Preferences for improved health examined in 1,024 patients with rheumatoid arthritis: Pain has highest priority




To examine preferences for improved health in patients with rheumatoid arthritis (RA).


A survey was mailed to patients with RA enrolled in a county-based register. The questionnaire comprised a variety of health status measures (Medical Outcome Study Short Form-36, Arthritis Impact Measurement Scales 2 [AIMS2], Modified Health Assessment Questionnaire, and visual analog scale for pain and fatigue). The patients were asked to check 3 of 12 areas in which they would most like to see improvement (item 60 AIMS2). The number of respondents was 1,024 (mean age/disease duration 63.4/12.7 years, 78.7% female).


Pain was the preferred area for improvement in all subgroups of patients. Preference for improvement in pain was associated with lower age, higher levels of perceived pain, and lower scores for self efficacy related to pain. One-third of the patients with this preference did not report use of pain-relieving medication.


Pain is the area of health in which almost 70% of the patients would like to see improvement. This study suggests that more attention should be paid to the examination of patient preferences for improvement in health.


Perceptions of patients' health status and need for care differ between health professionals and patients (1–3). The lack of patient-centered approach in assessing need for care may lead to ignorance of key symptoms and issues preferred by individuals (4). Attempts have been made to implement patients' preferences into some of the instruments used for detecting change in broad areas of function and performance of activities (5–7). However, few studies of rheumatoid arthritis (RA) have addressed the patients' preferences for improvement within multiple areas of health. Awareness of patients' preferences may offer health professionals the opportunity to include patients' own values and preferences when performing patient care (8). This again may lead to more effective and efficient delivery of health care (9).

Also of interest is that assessment of health care needs is one of the initial activities of the Bone and Joint Decade (10). Furthermore, a new patient rights law has been passed in Norway that confirms patients' rights for information, consent, and contribution in their own health care (11). This signifies a trend toward giving patients' opinions and preferences increased impact on priorities in health care performance. Reliable information about patients' needs and preferences is needed to create a more thorough basis for health care.

The main purpose of this study was to determine in which areas of health a large community-based sample of patients with RA would prefer to see improvement. Secondly, we aimed to compare patients with and without preferences for improvement in areas of health regarding demographic variables, self reported health status, self efficacy, and current use of medication.


Setting and patients.

The Oslo RA register served as the basis for this study. It was established in 1994 and includes patients with RA (12) who have a residential address in Oslo; it has been shown to have a completeness of 85% (13, 14). At the time of data collection it included 1,552 patients.

Data collection.

Data collection was performed via a survey mailed in 1994 to all patients enrolled in the register. The 30-page questionnaire asked respondents for a variety of information. The study variables included demographics (age, sex, years of education, disease duration, number of comorbidities, employment status), health status measures (Arthritis Impact Measurement Scales 2 [AIMS2]; Modified Health Assessment Questionnaire [MHAQ], Medical Outcomes Study Short Form-36 [SF-36]; pain, fatigue, and patient global assessment of disease activity), self efficacy, and current use of medications (15–18).


AIMS2 (15) is a revised and expanded version of the original AIMS (19). It includes 57 items that may be broken into 12 scales, representing 12 areas of health: mobility (5 items), walking and bending (5 items), hand and finger function (5 items), arm function (5 items), self-care tasks (4 items), household tasks (4 items), social activity (5 items), support from family and friends (5 items), arthritis pain (5 items), work (4 items), level of tension (5 items), and mood (5 items). The scales may be combined into a 5-component model of health dimensions reflecting the physical dimension (mobility level, walking and bending, hand and finger function, arm function, self-care tasks, household tasks), social interaction (social activity and support from family and friends), symptoms (pain), role (work), and affect (level of tension, mood). Item numbers 58 and 59 deal with satisfaction with and arthritis impact on problems in each of the 12 health areas. In item number 60, the main focus of this study, the patients are asked to report the 3 areas of health in which they would most like to see improvement (Table 1).

