JOINT ASSESSMENT IN RHEUMATOID ARTHRITIS

Transcription

1 British Journal of Rheumatology 1996;35(suppl.2):14-18 JOINT ASSESSMENT IN RHEUMATOID ARTHRITIS D. L. SCOTT and D. A. HOUSSIEN Academic Rheumatology Unit, King's College School of Medicine and Dentistry, London SUMMARY Determining the number of swollen joints and tender joints is a key component in the clinical assessment of rheumatoid arthritis (RA). There have been a series of investigations carried out in the last decade, which have defined the best ways to measure joint inflammation and have identified which joints should be evaluated. There is not complete agreement on the optimal joint count, but two approaches are widely used. These comprise counting either 66/68 or 28 joints. The main difference is that the 28-joint count excludes the joints in the feet. Both methods give similar information and are reproducible and valid. Tenderness and swelling should be measured separately. There are advantages and disadvantages associated with using the 28-joint count It has the benefit of simplicity and takes less time, although some potentially relevant clinical information about the feet may be lost There is general agreement that grading the severity of individual joint involvement is of limited advantage. Using weighted joint counts is also not widery accepted. Finally there is growing recognition of the need for training in the methods of assessing joints and the importance of international standardization. Joint counts are a component of the core clinical data set for RA and will continue to play a key role in the foreseeable future. KEY WORDS: Rheumatoid arthritis. Joint assessment, Joint counts. RHEUMATOID arthritis (RA) is characterized by joint tenderness and swelling. Assessing its severity involves measuring the extent of tenderness and swelling. This apparently simple task is complex and requires careful clinical methodology. In this article we review the nature of joint swelling and tenderness, their assessment by standardized joint counts, how these should be used in conjunction with other clinical measures and their prognostic significance. JOINT SWELLING AND TENDERNESS Joint swelling is soft tissue swelling that is detectable along the joint margins. When a synovial effusion is present it invariably means the joint is swollen. Neither bony swelling nor deformity of the joints constitutes joint swelling. Fluctuation is a characteristic feature of swollen joints and it may influence the range of joint movement. Joint tenderness is the presence of pain in a joint at rest with pressure or on movement of the joint, e.g. on movement of the hip joints. Pressure to elicit tenderness should be exerted by the examiner's thumb and index finger sufficient to cause 'whitening' of the examiner's nail bed, which is a useful 'rule of thumb'. JOINT COUNTS A number of joint counts have been developed; these are detailed in Table I. The main joint counts are the ACR 66/68 counts for swollen and tender joints, the Ritchie articular index and the 28-joint count The 66/68-joint count [1] evaluates the following joints: upper temporomandibular, sternoclavicular, acromioclavicular, shoulder, elbow, wrist, meta- Correspondence to: Dr D. L. Scott, Clinical and Academic Rheumatology, King's College Hospital (Duhvich), East Duhvich Grove, Duhvich, London SE22 8PT, UK. carpophalangeal, proximal interphalangeal and distal interphalangeal; lower hip, knee, ankle, tarsus, metatarsophalangeal and interphalangeal. The Ritchie articular index [2] evaluates the following joints: upper temporomandibular, sternoclavicular, acromioclavicular, shoulder, elbow, wrist, metacarpophalangeal (as a group), proximal interphalangeal (as a group) and cervical spine; lower hip, knee, ankle, subtalar, midtarsal and metatarsophalangeal (as a group). The 28-joint count [3] evaluates the following joints: upper shoulder, elbow, wrist, metacarpophalangeal and proximal interphalangeal; lower knee. WEIGHTED JOINT COUNTS There are two main weighted counts: the Ritchie articular index [2] and the Thompson-Kirwan index [4]. The Ritchie index scores 53 joint groups and each are scored on a 0-3 scale. The Thompson-Kirwan index scores 38 joint groups, which are then weighted for their surface area. The knee is weighted more than other joints. There are theoretical advantages in using a weighted index, but the practical advantages seem modest and there may be greater inter-rater variability. COMPARISON OF JOINT COUNTS The various articular indices have been compared in a longitudinal study by Prevoo et al [5], who examined the validity and reliability of seven indices: the Ritchie articular index [2], a modified Ritchie articular index, the Thompson-Kirwan index [4], the 28-joint index, the 36-joint index, total tender joints and total swollen joints. They found that the validity and reliability of traditional joint indices did not differ substantially. Weighted joint indices seemed less valid andreliable. No joint index was superior for measuring disease activity. Taking simplicity into account, the 28-joint index [3], not graded and not weighted, was recommended as preferable. O 19% British Society for Rheumatology 14

