Abstract

Objective. To examine further the usefulness of a 30‐item disease‐specific quality of life (QoL) questionnaire in patients with rheumatoid arthritis (RA).

Methods. The Rheumatoid Arthritis Quality of Life (RAQoL) questionnaire was applied to two groups consisting of 210 and 300 patients with RA, one group with increasing difficulty in performing activities of daily living and one group with stable disease. The associations between the RAQoL and measures of utility, QoL, functional status and disease activity were evaluated. Factor analysis was carried out to investigate if one or more QoL dimensions could be distinguished within this questionnaire.

Results. Similar results regarding the association between the RAQoL and different sets of outcome measures were found in the two groups of patients. Regression analysis showed that about 75% of the variance of the RAQoL could be explained with variables of QoL, functional status and disease activity. Physical contact could be distinguished as a separate dimension within the RAQoL, in addition to the dimensions mobility/energy, self‐care and mood/emotion.

Conclusion. The RAQoL is a valid instrument for measuring QoL in different populations of patients with RA. Physical contact, a dimension that is not covered by other common instruments in RA, could be distinguished as a separate dimension within the questionnaire.

Rheumatoid arthritis (RA) is a chronic disease, which has a detrimental effect on many areas of life, including physical, psychological and social functioning. The ultimate aim of any therapeutic intervention in patients with RA is to achieve a better overall quality of life (QoL). Although the importance of QoL is broadly acknowledged, there is no consensus about how it should be measured.

Current instruments which are in use to measure QoL in patients with RA are either generic or specific to the disease. Generic instruments such as the Nottingham Health Profile (NHP) [1], the Sickness Impact Profile (SIP) [2] and the Short Form Health Survey‐36 (SF‐36) [3] are designed to compare health states among different disease populations. Because these instruments necessarily contain items that have to be applicable to different disease populations, it is conceivable that not enough attention is paid to issues that can be of special importance for patients with RA. A generic measure used in patients with RA may therefore have a reduced validity and sensitivity to clinically relevant changes [4].

Disease‐specific instruments are designed to measure QoL in a certain disease population. In RA, the most frequently utilized disease‐specific QoL instrument is the Arthritis Impact Measurement Scale (AIMS), which was revised (AIMS2) and adapted for use in different countries [511]. However, it was found that some patients find the AIMS2 too long and only 58% of patients believe that the available response options usually allow them to describe their situation [6]. Moreover, not all of the scales of this questionnaire seem sensitive to measure changes in health status [911].

Recently, Whalley et al. [4] developed the Rheumatoid Arthritis Quality of Life (RAQoL) questionnaire with the aim of developing a valid and reliable measure of QoL specific for patients with RA. Furthermore, the instrument should be simple to administer and complete and acceptable to the respondents. To satisfy these requirements, this questionnaire was based on the needs‐based model of QoL [12]. This 30‐item questionnaire appears to be valid and practical in use and has potential applications in cost‐effectiveness studies [13]. It has been validated against the NHP and a limited number of measures of disease activity and severity, in parallel in the UK and The Netherlands [13]. Also, the results of its sensitivity to change are promising [14].

Before applying this instrument on a larger scale, however, more information about the value of the RAQoL in relation to other aspects of QoL, such as social and emotional functioning and functional status, and to additional measures of disease activity is needed. Therefore, the aim of the present study was to extend the validation of the RAQoL in different populations of patients with RA.

Patients and methods

Patients

Data were gathered as part of two randomized controlled trials. In the first trial (study group 1), the cost‐effectiveness of three types of care for patients with RA was compared (in‐patient multi‐disciplinary care, multi‐disciplinary day care, and out‐patient care co‐ordinated by the rheumatologist and nurse practitioner). Two hundred and ten patients were recruited between December 1996 and January 1999, at the out‐patient clinic of the rheumatology departments of Leiden University Medical Center and five non‐academic hospitals within the region of Leiden. The inclusion criteria were: RA as defined by the 1987 revised American Rheumatism Association (ARA) criteria [15] and increasing difficulty in activities of daily living which were not solvable by the rheumatologist or a single health professional.

In the second trial (study group 2), the effect of an intensive exercise programme in RA was investigated. The participants were randomized to two groups. One group was allocated to follow an intensive exercise programme twice a week (1.25 h) for 2 yr, whereas the other group was to be treated by a physical therapist on doctor's prescription only. Recruitment took place in the Leiden University Medical Center, the academic hospital of the Free University in Amsterdam and two non‐academic hospitals. Patients with RA were invited to participate in the study by mail. After personalized screening for inclusion criteria, 300 patients were included in June 1997. The inclusion criteria were: age 20–70 yr, RA according to the 1987 revised ARA criteria [15], functional class I, II and III according to the 1991 American College of Rheumatology (ACR) revised criteria for the classification of functional status [16], stable dose of disease‐modifying drug in the past 3 months, able to cycle on a home trainer and no serious comorbidity. For both trials, the medical ethics committees of all participating hospitals approved the study protocols. All patients gave written informed consent.

