis and no difficulty also suggest that for those disabled indivi-
was encountered in patient comprehension duals who are still able to attempt a given
of the terminology used. Nevertheless, activity, the importance of being able to per-
it might be wondered whether the form the activity is similar to that of being
questionnaire in fact probed the severity able to perform other activities in the same
rather than the clinical importance of the and different dimensions.
patients symptoms. We do not believe this to At the present time no standard method
be the case since no significant correlation exists for evaluating patients with osteoar-
was detected between either age and scores thrifts either in clinical practice or in clinical
on any of the five dimensions or between trials. Given the nature of the disease and its
disease duration and scores on any of the many differences from rheumatoid arthritis,
five dimensions. If, in fact, the we believe that it is timely to attempt the
questionnaire had probed severity of symp- development of a multidimensional outcome
toms then given the insidiously progressive measure for use in patients with osteoarthri-
nature o f the disease a time-dependent in- tis of the hip and knee. To date we have
crease in symptomatology would have been identified the dimensionality and clinical im-
expected and a moderate level of correlation portance of a variety of pain and function
detected. As anticipated the prevalence o f items in a group of potential nonsteroidal
pain and physical disability did in fact in- anti-inflammatory drug trial subjects. Fur-
crease as a function of age although social ther work is required to assess the reliability,
and emotional forms of disability showed a construct validity and responsiveness of each
rather inconsistent relationship. The rela- item and to address the issues of scaling, ag-
tively low level of correlation between age gregation within- and across-dimensions and
and disease duration is not entirely unex- to establish the preferred method for statisti-
pected given the documented plateau-form cal analysis. The value of using multiple
age-prevalence profile of certain types of items on several dimensions versus one or a
osteoarthritis (45) and the highly variable in- few items on a restricted number of di-
terval between the age of onset of disease mensions, in discriminating between an ac-
and the age of onset of symptoms (46). tive drug and a placebo or, between two ac-
Finally, it is only to be expected in a disease tive drugs cannot be assessed from the cur-
in which pain leads to disability that pain rent data but is the subject of ongoing re-
scores and disability scores would be moder- search. Nevertheless, the items which have
ately correlated. been identified serve as a useful battery of
This study highlights the multidimension- questions having both face and content val-
ality of discomfort and disability in patients idity which can be used to evaluate indivi-
with osteoarthritis of the hip and/or knee. dual patients or groups of patients with os-
Health care providers and clinical investiga- teoarthritis for descriptive purposes.
tors, therefore, require the assessment not
240 N. Bellamy, W . W . Buchanan

REFERENCES

1. Baum, J. A review of the psychological aspects of 18. Lansbury, J. A method for summation of the syste-
rheumatic disease. Semin Arthritis Rheum. 1982, mic indices of rheumatoid activity. Am J Med Sci.
11,352-361. 1956, 232, 300.
2. Cumey, H.L.F. Osteoarthritis of the hip joint and 19. Lansbury, J. Report of a three-year study on the
sexual activity. Ann Rheum Dis. 1970, 29, 488-493. systemic and articular indexes in rheumatoid arthri-
3. Lunghi, M.E., Miller, P.M., McQnillan, W.M. tis - Theoretical and clinical considerations. Arthri-
Psychological factors in osteoarthrosis of the hip. J tis Rheum. 1958, 1, 505-522.
Psychosom Res, 1978, 22, 55-63. 20. Lansbury, J., Baler, H.N., McCracken, S. Statisti-
4. Bellamy, N., Buchanan, W.W. Outcome measure- cal study of variation in systemic and articular in-
ment in osteoarthritis clinical trials: The case for dices. Arthritis Rheum. 1962, 5, 445-456.
standardisation. Clin Rheum~itol, 1984, 3, 293-303. 21. Lansbury, J., Mueller, E.E. A numerical method
5. Convery, F.R., Minteer, M.A., Amiel, D., Con- for summing up total deformity, Am J Med Sci.
nett, K.L. Polyarticular disability : A functional as- 1958, 235, 154.
sessment. Arch Phys Med Rehabil, 1977, 58, 494- 22. Lee, P., Jasani, M.K. Dick, W.C., Buchanan,
499. W.W. Evaluation of a functional index in rheu-
6. Deniston, O.L., Jette, A. A functional status as- matoid arthritis. Scand J Rheurnatol. 1973, 2, 71-
sessment instrument : Validation in an elderly pop- 77,
niation. Health Serv Res. 1980, 15, 21-34. 23. Lee, P., Webb, J., Anderson, J., Buchanan, W.W.
7. Doyle, D.V., Dieppe, P.A., Scott, J., Huskisson, Method of assessing therapeutic potential of anti-
E.C. An articular index for the assessment of os- inflammatory anti-rheumatic drugs in rheumatoid
teoarthritis. Ann Rheum Dis. 1981, 40, 75-78. arthritis. Br Med J 1973, 2, 685-688.
8. Eberl, D.R., Fasching, V., Rahlafs, V., Schieyer, 24. Lequesne, M. European guidelines for clinical trials
I., Wolfe, R. Repeatability and objectivity of va- of new antirheumatic drugs. Eular Bulletin. 1980, 9
rious measurements in rheumatioid arthritis. At- (Suppl 6), 171-175.
thrifts Rheum. 1976, 19, 1278-1286. 25. Meenan, R.F., Gertman, P.M., Mason, J.H.
9. Empire Rheumatism Council. Multicentre con- Measuring health status in arthritics. The arthritis
trolled trial comparing cortisone acetate and acetyl impact measurement scales. Arthritis Rheum. 1980,
salicylic acid in the long-term treatment of 23, 146-152.
rheumatoid arthritis. Ann Rheum Dis. 1955, 14, 26. Ritchie, D.M., Boyle, J.A., MeInnes, J.M., Jasani,
353-370. M.K., Dalakos, T.G., Grieveson, P., Buchanan,
10. Fries, J.F., Spitz, P., Kralnes, R.G., Homan, H.R. W.W. Clinical studies with and articular index for
Measurement of patient outcome in arthritis. Ar- the assessment of joint tenderness in patients with
thritis Rheum. 1980, 23, 137-145. rheumatoid arthritis. 1968, Q J Med. 147, 393-406.
11. Jette, A.M. Functional capacity evaluation: An 27. Smythe, H.A., Helewa, A., Goldsmith, C.H. Inde-
empirical approach. Arch Phys Med Rehabil. 1980, pendent assessor and pooled index as techniques
61, 85-89. for measuring treatment effects in rheumatoid ar-
12. Jette, A.M. Functional status index: Reliability of thritis. J Rheumatol, 1977, 4, 144-152.
a chronic disease evaluative instrument. Arch Phys 28. Steinbrocker, O., Traeger, C.H., Batterman, R.C.
Meal Rehabil. 1980, 62, 395-401. Therapeutic criteria in rheumatoid arthritis.
13. Jette, A.M., Deniston, O.L. Interobserver reliabili- JAMA, 1949, 140, 659-662.
ty of a functional status instrument. J Chronic Dis. 29. Taylor D. A table for the degree of involvement in
1978, 31,573-580. chronic arthritis. Can Med Assoc J. 1977, 36, 608-
14. Katz, S., Akpom, C.A. A measure of primary 610.
sociohiological functions. Int J Health Serv. 1976, 30. The Co-operating Clinics Committee of the Ameri-
6, 493-507. can Rheumatism Association. A seven-day variabil-
15. Katz, S., Downs, T.D., Cash, H.R., Grotz, R.C. ity study of 449 patients with peripheral
Progress in development of the index of ADL. Ge- rheumatoid arthritis. Arthritis Rheum. 1965, 8,
rontologist. 1970, 1, 20-30. 302-335.
16. Katz, S., Ford, A.B., Moskiwitz, R.W., Jackson, 31. Tugwell, P.X., Bombardier, C., Chambers, L.W.
B.A., Jaff, M.W. Studies of illness in the aged. MACTAR Functional assessment questionnaire.
The index of ADL: A standardised measure of Monograph - McMaster University - Department of
biological and psyehosomal function. JAMA. Clinical Epidemiology and Biostatistics, 1981.
1963, 914-919. 32. Chambers, L.W. McMaster Health Index
17. Keitel, W., Hoffman, H., Weber, G. Ermitflung Questionnaire (MHIQ). Monograph - McMaster
der prozentualen Funktionsminderung der Gelenke University - Department of Clinical Epidemiology
durch einen Bewengungsfunkionstest in der and Biostatistics. 1980.
Rheumatologie. Dtsch Gesundheitsw. 1971, 26, 33. Moskowitz, R.W. Clinical and laboratory findings
1901-1903. in osteoarthritis. In: Arthritis and Allied Con-
 A preliminary evaluation of the dimensionality 241