2 SCOTT AND HOUSSIEN: JOINT ASSESSMENT IN RA 15 Index ACR tender joint count [1] ACR swollen joint count [1] Ritchie articular index m 1*1 Total tender joint count [3] Total swollen joint count [3] Thompson and Kirwan index [4] 36-joint count [3] 28-joint count [5] TABLE I Different types of joint counts Number of joints Grading Comment Swollen joints only Swollen joints only Combination index of tenderness and swelling with weighting Separate measures of tender and swollen joints Separate measures of tender and swollen joints EXPERIENCE WITH THE 28-JOINT COUNT How many joints need to be measured? Is the reduced joint count sufficient? These issues were examined in detail in two recent papers by Fuchs and Pincus [6] and Smolen et al [7]. Fuchs and Pincus investigated whether quantitative assessment of a reduced number of joints by the 28-joint count provides information equivalent to that obtained by the traditional 60-joint evaluation. They were most interested in detecting changes in patients participating in clinical trials. They looked at changes in quantitative joint scores of patients in three clinical trials. These were compiled and compared with changes in quantitative scores derived using a reduced number of joints. Effect sizes, calculated as mean change in joint score/standard deviation of joint score, were similar for the initial 60-joint and reduced 28-joint scores. The reduced joint count scores gave significant changes for clinical trials involving as few as 15 patients. They concluded that 28-joint count scores may be used in clinical trials without decreasing the ability to detect change over time. They also felt that reduced joint counts were more appropriate for quantitative assessments in the routine care of patients with RA. Smolen et al. also investigated the validity of the 28-joint count in comparison to the 66/68-joint count. They used data from 735 prospectively studied RA patients. They found that joints included in the 28-joint count were more commonly involved than other joints. They also noted that findings from the 28-joint count correlated highly with those from the 66/68-joint count in all analyses. They considered that the 28-joint count is a reliable and valid measure for joint assessment, is easier to perform than the 66/68-joint count and addresses the joints that are critically involved. Prevoo et al. [8] showed that the 28-joint count can be used in calculations of disease activity. The matter is, however, by no means clearly decided. Johnson [9] has continued to express concern about the 28-joint count. In many instances in routine clinical practice there will be important information from the joints in the feet relevant to management decisions; these joints should not be ignored even if they are not always counted. VARIABILITY IN ASSESSMENTS OF JOINTS Variability presents a problem in clinical assessments; this especially applies to joint assessment in RA. The first study to examine this variability looked at the extent of the inter-rater error. Hart et al. [10] used the Ritchie articular index for this purpose. Three physicians recorded independent joint scores on each of 18 patients, examined in random order. They found that close agreement was achieved on global Ritchie articular index scores and that raters can achieve reasonable agreement on the absolute presence or absence of tenderness of individual joints. By contrast, inter-rater agreement hardly exceeds the chance level when degree of tenderness is independently assessed. The grading system of the Ritchie articular index may thus be implicated as an important source of the instrument's inter-rater error. Lewis et al [11] examined the reliability of assessing joint tenderness in RA. Reliability both within and between observers was measured and the 95% confidence intervals for a change in score calculated. Forty-two patients were assessed by two metrologists over a 14 week period using three subjects per week. Results showed close agreement within and between metrologists. The 95% confidence intervals for repeat measures by the same metrologist was ±8.5 and ±12 by a different metrologist. Taking a change in the score of 5 as clinically significant, the 95% confidence intervals for a clinically significant change would be ±14 with one observer and ±17 when a different observer repeated the assessment. They recommended that other centres should establish their own confidence intervals for a change in scores when undertaking clinical trials Using a Latin square design, Thompson et al. [12] studied the examination of four patients with RA by four rheumatologists. Joints were scored for tenderness and inflammation. The Ritchie index, a modified Ritchie index and two other similar approaches were calculated from the raw data. They found that an articular index consisting of a simple count of tender joints (a modified Ritchie), or a simple count of tender or swollen joints, is the most reproducible with multiple observers. They concluded that these indices are potentially the most appropriate for multicentre clinical trials. Variation across European centres was highlighted in two collaborative investigations reported by Scott et al [13, 14]. They examined a variety of clinical measures and found that determining the number of swollen joints and the number of tender joints were the most