Methods

In each trial there were two experienced assessors who carried out all assessments and who were blind to the patients’ treatment status. In study group 1 the assessments were carried out at baseline, weeks 6, 12, 26, 52, 78 and 104, and in study group 2 at baseline and every 3 months thereafter during the following 2 yr. For the present validation study of the RAQoL, only the data gathered at baseline were used.

Sociodemographic characteristics and clinical data.

The following data were recorded at baseline: sex, age, disease duration, rheumatoid factor, status of living (living alone or living with a partner) and level of education (categorized as low: up to and including lower technical and vocational training; medium: up to and including secondary technical and vocational training; and high: up to and including higher technical and vocational training and university).

QoL measures.

Two measures of QoL were used, the RAQoL and the RAND. The RAQoL questionnaire consists of 30 items with a yes/no (1/0) response format [13]. The overall score is the sum of the individual item scores, with a lower score indicating better QoL (range 0–30). The RAND 36‐item Health Survey 1.0 (RAND) contains subscales for physical functioning, social functioning, role limitations physical problem, role limitations emotional problem, mental health, vitality, pain and general health perception. Each scale generates a score from 0 to 100, with higher scores indicating better health. The RAND can be converted to two summary scales: the physical and mental component summary scales [3]. The RAND includes the same items as the Medical Outcomes Study Short‐Form (SF36) and although the scoring procedures are somewhat different, the effects of these on final scores are minimal [17]. The RAQoL and the RAND have been validated for use in The Netherlands [13, 18].

Utilities.

Two measures of utility, the time trade‐off (TTO) and the rating scale (RS), were used in study group 1. In study group 2, the Euroqol and the visual analogue scale (VAS) of the Euroqol were used. The TTO is a method in which patients are asked how many years (x) in perfect health they would consider equivalent to their remaining life expectancy (y) in their current health state [19]. The utility is then calculated as x/y. A visual aid is used with this technique. To determine the life expectancy of a patient, the age‐specific life expectancy from the general population was used [20]. The RS is a method in which patients are asked to place two so‐called marker health states and subsequently their own current health state, on a 0–10 cm vertical line anchored by the ‘worst imaginable health state’ (bottom) and perfect health (top) [21]. For the description of the marker health states we used the McMaster dimensions with modifications as described by Bakker et al. [22]. In the Euroqol, utilities are obtained indirectly by means of a health classification system [23]. In this system, healthy members of the community have provided assessments of the value of particular health state descriptions. The Euroqol consists of five domains of three statements each. The domains are: mobility, self‐care, usual activities, pain/discomfort and anxiety/depression. The utilities were calculated by use of the data of Dolan [24]. The VAS of the Euroqol is comparable with the above‐described RS, however no marker states were used.

Functional measures.

Functional ability was measured by: (1) The Dutch version of the Health Assessment Questionnaire (HAQ) [25]. (2) The grip strength test (mean grip strength of the right and left hands) [Accoson vigorimeter (study group 1) and Martin vigorimeter [26] (study group 2)]. In study group 1, the values of grip strength (mmHg) were converted to values in kPa. (3) The walk test (50‐ft walking time) [27].

Measures of disease activity.

Measures of disease activity included: (1) The patient's global assessment of disease activity, pain and morning stiffness, all measured on a VAS (range 0–10 cm). The anchors on the left were no disease activity, no pain and no stiffness, and on the right were worst imaginable disease activity, severe pain and severe stiffness, respectively. (2) The physician's estimation of disease activity measured on a VAS (range 0–10 cm) with no disease activity on the left anchor and worst imaginable disease activity on the right anchor. (3) The number of swollen joints (0=no swelling and 1=swelling) using the 28 joint score [28]. (4) The number of tender joints, including the same joints (0=no pain, 1=painful). (5) The erythrocyte sedimentation rate (ESR). From the patient's global assessment of disease activity (VAS), swollen and tender joint scores and the ESR, the disease activity score (DAS) was calculated using the following formula [29]: 0.555√(tender joints)+0.284√(swollen joints)+0.7 ln(ESR)+0.0142(VAS patient's global assessment of disease activity).

Analysis and statistical methods.

To investigate if there were differences in patient characteristics, measures of QoL, functional status and disease activity between the two study groups, the Mann–Whitney U‐test or χ2 test was used where appropriate. The associations between patient characteristics and the RAQoL were estimated by calculating Pearson's correlation coefficients. The difference in RAQoL scores between male and female patients was measured by the Mann–Whitney U‐test.

The associations between the RAQoL and utilities and measures of QoL (construct validity) and between the RAQoL and measures of function and disease activity (content validity) were evaluated by calculating Pearson's correlation coefficients. The differences between the associated regression coefficients of similar outcome measures performed in both study groups were tested for statistical significance by regression analysis. To investigate a possible influence of patient characteristics on the associations, all procedures were repeated by controlling for these characteristics. The ability of the RAQoL to discriminate between worse and better health state outcomes was studied by dividing the following measures into four categories: the summary scales of the RAND (physical and mental), the HAQ, the DAS and the VAS pain. These outcome measures were divided into four parts according to quartiles, except for the DAS for which the categories were as follows: 0–2.0, 2.01–4.0, 4.01–6.0 and 6.01–higher. To determine the contribution of several sets of variables to the explained variance of the RAQoL, a multiple regression analysis was performed with the following sets of variables: patient characteristics, measures of utility, QoL, functional status and disease activity. The dependent variable was the RAQoL.

