Evaluation of the 1987 revised criteria for rheumatoid arthritis in a cohort of newly diagnosed female patients.код для вставкиСкачать
1042 ~ -~ ~ .- BRIEF REPORT EVALUATION OF THE 1987 REVISED CRITERIA FOR RHEUMATOID ARTHRITIS IN A COHORT OF NEWLY DIAGNOSED FEMALE PATIENTS CARIN E. DUGOWSON, J. LEE NELSON, and THOMAS D. KOEPSELL The revised criteria for the diagnosis of rheumatoid arthritis (RA) were presented in 1987. We tested these criteria on 135 women from a population-based study of patients with newly diagnosed RA. None of the 19 women with probable RA and 100 of the 116 women with definite RA met the 1987 criteria. The 1987 criteria appear to be less sensitive or more specific than the criteria formulated in 1956. In 1956, a committee of the American Rheumatism Association (ARA) proposed criteria for the diagnosis of rheumatoid arthritis (RA) ( I ) . These criteria were revised in 1958 (2). This classification system has been used to define RA for the last 30 years. Patients who met 3 or 4 criteria were diagnosed as having probable KA, those who met 5 or 6 criteria were diagnosed as having definite RA, and those who met 7 or more criteria were diagnosed as having classic RA. A list of multiple exclusions was provided. Compared with earlier studies, those studies conducted in the last 30 years using these criteria have provided useful information on the clinical status of the study subjects. Nevertheless, as time has passed, From the University of Washington and the Fred llutchinson Cancer Research Center, Seattle, Washington. Supported by NICHD contract NOI-HD-62914. Carin E. Dugowson. MD. MPH: Assistant Professor of Medicine and Adjunct Assistant Professor of Epidemiology, University of Washington: J. Lee Nelson. MD: Assistant Member, Associate in Clinical Research, Fred Hutchinson Cancer Research Center. and Clinical Assistant Professor, University of Washington; Thomas D. Koepsell, MD. MPH: Professor of Epidemiology, University of Washington. Address reprint requests to C a m E. Dugowson. MD, MPH. I124 Columbia Street. MP381, Seattle. WA 98104. Submitted for publication October 2. 1989; accepted in revised form January 25. 1990. Arthritis and Rheumatism, Vol. 33, No. 7 (July 1990) several shortcomings in these criteria have become apparent. Despite the absence of a “gold standard,” the consensus among rheumatologists has been that patients who met criteria for probable RA often seemed to have a different disease than those with classic or definite RA. In addition, the criteria involving synovial tissue or fluid samples were almost never used and, thus, rarely contributed to the diagnosis. Several more commonly diagnosed disorders. including seronegative spondylarthropathy and crystal deposition disorders, exclude patients who previously would have been diagnosed as having RA. New approaches to therapy and increased interest in the measurement of outcomes and prognosis in RA have made these shortcomings increasingly apparent. The ARA therefore mandated a revision of the 1958 criteria. The ncw 1987 criteria madc several changes (see Figure 1). The total number of criteria was decreased to 7; the presence of any 4 criteria now constitutes a diagnosis of RA. The long list of exclusions was deleted, and the use of levels of certainty was eliminated. One new item (3 or more swollen joints) and a revised item (arthritis of specified joints) complete the new list. This revision is an important change in our approach to the classification of RA. Criteria help define a disease and influence the findings of studies of etiology, pathogenesis, and therapy. ‘To help evaluate the likely impact of these new criteria, we applied them to a cohort of women with RA who had recently been diagnosed according to the 1958 criteria. Patients and methods. We are currently conducting a prospective, population-based case-control study of all women with newly diagnosed RA who live in King County, Washington (Seattle and surrounding 1043 BRIEF REPORTS 1958 Criteria 1987 Criteria Morning stiffness‘ Positive RF: Titer: Date observed: Figure 1. Comparison of 1958 and 1987 diagnostic criteria for the classification of rheumatoid arthritis (RA)Stippled areas indicate cnteria with no comparable item i n the other set of critena. a = specified cnteria present for at least 6 weeks; b = right or left proximal interphalangeal (PIP). metacarpophalangeal (MCP), wrist, elbow, knee, ankle. and metatarsophalangeal (MTP) joint RF = rheumatoid factor. area) or who receive medical care at Group Health Cooperative of Puget Sound, a large Seattle-based prepaid health plan organization. Women are referred to the study by all practicing rheumatologists, internists, and family practitioners in the area. Multiple methods are used to enhance completeness of ascertainment. The patients described in this report were the first 135 qualifying patients referred to this study. All were women, ages 18 to 64, who were either residing in King County, Washington or were members of Group Health Cooperative at the time of diagnosis. Their diagnosis date was no earlier than November 15. 