Table 1. Areas of health examined for patient preferences for improvement (question 60 in Arthritis Impact Measurement Scales 2)
EnglishNorwegian translation
Please check (x) up to three areas in which you would most like to see improvement:Kryss av opp til tre områder der du helst vil oppleve forbedring:
Mobility level (example: do errands)Rørlighet (for eksempel gjøre ærender)
Walking and bending (example: climb stairs)Å gå og bøye seg (for eksempel gå i trapper)
Hand and finger function (example: tie a bow)Rørlighet i hender og fingre (for eksempel knytte en knute)
Arm function (example: comb hair)Rørlighet i armene (for eksempel gre håret)
Self-care (example: take bath)Selvhjulpenhet (for eksempel ta et bad)
Household tasks (example: housework)Husarbeid (for eksempel rengjøring)
Social activity (example: visit friends)Kontakt med andre mennesker (for eksempel besøke venner)
Support from family (example: help with problems)Støtte fra familie og venner (for eksempel hjelp med problemer)
Arthritis pain (example: joint pain)Revmatiske smerter (for eksempel leddsmerter)
Work (example: reduce hours)Arbeid (for eksempel redusert arbeidstid)
Level of tension (example: felt tension)Anspenthet (for eksempel amper)
Mood (example: down in dumps)Sinnsstemning (for eksempel nedfor)

MHAQ is a shortened version of the Health Assessment Questionnaire (HAQ) (20) developed for patients with RA and osteoarthritis. It focuses on physical function (16).

SF-36 is a 36-item measure of 8 different dimensions of health status represented in 8 scales with scores of 0–100: physical functioning (10 items), physical role (4 items), bodily pain (2 items), general health (5 items), vitality (4 items), social functioning (2 items), emotional role (3 items), and mental health (5 items). It also contains 1 item measuring a self evaluated change in health status. The Norwegian version has been validated and found to perform well in patients with RA (17, 21).

Joint pain and fatigue were measured on 100-mm visual analog scales, and patient global assessment of disease activity was measured on a scale of 1–5.

Bandura (22) has suggested that self efficacy is understood as people's judgements of their capabilities to organize and execute courses of action required to attain designated types of performance. It is concerned not with the skill one has, but with the judgement of what one can do with whatever skill one possesses. Self efficacy for pain and other symptoms was measured on Lorig's scale with the range of 10–100 (18). Current use of medications was self reported on a 1-page checklist of the brand names of all available antirheumatic drugs at the time of the study, including different nonsteroidal antiinflammatory drugs (NSAIDs), analgesics (paracetamol, codeine), disease-modifying antirheumatic drugs (DMARDs), and corticosteroids.

Statistical analyses.

Distribution of responses to question 60 in AIMS2 (Table 1) was examined across subgroups based on demographic variables. Clinical and demographic variables were compared between patients with and without preferences for improvement in the various areas of health by using two-sample t-tests (continuous variables) and chi-square tests (categorical variables). P-values are marked when significance remains after Bonferroni correction for chance significance, (i.e., each P-value multiplied by the number of separate analyses.) Multivariate associations were explored by logistic regression analyses. Data were analyzed using SPSS software version 8.0 (SPSS, Chicago, IL).


The number of respondents was 1,024 (66%), mean (SD) age was 63.4 (14.8) years, mean (SD) disease duration 12.7 (11.1) years, and 78.7% were women. Respondents were somewhat younger (mean difference 5.4 years) than nonrespondents, and had shorter disease duration (mean difference 1.9 years), whereas there were no differences in distribution of sex and presence of rheumatoid factor.

Seven-hundred two (68.6%) patients reported pain as a preferred area for improvement. Priority for mobility was reported by 245 (23.9%), walking and bending by 341 (33.3%), hand and finger function by 457 (44.6%), and arm function by 189 (18.5%). Within the other areas, the numbers were as follows: self care 122 (11.9%), household tasks 257 (25.1%), social activity 135 (13.2%), support from family 53 (5.2%), work 92 (9.0%), level of tension 89 (8.7%), and mood 177 (17.3%) (Figure 1).

Figure 1.

Distribution of preferences for improvement in different areas of health. Mob = mobility; fing = finger; Soc.act = social activity; Supp = support; Tens = tension.

The distribution of preference for improvement in the various health areas differed somewhat according to age, with a relatively higher preference for improvement in pain, work, and mental conditions in the younger respondents, and physical functioning in the elderly. There was no difference in preference between patients on disablement benefit and those who worked full time. However, pain was consistently the area of health that had highest priority for improvement in all subgroups (Table 2).