The goodness of fit of the regression model was assessed by inspection of the distribution of the residuals. The normality of this distribution and the homoscedasticity were checked visually. Multicollinearity was checked by inspecting variance inflation factors. First, we studied the association with the RAQoL for the five sets of variables separately to reduce the problem of multiple testing (method enter). Second, within each set of variables, the contribution of individual variables was calculated (stepwise forward). Subsequently, the variables out of the five different sets of variables that contributed significantly to the explained variance (P<0.001) were entered into one final model (stepwise forward).

A principal components analysis was carried out to investigate the possibility of distinguishing one or more QoL dimension within the 30‐item RAQoL questionnaire. Subscales were constructed on the basis of the principal components analysis. Only factors with an eigenvalue >1.0 were considered. We rotated the principal components to an orthogonal simple structure using the varimax method, to facilitate interpretation [30]. Cronbach's alpha was used to estimate the internal consistency, a measure of the reliability of these subscales [31].

Results

Patient characteristics

The sociodemographic and clinical characteristics of the patients are shown in Table 1. The patients in study group 1 were significantly older, had a shorter disease duration, a lower percentage of positive rheumatoid factor, and had a lower education level than the patients in study group 2 (P<0.001). Table 2 shows the mean or median values of measures of QoL, utility, functional status, and disease activity of the patients in both study groups. The differences between the two groups were statistically significant for all measures, with patients in group 2 overall having a better health status than patients in group 1 (P<0.001).

Table 1.

Characteristics of 510 patients with RA


 
Study group 1 (n=210)
 
Study group 2 (n=300)
 
Median (range)   
Age (yr) 59 (23–85) 54 (25–70)* 
Duration of disease (yr) 2.2 6.0 
Median (range) (0.0–46.6) (1.0–43.0)* 
No. of patients (%)   
Female 158 (75) 237 (79) 
With positive rheumatoid factor 144 (69) 264 (88)* 
Living alone 42 (20) 69 (23) 
Education levela   
   Low 113 (54) 106 (35)* 
   Medium 72 (34) 108 (36) 
   High 25 (12) 80 (27)* 
   Unknown – 6 (2) 

 
Study group 1 (n=210)
 
Study group 2 (n=300)
 
Median (range)   
Age (yr) 59 (23–85) 54 (25–70)* 
Duration of disease (yr) 2.2 6.0 
Median (range) (0.0–46.6) (1.0–43.0)* 
No. of patients (%)   
Female 158 (75) 237 (79) 
With positive rheumatoid factor 144 (69) 264 (88)* 
Living alone 42 (20) 69 (23) 
Education levela   
   Low 113 (54) 106 (35)* 
   Medium 72 (34) 108 (36) 
   High 25 (12) 80 (27)* 
   Unknown – 6 (2) 

aLow, up to and including lower technical and vocational training; medium, up to and including secondary technical and vocational training; high, up to and including higher technical and vocational training and university.

*P<0.001, study group 1 vs 2 (Mann–Whitney U‐test or χ2 test).

Table 2.

Measures of QoL, utility, functional status and disease activity in 510 patients with RA. Results are presented as means (standard deviation)


 
Study group 1 (n=210)
 
Study group 2 (n=300)
 
QoL   
   RAQoL (0–30) 16.4 (7.0) 11.3 (6.9) 
   RAND (0–1.0)   
      Physical functioning 39.0 (24) 56.7 (21) 
      Social functioning 59.9 (28) 76.0 (23) 
      Role limitations physical 16.1 (29) 46.7 (42) 
         functioning   
      Role limitations emotional 49.3 (45) 74.2 (39) 
         problems   
      Mental health 62.6 (19) 73.7 (17) 
      Vitality 43.9 (19) 56.1 (18) 
      Pain 39.8 (21) 58.0 (20) 
      General health perception 39.9 (18) 49.9 (19) 
      Summary scale mental 57.3 (25) 74.6 (22) 
      Summary scale physical 31.7 (20) 53.8 (24) 
Utility   
   TTO (0–1) 0.71 (0.3)    – 
   RS (0–100)/100 0.57 (0.2)    – 
   Euroqol (−0.5–1.0)a  – 0.69 (−0.2–1.0) 
   VAS health status (Euroqol)  – 0.66 (0.2) 
      (0–100)/100   
Functional status   
   HAQ (0–3) 1.40 (0.7) 0.75 (0.6) 
   Grip strength (kPa) 21.7 (10) 35.6 (21) 
   50‐ft walking time (s)a 11.0 (5–45) 12.1 (8–39) 
Disease activity   
   Disease activity VAS 39.8 (20) 25.7 (20) 
      (investigator) (0–100)   
   Pain VAS (0–100) 50.9 (24) 36.1 (25) 
   Morning stiffness VAS (0–100) 53.7 (28) 32.4 (29) 
   DAS 5.62 (1.2) 4.28 (1.3) 

