Disease Activity Score for Children With Juvenile Dermatomyositis: Reliability and Validity Evidence

Transcription

1 Arthritis & Rheumatism (Arthritis Care & Research) Vol. 49, No. 1, February 15, 2003, pp 7 15 DOI /art , American College of Rheumatology ORIGINAL ARTICLE Disease Activity Score for Children With Juvenile Dermatomyositis: Reliability and Validity Evidence RITA K. BODE, 1 MARISA S. KLEIN-GITELMAN, 2 MICHAEL L. MILLER, 2 TOULA S. LECHMAN, 3 AND LAUREN M. PACHMAN 2 Objective. To assess the reliability and validity of the Disease Activity Score (DAS), an instrument used to evaluate children with juvenile dermatomyositis (JDM). Methods. Psychometric study of internal consistency, reliability, rater agreement, and the relationship with measures of muscle strength and disability was conducted. Results. The DAS ratings are internally consistent (reliability 0.89) and describe a wide range of disease activity. The pediatric rheumatologists in this study agree on the presence of most of the disease indicators. Their disagreements tend to cancel each other, resulting in highly correlated (r 0.79) overall measures across raters. Estimates of muscle weakness using the DAS and ratings of muscle strength obtained independently from therapists are highly related (r 0.77), but estimates of disease activity and disability are weakly related (r 0.20). Conclusion. The DAS exhibits evidence of good reliability and validity. The combination of skin and muscle strength assessments makes this easily administered instrument a useful addition in the evaluation of children with JDM. KEY WORDS. Juvenile dermatomyositis; Disease activity; Reliability; Validity. INTRODUCTION 1 Rita K. Bode, PhD: Northwestern University Medical School, Chicago, Illinois; 2 Marisa S. Klein-Gitelman, MD, Michael L. Miller, MD, Lauren M. Pachman, MD: Northwestern University Medical School and Children s Memorial Hospital, Chicago, Illinois; 3 Toula S. Lechman, MBA: Children s Memorial Hospital, Chicago, Illinois. Address correspondence to Rita K. Bode, PhD, Rehabilitation Institute of Chicago, 345 East Superior Street, Chicago, Illinois Submitted for publication August 28, 2001; accepted in revised form April 29, Juvenile dermatomyositis (JDM) is the most common of the pediatric inflammatory myopathies, with an annual incidence of 3.2 per 1 million children in the US (1) and a mean age at diagnosis of 6.3 years (SD 3.4 years) (2,3). There are few validated assessments of either clinical (4) or laboratory-based (5) data that reliably evaluate the disease activity in children diagnosed with JDM. The assessments that are available focus on muscle strength and do not quantify levels of cutaneous involvement. Because the measurement of muscle strength is problematic in children who are younger than 4 years, such assessments are typically used for examinations of children who are older. As a result of this dearth of information, recent publications have focused on identifying the important variables in assessing disease activity as well as tissue damage (6 9). Recently, we have documented that the cutaneous manifestations in JDM are strongly associated with evidence of systemic derangement of blood vessel morphology, thus strengthening the importance of the inclusion of the extent and severity of skin involvement in the assessment of disease activity (10). Several laboratory-based methods are now available as markers for disease activity, but no single test has yet proven to be a reliable indicator (5,8,11). Thus, there is an urgent need for a disease activity assessment that can measure more than muscle strength and be used to evaluate children of all ages who are diagnosed with JDM. The Juvenile Myositis Clinic at Northwestern University Medical School s Children s Memorial Hospital (Chicago, Illinois) has evaluated more than 210 children with JDM during the past 2 decades. In this time, a rapid assessment of each child s clinical status has evolved and been described as a measure of disease activity (to differentiate it from disease chronicity, which might be measured by indicators such as range of motion loss or calcifications). This assessment, the Disease Activity Score (DAS), is performed by a single physician during a physical examination and has been shown to be highly related to sensitive indicators of immune-mediated vascular damage (5,12 18). Also, the DAS reflects clinical and immunologic activation, even when the usual labo- 7