Table 2. Number of patients (%) in demographic subgroups with preferences for improvement in areas of health
 No.MobilityWalking/bendingHand/fingerArmSelf careHouseSocial activitySupportPainWorkTensionMood
  • *

    P < 0.05 chi-square test between subgroups.

  • P < 0.01 chi-square test between subgroups.

All1024245 (23.9)341 (33.3)457 (44.6)189 (18.5)122 (11.9)257 (25.1)135 (13.2)53 (5.2)702 (68.6)92 (9.0)89 (8.7)177 (17.3)
 Female806190 (23.6)258 (32.0)349 (43.4)149 (18.5)101 (12.5)236 (29.3)111 (13.8)45 (5.6)559 (79.7)73 (9.1)69 (8.6)145 (18.0)
 Male21855 (25.2)83 (38.1)107 (49.1)40 (18.3)21 (9.6)21 (9.6)24 (11.0)8 (3.7)142 (65.1)19 (8.7)20 (9.2)31 (14.2)
Age groups (years)             
 <4010821 (19.4)32 (29.6)43 (39.8)11 (10.2)4 (3.7)16 (14.8)13 (12.0)4 (3.7)80 (74.1)27 (25.0)17 (15.7)22 (20.4)
 40–499315 (16.1)25 (26.9)37 (39.8)11 (11.8)7 (7.5)17 (18.3)19 (20.4)5 (5.4)69 (74.2)21 (22.6)14 (15.1)23 (24.7)
 50–5915631 (19.9)55 (35.3)68 (43.6)38 (24.4)13 (8.3)39 (25.0)15 (9.6)10 (6.4)117 (75.0)26 (16.7)15 (9.6)28 (17.9)
 60–6926357 (21.7)96 (36.5)124 (47.1)52 (19.8)35 (13.3)73 (27.8)31 (11.8)10 (3.8)178 (67.7)12 (4.6)22 (8.4)43 (16.3)
 70–7931695 (30.1)105 (33.2)148 (46.8)58 (18.4)48 (15.2)89 (28.2)44 (13.9)17 (5.4)206 (65.2)3 (0.9)17 (5.4)47 (14.9)
 >798826 (29.5)*28 (31.8)37 (42.0)19 (21.6)*15 (17.0)23 (26.1)13 (14.8)7 (8.0)52 (59.1)*3 (3.4)4 (4.5)14 (15.9)
Employment status in patients <65 years of age             
 Employed29560 (20.3)101 (34.2)117 (39.7)53 (18.0)23 (7.8)65 (22.0)37 (12.5)15 (5.1)225 (76.3)49 (16.6)36 (12.2)53 (18.0)
 Disablement benefit465 (10.9)16 (34.8)24 (52.2)9 (19.6)3 (6.5)12 (26.1)8 (17.4)5 (10.9)31 (67.4)5 (10.9)8 (17.4)7 (15.2)

In general, preference for improvement in one area of health was associated with worse reported level of health in the same area, across the different health status instruments (data not shown). For example, patients with preference for improvement in pain reported more severe pain than those not having pain as a preferred area for improvement. Likewise, greater fatigue and worse global health were seen in patients with preference for pain improvement (Table 3). However, levels of physical disability, mental health, and social functioning (data not shown) were similar in groups with and without preference for improvement in pain.

Table 3. Comparison of patients with and without preference for improvement in pain regarding demographic variables, health status, self efficacy, and current use of medications*
VariablesWith preference (n = 701)Without preference (n = 323)P
  • *

    Values are mean ± SD unless otherwise indicated. AIMS2 = Arthritis Impact Measurement Scales 2; MHAQ = Modified Health Assessment Questionnaire; NSAIDs = Nonsteroidal antiinflammatory drugs; SF-36 = Medical Outcomes Study Short Form-36; VAS = visual analog scale.

  • Remains significant after Bonferroni correction, i.e., each P value has been multiplied by the number of separate analyses (n = 21).