 
Study group 1 (n=210)
 
Study group 2 (n=300)
 
QoL   
   RAQoL (0–30) 16.4 (7.0) 11.3 (6.9) 
   RAND (0–1.0)   
      Physical functioning 39.0 (24) 56.7 (21) 
      Social functioning 59.9 (28) 76.0 (23) 
      Role limitations physical 16.1 (29) 46.7 (42) 
         functioning   
      Role limitations emotional 49.3 (45) 74.2 (39) 
         problems   
      Mental health 62.6 (19) 73.7 (17) 
      Vitality 43.9 (19) 56.1 (18) 
      Pain 39.8 (21) 58.0 (20) 
      General health perception 39.9 (18) 49.9 (19) 
      Summary scale mental 57.3 (25) 74.6 (22) 
      Summary scale physical 31.7 (20) 53.8 (24) 
Utility   
   TTO (0–1) 0.71 (0.3)    – 
   RS (0–100)/100 0.57 (0.2)    – 
   Euroqol (−0.5–1.0)a  – 0.69 (−0.2–1.0) 
   VAS health status (Euroqol)  – 0.66 (0.2) 
      (0–100)/100   
Functional status   
   HAQ (0–3) 1.40 (0.7) 0.75 (0.6) 
   Grip strength (kPa) 21.7 (10) 35.6 (21) 
   50‐ft walking time (s)a 11.0 (5–45) 12.1 (8–39) 
Disease activity   
   Disease activity VAS 39.8 (20) 25.7 (20) 
      (investigator) (0–100)   
   Pain VAS (0–100) 50.9 (24) 36.1 (25) 
   Morning stiffness VAS (0–100) 53.7 (28) 32.4 (29) 
   DAS 5.62 (1.2) 4.28 (1.3) 

aMedian values (range).

For all outcome measures the difference between study groups 1 and 2 was statistically significant, P<0.001 (Mann–Whitney U‐test).

Validity

Associations between the RAQoL and patient characteristics.

In study group 1, being older, a longer disease duration and female sex were significantly associated with higher RAQoL scores (worse QoL) (P<0.05) (Table 3). In study group 2, female sex was also associated with higher RAQoL scores. However, in contrast with study group 1, being older was associated with lower RAQoL scores (P<0.05) (Table 3). The mean (standard deviation) RAQoL scores in males were 13.3 (7.3) and 8.5 (6.5) and in females were 17.4 (6.6) and 12.1 (6.8) in study groups 1 and 2, respectively. The difference in RAQoL scores between male and female patients was statistically significant in both groups (P<0.001).

Table 3.

Pearson's correlation coefficients between the RAQoL and patient characteristics and health status outcomes in 510 patients with RA


 
Study group 1
 
Study group 2
 
Patient characteristics   
   Age 0.23 −0.14d 
   Duration of disease 0.15 −0.01c 
   Sexa −0.25 −0.21 
   Rheumatoid factor 0.03c −0.01c 
   Status of livingb −0.06c −0.04c 
   Education level −0.03c −0.12c 
Utilities   
   TTO −0.37  – 
   RS −0.63  – 
   Euroqol – −0.62 
   VAS Euroqol – −0.51 
QoL   
   RAND (0–100)   
      Physical functioning −0.68 −0.69 
      Social functioning −0.65 −0.66 
      Role limitations physical functioning −0.39 −0.59 
      Role limitations emotional problems −0.43 −0.45 
      Mental health −0.56 −0.53 
      Vitality −0.57 −0.62 
      Pain −0.60 −0.62 
      General health perception −0.53 −0.56 
      Summary scale mental −0.64 −0.61 
      Summary scale physical −0.68 −0.74 
Functional status   
   HAQ 0.75  0.72 
   Grip strength −0.55 −0.31d 
   Walk test 0.42  0.41 
Disease activity   
   Disease activity VAS (investigator) 0.33  0.36 
   Pain VAS 0.52  0.46 
   Morning stiffness VAS 0.34  0.40 
   DAS 0.47  0.36 