3 16 CONTROLLING PROGRESSION OF SEVERE RA stable clinical methods. Nevertheless there was considerable inter-centre variation. A further study by Scott et al. [15] examined the extent of variation in measurement of joint swelling and tenderness and evaluated the impact of training to standardize methods. Eight assessors examined a group of RA patients before and after training in clinical methods. There was extensive variability in determining numbers of swollen and tender joints. Coefficients of variation for articular indices recorded by the eight observers in individual patients were high, indicating considerable differences between assessors. The mean coefficients of variation fell after training the assessors. Training had a major impact on the assessment of the numbers of swollen joints, which these increased by a mean of 32% and also the number of tender joints, which increased by 41%. These results highlight the extent of variation in clinical assessment of RA and stress the advantages of training, which leads to increased sensitivity of measurement. THE EFFECTS OF PATIENTS USING A SELF-REPORTED ARTICULAR INDEX A self-report measure of rheumatoid arthritic activity was developed by Stewart et al. [16] based on the articular index described by Thompson et al [12]. The self-report measure was tested for reliability, validity and sensitivity to therapeutic change. Concurrent validity was demonstrated by a significant correlation between the self-report measure and a standard Thompson index completed by a rheumatologist. Test retest reliability was demonstrated by significant correlations between repeated administrations of the self-report form in two independent samples. The selfreport measure demonstrated sensitivity to therapeutic change when completed by patients both before and after intra-articular corticosteroid injection. Further studies showing the value and reliability of selfreporting by patients have been undertaken by Mason et al [17] and Stucki et al. [18]. The main problem is relating self-reported indices from patients to more conventional clinical assessments of joint involvement. PATIENTS' PERSPECTIVES Concordance between the clinician's and the patient's assessment of the patient's physical and mental functioning was examined in 166 RA patients by Kwoh et al [19]; physicians and their patients with RA often disagreed in their assessment of the degree of physical and mental impairment that the patient experiences. Various demographic, medical and psychological factors influence patients' perceptions. This was examined by Parker et al [20] in 135 patients with RA. Patients were examined using the systemic index, articular index, McGill Pain Questionnaire, Symptom Checklist-90-R and a pain visual analogue scale. Multiple regression analyses found no significant relationships between pain and the medical variables. Age, income and selected psychological variables were significantly correlated with pain. The greatest management challenge occurred in patients who were TABLEn Core data set agreed by OMERACT, ACR, EULAR, and ILAR for assessing RA Core assessments Number of swollen joints Number of tender joints Pain assessed by the patient Patient's global assessments of disease activity Physician's global assessments of disease activity Laboratory evaluation of an acute-phase reactant (erythrocyte sedimentation rate, C-reactive protein or equivalent) Self-administered functional assessment (such as the Health Assessment Questionnaire) middle-aged, living on limited incomes and experiencing major stresses in everyday life. These patients were worry-prone and felt isolated and lacking in social support. MENTAL STATE There is considerable evidence that psychological factors influence the joint count. One study of 50 patients reported by Salaffi et al. [21] showed joint count and tenderness (Ritchie's index) correlated in a highly significant manner with depression measured using validated assessment methods. Pain is an important component of joint tenderness. Parker et al [20] have demonstrated that depression in RA cases influences joint pain and can thus affect the joint count. But, conversely, pain does not affect depression. A larger study of >700 RA patients by Wolfe and Hawley [22] suggested that disease activity in RA, which includes the assessment of the joint count, has a small influence on depression; this is in theregionof 20%, but it is not marked. Pain seems to be the most important component. GENETIC VARIATIONS Variations in the articular index could have a cultural basis. The features of RA were compared in 108 Greek and 107 British patients with RA by Drosos et al [23], who found that British patients had more severe articular involvement than did Greeks, as judged by the duration of morning stiffness, grip strength, and the numbers of swollen and tender joints. Although genetic and environmental factors may be responsible for such differences in disease expression between these European populations with RA, some of the variation may be related to the clinical methods of measuring joint involvement. CLINICAL END-POINTS Outcome measures should be based on the core data set. This was proposed at Maastricht [24] and this is accepted by EULAR [25], WHO, ILAR, ACR [26, 27] and other groups. The core data set is shown in Table II. It includes tender joint counts, swollen joint counts,