2 8 Bode et al ratory indicators of disease activity (e.g., serum levels of muscle-derived enzymes) are within normal ranges. For example, an increased percent circulating CD19 B cells are associated with an increased DAS (12,13) as well as other indicators of inflammation (3,14 19). To date, the DAS has mainly been used by experienced rheumatologists who work in the Division of Immunology/Rheumatology at Northwestern s Children s Memorial Hospital. The purpose of this study was to determine if this rapid assessment might be useful to physicians, especially rheumatologists, who might see few children with JDM. Practitioners who are not familiar with JDM and who wish to evaluate disease activity using the DAS, both initially and after therapy has been started, need to be aware of specific disease indicators and be able to recognize the presence of those indicators. For such an assessment to be useful in detecting JDM activity, different physicians would have to come to the same conclusion regarding the presence of the indicators. To evaluate whether the DAS provides consistent information, inter-rater agreement and other evidence of reliability are needed. For a new instrument to be useful in clinical applications, it is also important to know how the information that the tool provides compares with other signs of the disease and/or its consequences. An indicator of disease activity that is typically part of the physical examination of children with JDM is an overall rating of muscle weakness. Similarity of muscle strength ratings by rheumatologists using this instrument and results of physical examination by experienced occupational and physical therapists (OT/PT) would provide evidence of the validity of this disease activity measure. Furthermore, if the examiner perceives disease activity to be different than general disability, it would be expected that measures of disability and disease activity would not be correlated. The objectives of our study were to determine the psychometric characteristics of the Disease Activity Score, namely: 1) its internal consistency, 2) reliability of DAS measures across raters, 3) identification of JDM indicators that may need enhanced descriptions or characterization, and 4) validity of DAS measures relative to other indicators of disease consequences. The assumptions implicit in these objectives include the following: 1) all variables comprising the DAS work together to describe clinical evidence of disease activity; 2) the ordering of items from most- to least-commonly observed matches clinical experience with the illness; 3) the overall ratings will differentiate distinct strata of disease activity into which each evaluated child can be reliably classified; 4) measures of muscle strength based on physician ratings will be at least moderately correlated to ratings provided by OT/PT; and 5) the overall relationship between measures of disease activity and general disability two distinct items will be low to moderate. PATIENTS AND METHODS Patient population. The convenience sample consisted of 44 children who attended a pediatric rheumatology clinic at Children s Memorial Hospital (Chicago, Illinois), with a total of 58 visits among them. These children were previously diagnosed with definite JDM, according to the criteria of Bohan and Peter (20). The sample was predominately white (88%) and female (74%), which reflected national data for this disease (2). Mean age at the time of visit was 11.3 years (range years). In this sample, 2 children were younger than or equal to 4 years of age at the time of evaluation. Informed consent was obtained for the evaluation of the patient and laboratory data. Instrument. The DAS was developed to quantify disease presence and progression over time. We have used it to compare clinical and laboratory indicators of JDM presence, even though the tool requires no knowledge of laboratory results (12 18). It was selected from a field of contenders at a consensus conference of the Pediatric Rheumatology International Trials Organization in Pavia, Italy (March 29 April 4, 2001). The directions and the evaluation sheet for initial assessment of disease activity and response to therapy in international studies of drug efficacy in children with JDM are presented in Appendix A. The DAS consists of 19 items, resulting in a score of 0 20: 10 items are scored dichotomously (the indicator is present or not), and 3 polychotomously (rating severity levels or extent to which the indicator is present). The equal contribution of items relating to muscle and skin reflects their equal importance in the disease pathophysiology. The presence or absence of weakness is assessed via 8 variables (each are counted), specifically the following: neck flexor muscles, abdominal muscles (doing a sit up and clearing one s scapula), upper extremity proximal muscles, lower extremity proximal muscles, Gower s sign, abnormal gait, difficulty swallowing, and nasal speech. The presence or absence of vasculitis is assessed by determining the presence of any 1 of the following: eyelid erythema, eyelid vessel dilation, eyelid thrombosis, nailfold erythema, nailbed telangiectasia, dilation of blood vessels on the palate, and other vasculitis (such as patches of persistent vascular dilation on the limbs). The other category can be used to identify the consequence of active vasculitis, ulceration. The child s overall functional status and the presence of rashes are rated using polychotomous scales: functional status comprises a 4-point scale, ranging from normal function to severe limitations; the distribution of the involved skin is rated on a 4-point scale, ranging from none to generalized, while the severity of the skin involvement is rated on a 5-point scale, ranging from absent to severe. Gottron s papules are rated on a 4-point scale, ranging from absent to severe, and include evidence of atrophic lesions (which usually disappear entirely, but can sometimes flare ). Other assessments were conducted as part of the routine clinical evaluation. The Childhood Health Assessment Questionnaire (CHAQ) (21,22) was completed by a parent of each child in this sample. The CHAQ is an established outcome measure that has been validated for children with JDM (8,9). Testing by the team s OT/PT, a component of standard medical care, follows manual muscle testing (MMT) guidelines (23,24) and was performed at each clinic visit.