 
Study group 1
 
Study group 2
 
Patient characteristics   
   Age 0.23 −0.14d 
   Duration of disease 0.15 −0.01c 
   Sexa −0.25 −0.21 
   Rheumatoid factor 0.03c −0.01c 
   Status of livingb −0.06c −0.04c 
   Education level −0.03c −0.12c 
Utilities   
   TTO −0.37  – 
   RS −0.63  – 
   Euroqol – −0.62 
   VAS Euroqol – −0.51 
QoL   
   RAND (0–100)   
      Physical functioning −0.68 −0.69 
      Social functioning −0.65 −0.66 
      Role limitations physical functioning −0.39 −0.59 
      Role limitations emotional problems −0.43 −0.45 
      Mental health −0.56 −0.53 
      Vitality −0.57 −0.62 
      Pain −0.60 −0.62 
      General health perception −0.53 −0.56 
      Summary scale mental −0.64 −0.61 
      Summary scale physical −0.68 −0.74 
Functional status   
   HAQ 0.75  0.72 
   Grip strength −0.55 −0.31d 
   Walk test 0.42  0.41 
Disease activity   
   Disease activity VAS (investigator) 0.33  0.36 
   Pain VAS 0.52  0.46 
   Morning stiffness VAS 0.34  0.40 
   DAS 0.47  0.36 

a1=female; 2=male.

b1=living alone; 2=living together with one or more persons.

The associations between the RAQoL and outcome measures were all statistically significant (patient characteristics: P<0.05; all other outcome measures: P<0.01).

cNo statistically significant associations were found.

dA statistically significant difference of correlation coefficients was found between study group 1 and study group 2 (P<0.001) (regression analysis).

Construct validity.

Moderate to high associations were found between the RAQoL and measures of utility and QoL (P<0.01) (Table 3). The differences between correlation coefficients of study group 1 and study group 2 were not statistically significant. The correlations indicated that a worse outcome concerning utilities and QoL was associated with higher RAQoL scores.

Content validity.

Moderate to high associations were found between the RAQoL and measures of functional status and disease activity (P<0.01) (Table 3, Fig. 1). In both study groups, all correlation coefficients between the RAQoL and the different outcome measures indicated that a worse outcome concerning functional status and disease activity was associated with higher RAQoL scores. Except for grip strength, there were no differences between correlation coefficients of study groups 1 and 2. Regression analysis showed that patient characteristics did not have an appreciable effect on these correlation coefficients. In study group 1, for example, controlling for age and sex, the partial correlation between the RAQoL and the HAQ changed from 0.75 to 0.72 and from 0.75 to 0.71, respectively. An illustration of the discriminating properties of the RAQoL for four categories of the HAQ, the VAS pain, the DAS and the summary scales of the RAND is shown in Fig. 2. A trend towards higher scores in categories corresponding to worse health states was found for all outcome measures.

Multiple regression analysis (method enter) showed that the set of variables of measures of QoL, functional status, utility and disease activity, and the patient characteristics independently explained 69, 58, 42, 32 and 12%, respectively, of the variance of the RAQoL in study group 1 and 69, 52, 46, 26 and 10%, respectively, in study group 2. After multiple regression analysis (stepwise forward) within each set of variables it appeared that in study group 1 the following variables contributed significantly to the explained variance (P<0.001): sex, age, the RS, the TTO, five of the eight dimensions of the RAND (physical and social functioning, mental and general health, pain), the HAQ, grip strength, VAS pain and the DAS. In study group 2, sex, level of education, age, the Euroqol and the VAS of the Euroqol, all dimensions of the RAND except the dimension ‘role limitation emotional problems’, the HAQ, walk test and the VASs of pain, disease activity (assessor) and morning stiffness. Stepwise forward analysis of these 13 and 17 variables, respectively, into one model resulted in a significant contribution to the explained variance of the RAQoL of seven variables in both groups (P<0.001) (Table 4). The total explained variance of the RAQoL with all these factors was 75.9 and 73.4% in study groups 1 and 2, respectively. The HAQ and some subscales of the RAND 36 (mental, social and physical) were responsible for most of the explained variance.

Fig. 1.

Scatter plot showing the association between RAQoL scores and HAQ scores in the patients of study group 1 (solid diamonds) and study group 2 (open diamonds). Pearson's correlation of the values in the patients of study group 1 (– – – – –) and study group 2 (‐ ‐ ‐ ‐ ‐ ‐) revealed an r of 0.75 and 0.72, respectively.

Fig. 1.

Scatter plot showing the association between RAQoL scores and HAQ scores in the patients of study group 1 (solid diamonds) and study group 2 (open diamonds). Pearson's correlation of the values in the patients of study group 1 (– – – – –) and study group 2 (‐ ‐ ‐ ‐ ‐ ‐) revealed an r of 0.75 and 0.72, respectively.

Fig. 2.

The discriminating properties of the RAQoL for four categories of the HAQ, the VAS pain, the DAS and the summary scales mental health and physical health of the RAND‐36. Means (standard deviation) of RAQoL scores are given.

Fig. 2.

The discriminating properties of the RAQoL for four categories of the HAQ, the VAS pain, the DAS and the summary scales mental health and physical health of the RAND‐36. Means (standard deviation) of RAQoL scores are given.

Table 4.