4 SCOTT AND HOUSSIEN: JOINT ASSESSMENT IN RA > 6 tender joints 17 TABLEm Relationship of number of tender and number of swollen joints to disease duration and patients' global assessments > 6 swollen Joints Disease duration ESR > 30 mnvh All three Variable Tender joints Swollen joints Correlation Patients' global i Variable Tender joints Swollen joints Correlation Data are from 153 current clinic attendees with RA. Fio. 1. Disease activity from 153 clinic attendees with RA. PROGNOSTIC MARKERS Many factors predict RA outcome, including disease severity and the number of swollen and tender joints. Kaarela [30] evaluated predictors of outcome in 442 patients with recent arthritis and compared 22 variables recorded at the onset of arthritis. Destructive arthritis was indicated at the onset of inflammatory synovitis by symmetrical polyarthritis of peripheral joints, serum rheumatoid factor, morning stiffness, high ESR and old age. Patients with the most severe RA have the worst prognosis. Single measures of disease activity are relatively weak predictors of outcome; a combination of measures or multiple values are better. Hassell et al [31] evaluated disease activity measured annually over 7 yr in 127 patients. They studied the relationship of serial measures of disease activity (looking at the area under the curve for each variable followed with time) to outcome, measured radiologically, functionally and by global assessment. There were significant correlations between persistent disease activity and radiographic APPLICATION IN ROUTINE CLINICAL PRACTICE What happens if articular indices are routinely measured in a specialist clinic? The information from our own clinic has been quite surprising to us, though undoubtedly relatively commonplace. Firstly, many patients on long-term therapy with slow-acting anti-rheumatic drugs such as gold and methotrexate have active RA despite therapy. Secondly, the measures of joint swelling and tenderness correlate closely with the patient's general well-being. We have analysed how these changes affect patients seen in our own centre. Our experience with 153 outpatients with RA [34] is summarized in Fig. 1 and Table III. The key question is whether we should routinely measure the joint count. The advantage is the added dimension of relevant clinical data it gives. The disadvantage is the time taken. We estimate that it adds 5-10 min to each routine consultation. Whether the benefits of measurement make the costs of collecting the information worthwhile remains to be demonstrated. This may be where self-reported joint counts undertaken by patients become most important. patient's assessment of pain, global assessments of disease activity and laboratory evaluation of an acute-phase reactant. Using a small number of end-point measures can improve RA clinical trials and clinical practice. Van der Heide et al. [28] reviewed 32 clinical trials on six slow-acting anti-rheumatic drugs. They noted that it is common practice to evaluate multiple outcome measures. They suggested that the number of measures should be reduced to those considered important, sensitive to change and able to differentiate between drugs. A joint count, assessment of pain, a questionnaire on functional status and measurement of erythrocyte sedimentation rate (ESR) are sufficient This fits in with the concept of the minimum data set. In these circumstances, what is a clinically important change? Using the core data set, Goldsmith et al [29] examined this idea, based on the opinion of rheumatologists. An important improvement consists of a change of 17% in the number of swollen joints. It is greater for other measures; for example, a 49% change is needed for measures of disability. deterioration. Van Zeben et al [32] looked at prognostic factors in a study of early diseases in 132 RA patients followed from an early phase of the disease (symptoms <5 yr) for 6 yr. They found that a combination of three commonly available variables (number of swollen joints, IgM-rheumatoid factor and the erosion score) could predict outcome quite well. Thus patients with many swollen and tender joints have a worse outcome. Pincus et al. [33] suggested joint count measures could predict long-term mortality. In a cohort of 75 patients with RA followed for up to 15 yr they found significant predictors of mortality included age, formal education level, joint count, activities- of-daily-living questionnaire scores, disease adjustment scores, morning stiffness, co-morbid cardiovascular disease, grip strength, modified walking time and button test. Five-year survival in patients with the poorest status according to these quantitative measures was 40-60%, comparable to the expected survival at that time of patients with three-vessel coronary artery disease or with stage 4 Hodgkin's disease. Simplified measures, including a count using only 28 joints and a questionnaire using only eight activities of daily living, were similar to the more elaborate traditional measures for predicting mortality.