ditions. Ed. McCarty, D.J. Philadelphia, Lea & Fe- Pain Measurement and Assessment. Ed. Melzack,
biger, 1979, 1161-1180. R. New York, Raven Press, 1985, 41-48.
34. Melzack, R. Concepts of pain measurement. In: 41. Brook, R.H., Ware, J.E., Davies-Avery, A., Ste-
Pain Measurement and Assessment. Ed. Melzack, wart, A.L., Donald, C.A., Rogers, W.K., Wil-
R. New York, Raven Press, 1985, 1-5. liams, K.N., Johnston, S.A. Overview of adult
35. International classification of impairments, disabil- health status measures fielded in Rand's health in-
ities and handicaps. Geneva, World Health Organi- surance study. Med Care, 1979, 17 (Suppl 7), 1-
zation. 1980, 7-174. 131.
36. Kremer, E.F., Block, A., Atkinson, J.H. Assess- 42. Deyo, R.A. Measuring functional outcomes in
ment of pain behaviour, factors that distort self- therapeutic trials for chronic disease. Controlled
report. In : Pain Measurement and Assessment. Ed. Clinical Trials. 1984, 5, 223-240.
Melzack, R. New York, Raven Press, 1985, 165- 43. Bellamy, N. The clinical evaluation of osteoarthd-
171. tis in the elderly. In : Clinics in Rheumatic Diseases.
37. Davis, G.C., Buchsbaum, M.S., Burney, W.E. Na- Ed. Kean, W.F. Eastbourne, W.B. Saunders. 1986,
loxone decreases diurnal variations in pain sensitiv- (In Press).
ity and somatoseasory-evoked potentials. Life Sci. 44. Carmines, E.G., Zeller, R.A. Reliability and validi-
1978, 23, 1449-1459. ty assessment. Beverley Hills, Sages Publications,
38. Simpson, H.W., Beltamy, N., Bohlen, J., Halberg, 1979, 5-71.
F. Double-blind field trial of a potential chronobio- 45. Forman, M.D., Malamet, R., Kaplan, E. A survey
tic (Quiadon). Int J Chronobiol. 1973, 1,287-311. of osteoarthritis of the knee in the elderly. J
39. Bellamy, N. Between-subjeet variability in response Rheumatol. 1983, 10, 282-287.
to antirheumatic drugs: Implications for clinical 46. Bellamy, N. Osteoarthritis - An evaiuative index
trials. Agents Actions. 1985, (In Press). for clinical trials. M.Sc. Thesis, Hamilton, Canada,
40. Melzack, R. The McGill pain questionnaire. In: McMaster University. 100-104.