Stepwise forward regression analyses with RAQoL as the dependent variable and other baseline assessments as independent variables

Independent variables
 
% of explained variance (cumulative)
 
Study group 1  
   HAQ 56.5 
   Mental health (RAND) 68.0 
   Social functioning (RAND) 71.2 
   Physical functioning (RAND) 72.9 
   DAS 74.2 
   General health perception (RAND) 75.2 
   Sex 75.9 
Study group 2  
   HAQ 50.4 
   Vitality (RAND) 65.6 
   Morning stiffness 67.9 
   Mental health (RAND) 69.8 
   Physical functioning (RAND) 71.6 
   General health perception (RAND) 73.0 
   Sex 73.4 
Independent variables
 
% of explained variance (cumulative)
 
Study group 1  
   HAQ 56.5 
   Mental health (RAND) 68.0 
   Social functioning (RAND) 71.2 
   Physical functioning (RAND) 72.9 
   DAS 74.2 
   General health perception (RAND) 75.2 
   Sex 75.9 
Study group 2  
   HAQ 50.4 
   Vitality (RAND) 65.6 
   Morning stiffness 67.9 
   Mental health (RAND) 69.8 
   Physical functioning (RAND) 71.6 
   General health perception (RAND) 73.0 
   Sex 73.4 

Internal consistency and dimensions of the questionnaire.

In both study groups, principal component analysis of the 30 items of the RAQoL resulted in eight factors with an Eigenvalue >1. On the basis of the scree plot and the fact that factors five to eight only contained one or two items with a loading >0.5, we analysed four factors. These four dimensions comprised eight, five, four and three items, respectively, in study group 1, and six, six, four and three items, respectively, in study group 2 (Table 5). Despite some differences in the nature and number of the items loading on the four dimensions between study groups 1 and 2, the dimensions could be labelled as follows: mobility/energy, self‐care, mood/emotion, and physical contact. A reliability analysis showed Cronbach's alpha ranging from 0.67 to 0.79 and 0.45 to 0.79 per dimension in study groups 1 and 2, respectively (Table 5). The total internal consistency of the RAQoL in both study groups was 0.90.

Table 5.

Factor analysis of the 30 items of the RAQoL

 Study group 1
 

 

 
Study group 2
 

 

 
Factor
 
Itemsa
 
Cronbach's alpha
 
% of variance
 
Itemsa
 
Cronbach's alpha
 
% of variance
 
Mobility/energy 3, 4, 6, 7, 10, 14, 17, 20 0.79 27 1, 6, 7, 10, 14, 20 0.79 27 
Self‐care 5, 11, 18, 22, 30 0.79  7 5, 8, 11, 18, 25, 30 0.73  6 
Mood/emotion 9, 16, 19, 28 0.67  5 9, 12, 16, 24 0.74  5 
Physical contact 2, 15, 29 0.68  5 2, 15, 26 0.45  4 
All items 30 0.90 – 30 0.90 – 
 Study group 1
 

 

 
Study group 2
 

 

 
Factor
 
Itemsa
 
Cronbach's alpha
 
% of variance
 
Itemsa
 
Cronbach's alpha
 
% of variance
 
Mobility/energy 3, 4, 6, 7, 10, 14, 17, 20 0.79 27 1, 6, 7, 10, 14, 20 0.79 27 
Self‐care 5, 11, 18, 22, 30 0.79  7 5, 8, 11, 18, 25, 30 0.73  6 
Mood/emotion 9, 16, 19, 28 0.67  5 9, 12, 16, 24 0.74  5 
Physical contact 2, 15, 29 0.68  5 2, 15, 26 0.45  4 
All items 30 0.90 – 30 0.90 – 

aFor a description of the items of the questionnaire, see [13].

Discussion

This study was designed to extend the validation of the RAQoL, a RA‐specific QoL questionnaire. We demonstrated that the RAQoL is a valid instrument for measuring QoL in different populations of patients with RA. The RAQoL has good discriminating properties and we showed that physical contact, a dimension of QoL that is not represented in the outcome measures currently in use in patients with RA, is represented within the questionnaire.

Moderate to high correlations were found between the RAQoL and measures of utility, QoL, functional status and disease activity. A multiple regression analysis showed that about 75% of the variance of the RAQoL could be explained by the HAQ, some scales of the RAND 36 (mental, social and physical) and the DAS. It appeared that the HAQ and scales of the RAND contributed most to the explained variance. Similar results were found in the two groups of patients who differed with respect to several patient characteristics and measures of QoL, disease severity and functional limitations. These results are in line with the findings of the study of de Jong et al. [13] in which the RAQoL was validated with sections of the NHP and measures of disease activity. In the present study, we also found moderate to high correlations between the RAQoL and measures of utility and functional status.

To analyse the construct of the RAQoL in more detail, the structure underlying the relationship among the items of the RAQoL was investigated. Physical contact (‘I'm afraid of people touching me’ and ‘I try to avoid shaking hands with people’), which is not covered by any of the currently used instruments, could be distinguished as a separate dimension, in addition to the dimensions energy/mobility, self‐care and mood/emotion. The reason why this dimension was found may be explained by the fact that in contrast to many other QoL questionnaires, the RAQoL was derived from the experiences of RA patients, following the needs‐based model of QoL [12]. In this model, QoL is defined as the extent to which RA interferes with the patient's ability to fulfil his or her needs. One of the advantages of using this model is that rather than asking about functions that are not relevant to all respondents it is possible to ask about the needs experienced by the respondents themselves. The results indicate that physical contact is an important aspect of QoL in patients with RA. Dimensions which were used in other QoL questionnaires were also identified in the RAQoL. The reliability of these dimensions and the percentage of variance were moderate. Therefore, we do not propose to divide the questionnaire into subscales.