1986. All women were examined by a board-certified rheumatologist (CED or JLN), and the findings of the joint examination were recorded in a standardized manner, keyed to the diagnostic criteria. Other information, including duration of morning stiffness, rheumatoid factor positivity and titer, duration of disease, and demographic information, was obtained directly from each patient and from her medical record. A feature was considered present if it was found either on examination by the study rheumatologist or on review of the patient’s ,medical record. Radiograph interpretation was taken from the radiologist’s report. Rheumatoid factor was assayed at the University of Washington clinical immunology laboratory using the Singer-Plotz latex agglutination method. The data were coded according to which of the 1958 and 1987 criteria for RA were met. We then compared the performance of the 2 criteria sets in this cohort. Results. Characteristics of the study subjects are summarized below. All patients were women, with a mean age of 44.6 years. Eighty-five percent of the patients were white, 6% were black, 4% were Asian, and 5% were of other races. The median duration of disease (the time between the onset of symptoms and the patient’s first physician visit) was 3.5 months. The number and percentage of women with each clinical feature of the 1958 criteria and the 1987 BRIEF REPORTS 1044 Table 1. Classification of rheumatoid arthritis (Rh) in 135 women by the 1958 and 1987 revised diagnostic criteria' Probable All patients RA (n = 135) (n = 19) 1958 criteria 1. Morning stiffness 2. Pain on motion or tenderness in at least I joint 3. Swelling of at least 1 joint 4. Swelling of at least I other joint 5 . Symmetric arthritis 6. Subcutaneous nodules 7. Radiographic changest 8. Serum rheumatoid factor 1987 criteria 1. Morning stiffness of 1 hour 2. Arthritis of 3 or more joint areas 3. Arthritis of the PIP. MCP. or wrist joints 4. Symmetric arthritis 5. Rheumatoid nodules 6. Serum rheumatoid factor 7. Radiographic changes in hand and/or wrist joint st Definite/ classic RA (n = 116) 134 (99) 14 (74) 18 (95) I14 (98) 116 (100) 135 (100) 19 (100) I16 (100) 130 (%) 14 (74) I I6 (100) 128 (95) 108 (80) 16 (12) 19 (29) 63 (47) 106 (91) 16 (14) 19 (29) 62 (53) 1 I5 (85) I05 (91) I05 (78) 101 (87) 130 (96) 116 (100) I08 (80) 16 (12) 63 (47) 106 (91) 16 (14) 62 (53) 19 (29) 19 (29) Values are the number (%). No patients had information on items 9, 10. or I I of the 1958 criteria (mucin clot. synovial biopsy. or nodule biopsy, respectively). PIP = proximal interphalangeal; MCP = metacarpophalangeal. t Erosions or periarticular osteopenia present. Percentages based on the 66 patients with documented radiographs. criteria in relation to their RA classification are shown in Table 1. Nineteen women met the 1958 criteria for probable RA. and 116 satisfied the criteria for definite/ classic RA. The 1958 criteria did not stipulate a minimum duration for morning stiffness. Accordingly. morning stiffness was coded as present by the 1958 criteria if this symptom was present at all. By this definition, morning stiffness was present in 98% of patients classified as having definite/classic RA and in 74% of those classified as having probable RA. Overall, 63 women (47%) had positive results on tests for rheumatoid factor, defined as a titer of z 1 :40. None of the women had information on items 9 (mucin clot), 10 (synovial histology). or 1 1 (nodule histology) of the 1958 criteria. Almost all women in both groups had 1 swollen joint and 1 tender joint; however, many more women with definiteklassic RA had symmetric arthritis (91%) and serum rheumatoid factor (53%) compared with those with probable R A (1 1% and 5%. respectively). None of the patients with probable RA had documented radiographic changes or subcutaneous nodules. Four of the 7 items in the 1987 criteria (symmetric arthritis, subcutaneous nodules, radiographic changes, and rheumatoid factor positivity) are identical to items in the 1958 criteria (Table I). The 20 women with morning stiffness lasting less than 1 hour were classified by the 1987 criteria as not having morning stiffness. The items retained from the 1958 criteria