Demographic variables   
 Age, years61.5 (14.9)64.3 (14.6)0.005
 % female79.776.40.25
 Years of education11.3 (3.3)11.6 (3.5)0.17
 Disease duration, years12.8 (11.1)12.5 (11.2)0.71
 Number of comorbidities1.10 (1.22)1.03 (1.28)0.42
Health status, physical   
 MHAQ score1.71 (0.57)1.67 (0.66)0.36
 AIMS22.89 (2.02)2.84 (2.28)0.76
 SF-36, physical functioning47.0 (25.0)48.5 (29.0)0.44
Health status, pain   
 AIMS25.99 (2.38)4.20 (2.83)<0.001
 VAS50.9 (21.5)35.3 (23.0)<0.001
 SF-36, bodily pain37.1 (19.4)49.4 (25.5)<0.001
Health status, fatigue   
 VAS52.1 (26.3)45.4 (29.0)0.001
 SF-36, vitality38.1 (21.3)43.0 (24.2)0.003
Health status, global   
 Patient global3.13 (0.91)2.67 (1.01)<0.001
 SF-36, general health40.3 (20.4)46.5 (24.0)<0.001
 AIMS24.77 (2.71)4.20 (2.92)0.004
Self efficacy   
 Pain50.2 (19.6)56.6 (22.3)<0.001
 Other symptoms60.8 (19.4)63.2 (20.5)0.08
Current use of medications   
 % NSAIDs49.436.0<0.001
 % Analgesics27.223.60.25
 % NSAIDs or analgesics64.654.40.002

The bivariate association between preference for improvement in pain and reported pain (Table 3) was further explored after dividing perceived pain intensity (AIMS2, VAS, SF-36 bodily pain) into quartiles (Table 4). It appeared that pain was reported as a preferred area for improvement in approximately 50% of the patients in the group with lowest pain intensity. About 15–20% of those with the most severe pain levels did not report pain as their preferred area for improvement. These bivariate associations between preference for improvement in pain and perceived pain intensity remained after adjusting for age, sex, and level of self efficacy (Table 4).

Table 4. The relationships between preference for improvement in pain and various levels of pain perception divided into quartiles Q1–Q4*
Levels of pain perceptionPercentages with preference for improvement in painMultivariate association
  • *

    OR = odds ratio; CI = 95% confidence interval; AIMS2 = Arthritis Impact Measurement Scales 2; SF-36 = Medical Outcomes Study Short Form-36; VAS = visual analog scale.

  • Logistic regression analyses with adjustment for age, gender, and level of pain self efficacy.

 Q266.72.101.44, 3.07
 Q376.43.432.24, 5.24
 Q484.56.223.77, 10.27
 Q272.42.461.71, 3.56
 Q370.82.211.36, 3.58
 Q480.13.732.40, 5.80
Pain VAS   
 Q268.82.211.49, 3.29
 Q377.23.462.25, 5.32
 Q484.55.873.61, 9.55

Lower scores for pain self efficacy were observed in patients with preference for improvement in pain (Table 3), and this association remained significant in logistic regression analyses after adjustment for pain intensity, as measured by VAS, AIMS2, or SF-36 (data not shown).

Those with preference for improvement in pain used, as expected, more NSAIDs, whereas use of analgesics, DMARDs, and corticosteroids was similar (data not shown). However, more than one-third of the patients in both groups did not report use of pain-relieving medication (Table 3).


The mandate of health professionals is to assess patients' needs for health care, and effective delivery is dependent on matching these needs. The frame of reference held by patients, particularly those with chronic disease, possesses valuable experience about their needs and how to meet them, either by self management or professional care. Thus, the patient should be a partner in the process of disease management, contributing at almost every decision and action level (9, 23). Awareness of the patients' preferences may be one salient premise for priorities in health care. This is further underlined by differences in perception of patients' health status and need for care between health professionals and patients (1–3). In this study, pain was reported as the preferred area for improvement in all subgroups of patients (Table 2). The distribution of preferences for improvement in areas of health in this study may encourage increased awareness and attention of pain in RA.

Gathering data from a large number of patients suggested to be representative of the total underlying patient population (13, 14) is one strength of the present study. Another strength is the comprehensiveness of the data collection that permits comparisons of preferences and perceptions across both disease-specific and generic instruments.