According to the results of the RAQoL in the present study, females had a worse QoL than males. This phenomenon is in accordance with studies in which other QoL measures were used, both in patients with RA and in patients with other diseases [32, 33].

Economic evaluation has become an increasingly important component of the evaluation of new drugs and technologies in rheumatology. The basic methods for undertaking economic evaluation have been described elsewhere [34, 35]. A cost‐effectiveness analysis is used to compare two or more therapies for the same condition, with the emphasis on one single definition of the effect. Because the score of the RAQoL is expressed as one single value, and the present study showed that high correlations between the RAQoL and important outcome measures were found, the RAQoL has potential applications as an overall measure of effect in studies in which a cost‐effectiveness analysis is involved.

In conclusion, the RAQoL is a valid instrument for measuring QoL in different populations of patients with RA. A dimension which is not covered by other instruments, namely physical contact, could be distinguished within this questionnaire. By its nature, the RAQoL appears to be a potentially valuable instrument in studies with economic evaluations. The first results on sensitivity to change look very promising [14]. Further studies looking at the sensitivity to change are currently in progress.

Correspondence to: G. J. Tijhuis, Department of Rheumatology C4‐R, Leiden University Medical Center, P.O. Box 9600, 2300 RC Leiden, The Netherlands.

We thank the participating rheumatologists of the following hospitals: Leiden University Medical Center (Leiden), the Academic Hospital of the Free University (Amsterdam), Jan van Breemen Instituut (Amsterdam), Groene Hart Ziekenhuis (Gouda), Kennemer Gasthuis (Haarlem), Spaarne Ziekenhuis (Haarlem), Reinier de Graaf Gasthuis (Delft), Rode Kruis Ziekenhuis (Den Haag), Leyenburg Ziekenhuis (Den Haag), Bronovo Ziekenhuis (Den Haag). In addition, we are indebted to all rheumatology nurse specialists from the participating hospitals, to Mrs I. Henkes, A. Jongma, B. Oud and I. Perquin for doing the assessments and to Mrs M. de Best‐Bosz, H. Ravensbergen‐van de Berg, A. van der Blom and H. Kupka for secretarial assistance. Studies 1 and 2 were financially supported by Het Nationaal Reuma Fonds, The National Association against Rheumatism of the Netherlands, grant no. 931 and the Dutch Fund for Investigative Medicine, grant no. 97–024.

References

1
Hunt SM, McKenna SP, McEwen J, Williams J, Papp E. The Nottingham Health Profile: subjective health status and medical consultations.
Soc Sci Med
 
1981
;
15
:
221
–9.
2
Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure.
Med Care
 
1981
;
19
:
787
–805.
3
Ware JEJ, Sherbourne CD. The MOS 36‐item short‐form health survey (SF‐36). I. Conceptual framework and item selection.
Med Care
 
1992
;
30
:
473
–83.
4
Whalley D, McKenna SP, de Jong Z, van der Heijde D. Quality of life in rheumatoid arthritis.
Br J Rheumatol
 
1997
;
36
:
884
–8.
5
Meenan RF, Gertman PM, Mason JH. Measuring health status in arthritis. The arthritis impact measurement scales.
Arthritis Rheum
 
1980
;
23
:
146
–52.
6
Meenan RF, Mason JH, Anderson JJ, Guccione AA, Kazis LE. AIMS2. The content and properties of a revised and expanded Arthritis Impact Measurement Scales Health Status Questionnaire.
Arthritis Rheum
 
1992
;
35
:
1
–10.
7
Riemsma RP, Taal E, Rasker JJ, Houtman PM, Van Paassen HC, Wiegman O. Evaluation of a Dutch version of the AIMS2 for patients with rheumatoid arthritis.
Br J Rheumatol
 
1996
;
35
:
755
–60.
8
Evers AW, Taal E, Kraaimaat FW et al. A comparison of two recently developed health status instruments for patients with arthritis: Dutch‐AIMS2 and IRGL. Arthritis Impact Measurement Scales. Impact of Rheumatic Diseases on General Health and Lifestyle.
Br J Rheumatol
 
1998
;
37
:
157
–64.
9
Archenholtz B, Bjelle A. Reliability, validity and sensitivity of a Swedish version of the revised and expanded Arthritis Impact Measurement Scales (AIMS2).
J Rheumatol
 
1997
;
24
:
1370
–7.
10
Anderson JJ, O'Neill A, Woodworth T, Haddad J, Sewell KL, Moreland LW. Health status response of rheumatoid arthritis to treatment with DAB486IL‐2.
Arthritis Care Res
 