along with the 2 new items of the 1987 criteria were positive more often among patients with definite RA than those with probable RA. Simultaneous arthritis in at least 3 of the 14 specified joint areas (item 2) was satisfied by 105 (78%) of all patients. However, while 87% of patients with definite RA met this requirement, only 21% of those with probable RA did. Swelling of at least 1 joint of the hand (excluding the distal interphalangeal joints; item 31, was present in 100% of the patients with definite RA and 74% of the patients with probable RA. This item is very similar to item 3 of the 1958 criteria, but the number of eligible joints is lower. Overall, 74% (100 of 135) of the women meeting the 1958 criteria for probable, definite, or classic RA also met the 1987 criteria for RA. None of the 19 patients with probable RA satisfied the 1987 criteria. Of those patients who met the 1958 criteria for definite RA, 86% (I00 of 116) were classified as having RA by the 1987 criteria. Characteristics of the women who were classified as having definite R A by the 1958 criteria but were excluded by the 1987 revision are shown in Table 2. The features in this group are markedly different from those in the group of women who met the 1987 criteria. in terms of morning stiffness (69% versus 95%), arthritis in specifiedjoints (25% versus 97%). and symmetric arthritis (50% versus 97%). Nodules were not commonly present in patients of either group. Discussion. One of the main goals of the 1987 revision of the diagnostic criteria for RA was to improve the specificity of the 1958 criteria by making several items more restrictive. Thus, a patient with bilateral swollen, painful knees would meet 4 items in the 1958 criteria for probable RA, whereas the same patient would meet only 2 items in the 1987 criteria and would therefore not qualify as having RA. The criteria committee found the criteria to be 91.2% sensitive and 89.3% specific for RA when diagnoses of expert rheumatologists were used as a benchmark. This repre- 1045 BRIEF REPORTS Table 2. Analysis of women with discordant eligibility for the classification of rheumatoid arthritis (RA) by the 1958 and 1987 criteria* 1. Morning stiffness of I hour 2. Arthritis of 3 or more joint areas 3. Arthritis of the PIP. MCP. or wrist joints 4. Symmetric arthritis 5. Rheumatoid nodules 6. Serum rheumatoid factor 7. Radiographic changes in hand and/or wrist jointst Eligible by 1987 criteria (n = 100) Ineligible by 1987 criteria (n = 16) 95 (95) 97 (97) II (69) 4 (25) 100 (100) 16 (100) 97 (97) 16 (16) 54 (54) 31 (47) 8 (50) 0 9 (56) 0 Values are the number (%). All women were classified as having definite RA by the 1958 criteria. Sixteen of these women did not meet the 1987 criteria for RA. PIP proximal interphalangeal: MCP = metacarpophalangeal. t Erosions or periarticular osteopenia present. Percentages based on the 66 patients with documented radiographs. sented an improvement in both characteristics. but particularly the specificity, over the 1958 criteria (3). Because the present study did not apply either the old or the new criteria to persons who were known not to have RA, numerical estimates of specificity cannot be directly obtained from our data. However, useful information can be gained by examining those study patients for whom the 1958 and 1987 criteria yielded different classifications. All 19 patients who had "probable" R A by the 1958 criteria failed to meet the 1987 criteria for RA. The label "probable" RA may be a misnomer, because the followup of such patients suggests that most of them probably do not have RA (4). If this were the case for the 19 patients with "probable" RA in the present study, then at least some of them would be false-positive by the 1958 criteria but true-negative by the 1987 criteria, suggesting that the 1987 criteria are more specific. Improving specificity can sometimes be achieved only by sacrificing sensitivity, however. In this regard, the 16 women in the present study who had "definite" or "classic" RA by the I958 criteria but were classified as not having RA by the 1987 criteria are of particular interest (Table 2). Most of these patients were referred to the study by rheumatologists. Treatment information was available on 1 1 of these 16 women. Seven of them were being treated with prednisone and/or a remittive agent, suggesting that the referring physician was sufficiently convinced of the clinical diagnosis of RA to have initiated therapy. It is possible that at least some of these 16 women may have had RA and were thus false-negative on the 1987 criteria. This. in turn, would suggest a possible problem with the