Questionnaires based on self-reported data rely on subjective statements that may be criticized as vulnerable to confounding factors and variance. Objective measurement has been the gold standard in evidence-based medicine. However, self-report questionnaires have been used as valid assessment tools for more than 20 years. Substantial data now suggest that they provide important and clinically useful information (24). This is of major importance because some dimensions, for instance pain, represent a subjective perception experienced by each individual. Because preference for improvement within one particular area of health was associated with worse health status in that area, it may be relevant to ask whether assessment of health status would be a sufficient indicator of patient preference. However, many patients with low pain scores reported pain as the area of health where they would like to see improvement (Table 4). This discrepancy indicates that assessment of patient preferences for improvement in health provide complementary information to traditional health status assessments, but further research is required to clarify this issue.

AIMS2 was used in this study because it is one of the few instruments that addresses patient preferences. Since 1992 it has been translated and validated in a number of languages for use in rheumatic diseases, and has been found valid, reliable, and responsive. It is, however, possible that AIMS2 is not the ideal instrument for a comparative analysis of preferences across different health dimensions. In AIMS2, 6 items represent physical functions, whereas only 1 item represents pain. The distribution of preferences could have been different if the 5 major dimensions of health in AIMS2 (physical functions, social functioning, pain, role, and affect) were represented by only 1 item each. Other studies have shown that patients with RA report both pain and loss of function as their most important problems (25, 26).

Patients' preferences are also addressed in the McMaster Toronto Arthritis Patient Preference Disability Questionnaire (MACTAR) (5) and in the Problem Elicitation Technique (PET) (6). The MACTAR and PET mainly ask the patients to indicate which functional ability they would most like to see improved (6, 7). Pain is seen as an obstacle to performance of activities but, in contrast to AIMS2, pain is not used as a separate and standardized outcome alternative (7). Both MACTAR and PET have been found to be responsive in clinical trials (6, 7), but feasibility is limited because they require an interviewer.

In comprehensive surveys mailed to a large number of elderly and disabled patients, nonrespondents may constitute a limitation (24). In the present study, nonrespondents were older than the respondents, indicating that a higher respondent rate would have influenced the results to some extent because the preferences for improvement in several areas of health were somewhat related to age (Table 2).

Pain as the preferred area for improvement in health was explicit and consistent across subgroups of patients. Comparable results have been found in other studies using question 60 of AIMS2 performed in different countries, at different times, and in smaller numbers of patients not necessarily reflecting the underlying population (15, 27).

We found that pain was the preferred area of improvement in full-time workers and patients on disablement benefit (Table 2), but pain intensity was reported to be significantly more severe in patients on disablement benefit (data not shown). The phenomenon of different health status and same preferences has also been found in other chronic diseases (28).

One surprising finding was that more than one-third of the patients with preference for improvement in pain did not use relevant symptom-modifying medications. The reason for this was not explored, but side effects of NSAIDs, contraindications with these drugs, and fear of future adverse reactions are plausible explanations. New symptom-modifying drugs with less toxicity, such as the cyclooxygenase 2 inhibitors (29, 30), have potential to be used by patients who, for medical reasons, have no or limited access to traditional NSAIDs.

Several studies have shown that low self efficacy and pain severity are closely related (31, 32). Self efficacy has also been found to be related to fewer pain behaviors (33), and self efficacy may be improved by self-management programs (34, 35). We found that low levels of self efficacy were related to preference for improvement in pain, even after adjusting for the level of perceived pain. This observation may hint at self efficacy as a therapeutic target in the management of pain, which is further supported by longitudinal studies indicating that self efficacy has a stronger pain-relieving impact on subsequent pain than pain has on the future self-efficacy level (36).

Medical treatment and patient education represent 2 polar but complementary approaches in the care of patients with RA. One is invasive, the other cognitive. Traditionally, nurses emphasize more educational and psychosocial care in their approach, whereas physicians make more routine use of state-of-the-art technology (37, 38). Pain in patients with RA is an area of health that may benefit from a multidisciplinary approach, because management depends on competency from various professions. Pain management as a main therapeutic target may have been left in a vacuum between professions. Nurses may consider pain as mainly the domain of physicians, as it is often handled successfully through decisions about pharmacologic interventions. Physicians, on the other hand, may focus more on physiologic issues concerning diagnoses and progression of diseases, as confirmed by laboratory findings (39). The challenges are to collaborate and to establish a pain-management program based on patients' preferences and health status.

In summary, this study highlights that pain is the preferred area for improvement in patients with RA, and suggests that more attention should be paid to the assessment of patients' preferences for improvement in health—both in research and in clinical practice.