1996
;
9
:
112
–9.
11
Pouchot J, Guillemin F, Coste J, Bregeon C, Sany J. Validity, reliability and sensitivity to change of a French version of the Arthritis Impact Measurement Scales 2 (AIMS2) in patients with rheumatoid arthritis treated with methotrexate.
J Rheumatol
 
1996
;
23
:
52
–60.
12
Hunt SM, McKenna SP. The QLDS: a scale for the measurement of quality of life in depression.
Health Policy
 
1992
;
22
:
307
–19.
13
de Jong Z, van der Heijde D, McKenna SP, Whalley D. The reliability and construct validity of the RAQoL: a rheumatoid arthritis‐specific quality of life instrument.
Br J Rheumatol
 
1997
;
36
:
878
–83.
14
Wells G, Boers M, Shea B et al. Sensitivity to change of generic quality of life instruments in patients with rheumatoid arthritis: preliminary findings in the generic health OMERACT study.
J Rheumatol
 
1999
;
26
:
217
–21.
15
Arnett FC, Edworthy SM, Bloch DA et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis.
Arthritis Rheum
 
1988
;
31
:
315
–24.
16
Hochberg MC, Chang RW, Dwosh I, Lindsey S, Pincus T, Wolfe F. The American College of Rheumatology 1991 revised criteria for the classification of global functional status in rheumatoid arthritis.
Arthritis Rheum
 
1992
;
35
:
498
–502.
17
Hays RD, Sherbourne CD, Mazel RM. The RAND 36‐item Health Survey 1.0.
Health Econ
 
1993
;
2
:
217
–27.
18
Van de Zee KI, Sanderman R. Het meten van de algemene gezondheidstoestand met de RAND‐36. Groningen: Rÿksuniversiteit,
1993
.
19
Torrance GW, Thomas WH, Sackett DL. A utility maximization model for evaluation of health care programs.
Health Serv Res
 
1972
;
7
:
118
–33.
20
Netherlands Central Bureau for Statistics. Vademecum of Health Statistics in The Netherlands. The Hague: SDU Uitgevrij,
1995
.
21
Robinson A, Dolan P, Williams A. Valuing health status using VAS and TTO: what lies behind the numbers?
Soc Sci Med
 
1997
;
45
:
1289
–97.
22
Bakker C, Rutten M, van Doorslaer E, Bennett K, van der Linden S. Feasibility of utility assessment by rating scale and standard gamble in patients with ankylosing spondylitis or fibromyalgia.
J Rheumatol
 
1994
;
21
:
269
–74.
23
EuroQol Group. EuroQol: a new facility for the measurement of health‐related quality of life.
Health Policy
 
1990
;
16
:
199
–208.
24
Dolan P. Modeling valuations for EuroQol health states.
Med Care
 
1997
;
35
:
1095
–108.
25
Siegert CE, Vleming L‐J, Vandenbroucke JP, Cats A. Measurement of disability in Dutch rheumatoid arthritis patients.
Clin Rheumatol
 
1984
;
3
:
305
–9.
26
Jones E, Hanly JG, Mooney R et al. Strength and function in the normal and rheumatoid hand.
J Rheumatol
 
1991
;
18
:
1313
–8.
27
Grace EM, Gerecz EM, Kassam YB, Buchanan HM, Buchanan WW, Tugwell PS. 50‐foot walking time: a critical assessment of an outcome measure in clinical therapeutic trials of antirheumatic drugs.
Br J Rheumatol
 
1988
;
27
:
372
–4.
28
Smolen JS, Breedveld FC, Eberl G et al. Validity and reliability of the twenty‐eight‐joint count for the assessment of rheumatoid arthritis activity.
Arthritis Rheum
 
1995
;
38
:
38
–43.
29
Prevoo  MlL, van ’t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LBA, van Riel PLCM. Modified disease activity scores that include twenty‐eight joint counts. Development and validation.
Arthritis Rheum
 
1995
;
38
:
44
–8.
30
Kleinbaum D, Kupper L.
Applied regression analysis and other multivariable methods
 . Boston: Duxbury Press,
1978
.
31
Nunnally J.
Psychometric theory
 . New York: McGraw‐Hill,
1978
.
32
Dowdy SW, Dwyer KA, Smith CA, Wallston KA. Gender and psychological well‐being of persons with rheumatoid arthritis.
Arthritis Care Res
 
1996
;
9
:
449
–56.
33
Mathieson CM, Logan‐Smith LL, Phillips J, MacPhee M, Attia EL. Caring for head and neck oncology patients. Does social support lead to better quality of life?
Can Fam Physician
 
1996
;
42
:
1712
–20.
34
Griffiths B, Emery P, Akehurst R. Economic evaluation in rheumatology: a necessity for clinical studies.
Ann Rheum Dis
 
1995
;
54
:
863
–4.
35
McCabe CJ, Akehurst RL. Health economics in rheumatology.
Baillieres Clin Rheumatol
 
1997
;
11
:
145
–56.

Comments

0 Comments