sensitivity of the revised criteria. It is important to note that all women in the present study had newly diagnosed RA. with a median disease duration of 3.5 months. In contrast, most patients studied by the AKA rheumatoid arthritis subcommittee for the 1987 criteria had established disease, with a median disease duration of 7.7 years. The subcommittee conducted a subanalysis on patients with disease of less than l year's duration and obtained an estimated sensitivity of 81%. which was lower than the 91% obtained for all their patients. These findings support the concerns expressed by Silman ( 5 ) in a recent commentary about possible shortcomings in the 1987 criteria with regard to identification of patients with early disease. The rheumatoid factor test can be negative in the early stages of RA. while the likelihood of a positive result has been shown to increase during the years following diagnosis (6). Thus, a patient with early disease is less likely to meet 1 of the 4 criteria needed to qualify for a diagnosis of RA. Even among our patients with "definite" or "classic" RA by the 1958 criteria. only 53% had a positive rheumatoid factor test result. In addition, RA typically affects an increasing number of joints over time, making it easier for a patient with longstanding disease to satisfy item 2 (arthritis in at least 3 qualifying joint areas) and item 4 (symmetry) in the 1987 criteria. Any comparison of alternative diagnostic criteria for RA is hampered by the lack of a universally accepted gold standard for diagnosis; without this, it may be particularly worthwhile to conduct careful followup studies of those patients whose diagnostic classification differs between the 1958 and the 1987 criteria. If these patients develop more clinical features of RA and eventually meet both sets of criteria. then concerns about the possible insensitivity of the 1987 criteria for early disease would be well founded. If, on the other hand, these patients have an evanescent syndrome and a benign long-term prognosis. such findings would providc reassurance that the restrictions built into the revised criteria have served chiefly to improve the specificity for clinically significant RA. Distinguishing between these 2 possibilities may be especially important in deciding whether such patients are to be included in studies of early remittive therapy. The present study suggests that the 1987 criteria BRIEF REPORTS are more restrictive than the 1958 criteria, and this may soon affect the interpretation of findings from epidemiologic investigations of RA. In monitoring future trends in the incidence and prevalence of RA. we will need to be cautious to distinguish between a true decline in disease frequency and artifactual changes due to a revision in the operational definition of RA. For this reason alone, it may be useful in future studies to gather sufficient data to allow both the 1958 and the 1987 criteria to be applied. Acknowledgments. We thank Dr. Mart Mannik and Dr. Bruce Gilliland for their review of the manuscript and Diane Rosman for design of the coding form. 3. 4. 5. REFERENCES 1. Ropes MW. Bennett GA, Cobb S, Jacox R, Jessar RA: Proposed diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 7:121-124, 1956 2. Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA: 6. 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958 Arnett FC, Edworthy SM, Bloch DA, McShane DJ. Fries JF. Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS, Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller JG, Sharp JT, Wilder RL, Hunder GG: The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988 O'Sullivan JB, Cathcart ES: The prevalence of rheumatoid arthritis: follow-up evaluation of the effect of criteria on rates in Sudbury, Massachusetts. Ann Intern Med 76:573-577, 1972 Silman AJ: The 1987 revised American Rheumatism Association criteria for rheumatoid arthritis. Br J Rheumatol 27:341-343, 1988 Jacoby RK. Jason MIV. Cosh JA: Onset, early stages, and prognosis of rheumatoid arthritis: a clinical study of 100 patients with 11-year follow-up. Br Med J 2:96-100, 1973 Erratum It has recently been brought to our attention that, in an article by Tate et al published in Arthritis and Rheumatism (Tate GA, Mandell BF, Karmali RA, Laposata M, Baker DG, Schumacher HR Jr, Zurier RB: Suppression of monosodium urate crystal-induced acute inflammation by diets enriched with gamma-linolenic acid and eicosapentaenoic acid. Arthritis Rheum 31 :1543-1551, 1988), one of the sources of grant support was not listed. Dr. Brian Mandell's work was supported by a grant from the Eastern Pennsylvania Chapter of the Arthritis Foundation.