5 18 CONTROLLING PROGRESSION OF SEVERE RA REFERENCES 1. Cooperating Clinics Committee of the American Rheumatism Association. A seven-day variability study of 499 patients with rheumatoid arthritis. Arthritis Rheum 1965;8J Ritchie DM, Boyle JA, Mclnnes JMetal Clinical studies with an articular index for the assessment of joint tenderness in patients with rheumatoid arthritis. Q J Med 1968^7: Fuchs HA, Brooks RH, Callahan LF, Pincus T. A simplified 28 joint quantitative articular index in rheumatoid arthritis. Arthritis Rheum : Thompson PW, Silman AJ, Kirwan JR, Currey HLF. Articular indices of joint inflammation in rheumatoid arthritis. Arthritis Rheum : Prevoo ML, van Riel PL, van't Hof MA et al Validity and reliability of joint indices. A longitudinal study in patients with recent onset rheumatoid arthritis. Br J Rheumatol ^ Fuchs HA, Pincus T. Reduced joint counts in controlled clinical trials in rheumatoid arthritis. Arthritis Rheum : 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: 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. Arthritis Rheum : Johnson K. Limitations in the usefulness of the 28-joint count. Arthritis Rheum : Hart LE, Tugwell P, Buchanan WW, Norman GR, Grace EM, Southwell D. Grading of tenderness as a source of interrater error in the Ritchie articular index. / Rheumatol 1985:12: Lewis PA, O'Sullivan MM, Rumfeld WR, Coles EC, Jessop JD. Significant changes in Ritchie scores. Br J Rheumatol 1988;27: Thompson PW, Hart LE, Goldsmith CH, Spector TD, Bell MJ, Ramsden ME Comparison of four articular indices for use in clinical trials in rheumatoid arthritis: patient, order and observer variation. J Rheumatol 1991;18: Scott DL, Panayi GS, van Riel PLCM, van de Putte LBA, Smolen J. Disease activity in rheumatoid arthritis: pre-liminary report of the Consensus Study Group of the European Workshop for Rheumatology Research. Clin Exp Rheumatol 1992;10: Scott DL, Panayi GS, van Riel PLCM. Variations between centres when assessing disease activity. Clin Rheumatol 1993;12J Scott DL, Choy E, Greeves A et al Standardising joint assessment in arthritis. Clin Rheumatol 1994;13: Stewart MW, Palmer DG, Knight RG, Highton J. A self report articular index: relationship to variations in mood and disease activity measures. Br J Rheumatol 1993; 32: Mason JH, Anderson JJ, Meenan RF, Haralson KM, Lewis-Stevens D, Kaine JL. The rapid assessment of disease activity in rheumatoid arthritis (RADAR) questionnaire: validity and sensitivity to change of a patient self-report measure of joint count and clinical status. Arthritis Rheum : Stucki G, Liang MH, Stucki S, Bruhlmann P, Michel BA. A self-administered rheumatoid arthritis disease activity index (RADAI) for epidemiologk research. Psychometric properties and correlation with parameters of disease activity. Arthritis Rheum 19953* Kwoh CK, O'Connor GT, Regan Smith MG et al Concordance between clinician and patient assessment of physical and mental health status. J Rheumatol 1992;19: Parker 1 Frank R, Beck N et al Pain in rheumatoid arthritis: relationship to demographic, medical and psychological factors. J Rheumatol 1988;15: Salafii F, Ferracrioli GF, Carotti M, Blasetti P, Cervini C. Disability in rheumatoid arthritis: the predictive value of age and depression. Rec Prog Med 1992;83: Wolfe F, Hawley DJ. The relationship between clinical activity and depression in rheumatoid arthritis. J Rheumatol 1993;2(h Drosos AA, Lanchbury JS, Panayi GS, Moutsopoulos HM. Rheumatoid arthritis in Greek and British patients. A comparative clinical, radiologic and serologic study. Arthritis Rheum : Tugwell P, Boers M. OMERACT conference on outcome measures in RA clinical trials: conclusion. J Rheumatol 1993^0: van Riel PLCM. Disease assessment and outcome in rheumatoid arthritis. Rheumatol Eur 1994; 23: Felson DT, Andersen JJ, Boers M et al The American College of Rheumatology preliminary core set of disease activity measures for the assessment of rheumatoid arthritis activity. Arthritis Rheum American College of Rheumatology Committee on Outcome Measures in Rheumatoid Arthritis Clinical Trials. Reduced joint counts in rheumatoid arthritis clinical trials. Arthritis Rheum : van der Heide A, Jacobs JW, Dinant HJ, Bijlsma JW. The impact of endpoint measures in rheumatoid arthritis clinical trials. Semin Arthritis Rheum 1992^1^ Goldsmith CH, Smythe HA, Helewa A. Interpretation and power of a pooled index. J Rheumatol 1993^0: Kaarela K. Prognostic factors and diagnostic criteria in early rheumatoid arthritis. Scand J Rheumatol 1985; 57(suppl.):l Hassell AB, Davis MJ, Fowler PD et al The relationship between serial measures of disease activity and outcome in rheumatoid arthritis. Q J Med 1993;86: van Zeben D, Hazes JM, Zwinderman AH, Vandenbroucke JP, Breedveld FC Factors predicting outcome of rheumatoid arthritis: results of a follow-up study. J Rheumatol 1993^0: Pincus T, Brooks RH, Callahan LF. Prediction of long-term mortality in patients with rheumatoid arthritis according to simple questionnaire and joint count measures. Ann Intern Med 1994; Houssien DA, Choy EHS, Berry H, Scott DL. Health status and joint damage divergence in rheumatoid arthritis. BrJRheumatol (suppl. l):107 (abstract).