3 Analysis of Disease Activity Score 9 Procedure. Three board-certified pediatric rheumatologists evaluated disease activity in this sample. Two of these physicians conducted independent physical examinations of each child, rating the presence (or extent/severity) of each disease activity indicator. Physician A evaluated the first 29 patients and, during the same visit, physician B examined each patient on her own. A few months later, physician B examined another 29 patients; similarly, during the same visit, physician C rated each patient on his own. The order of the physician examiners was varied to avoid patient fatigue, but the MMT was always performed before the physician examination. Analysis. The psychometric characteristics of the DAS were evaluated using Rasch analysis, which is based on a mathematical model developed by Georg Rasch (25). Since the early 1980s, it has been applied widely in the human sciences (26); specifically in the construction and validation of instruments in neurology (27), physical medicine and rehabilitation (28 30), pain (31), rheumatology (32), endocrinology (33), and geriatrics (34). The Rasch model for multiple rating scales was applied to the ratings in the DAS, the CHAQ, and the OT/PT muscle strength ratings to obtain interval measures for these instruments. The psychometric properties evaluated include 1) the extent to which the measures can distinguish persons with different levels of disease activity, and 2) the extent to which the items all measure a single theoretical construct. To evaluate the first property, the range of disease activity measures was compared with the error in estimating them (called person separation), conceptually similar to a t-test in which mean differences are compared with their standard error. The criterion for significant person separation is 2.00 or better. To evaluate the second property, the extent to which individual ratings deviated from expectation (called misfit) was used as evidence of validity. The criterion for item misfit is an infit mean square statistic (MnSq) greater than Additional evidence of validity was obtained through a comparison of the empirical ordering of items to the expected ordering of items based on clinical experience and previous research. The units of measure used in the analyses (called measures) are transformed to integer scales corresponding to the instrument s raw score range to improve their interpretability. Comparisons were made of the measures obtained for the DAS (total and separate skin and weakness components), the CHAQ, and therapist ratings of muscle strength to provide further evidence of validity. The overall measures for DAS and CHAQ were correlated using Pearson s coefficient, and the relationship between DAS measures and CHAQ scores was illustrated using a scatter plot. The DAS weakness measures were also correlated with the OT/PT muscle strength measures using Pearson s coefficient. To examine rater effects, an evaluation for which Rasch methodology is particularly useful, 2 sets of analyses were performed: 1 to identify individual items on which raters were differentially sensitive and 1 to estimate overall disease activity, controlling for differences in rater sensitivity. In addition to the use of Rasch methodology, traditional analyses were also conducted and both sets of results are reported. Rating scale analysis (35) using the Winsteps (36) computer program was conducted to identify individual items on which physicians were differentially sensitive. In this calibration, disease activity was estimated separately by each physician. To examine differences in rater sensitivity, the item calibrations from physician B were compared with the calibrations from physicians A and C to identify items on which there was disagreement regarding whether or not, or the extent to which, the indicator was present. If the physicians agreed on the presence of these indicators, the difference between the 2 estimates would be less than the combined errors of measurement; if the physicians disagreed on their presence, the difference would be greater than their combined errors. Inter-rater reliability was also examined using kappa coefficients; the criteria for agreement used in this study was a coefficient greater than 0.40 for good agreement and a coefficient of 0.20 for marginal agreement. A many-faceted rating scale analysis (37) using the Facets (38) computer program was conducted to control for rater effects in the estimation of overall disease activity. As part of this analysis, a chi-square test was conducted to determine whether the physicians significantly differed in the ratings they provided for the group of patients as a whole. The relationship between the separate estimates of disease activity was examined by comparing the difference between the ratings of the clinicians to their combined error of measurement, to identify patients for whom the 2 estimates of disease activity were significantly different. The relationship of overall measures for physician B (the overlapper) versus physicians A and C was also examined using Pearson correlation; the criterion for a strong relationship is a correlation coefficient greater than RESULTS Scales Table 1. Calibration results* Scale means (SD) Scale ranges Separation index DAS total score (4.53) DAS skin scale 4.80 (2.46) DAS weakness scale 3.86 (3.51) CHAQ 6.80 (8.40) CHAQ score 0.16 (0.44) 0 3 OT/PT muscle strength ratings (8.85) * DAS disease activity score; CHAQ Childhood Health Assessment Questionnaire; OT/PT occupational and physical therapists. Rasch units of measurement transformed to raw score range. Indicator of scale reliability. The psychometric characteristics of the DAS, CHAQ, and OT/PT muscle strength ratings are described below and the summary statistics (means, standard deviations, and score ranges) for each of the scales are presented in Table 1.

4 10 Bode et al Figure 1. Comparison of disease activity score measures and CHAQ scores. Symbols reflect the density of cases at each plot point. 1 case; 2 cases; 3 cases, etc. Reliability. Table 1 also contains the reliability information for these instruments. The set of indicators in the DAS produces a reliable estimate of disease activity (person separation 2.80 as compared with the criterion of 2.00) that distinguishes at least 3 distinct strata of disease activity in the sample: high, average, and low. The separate skin and weakness measures were less reliable, suggesting that both components are needed to adequately measure disease activity. The set of OT/PT ratings of muscle strength produces measures with excellent reliability that distinguishes 7 distinct strata of patients in the sample. However, for this sample, the CHAQ produces measures with a less-than-acceptable level of reliability that barely distinguishes 2 strata in the sample. Approximately half of the sample reported no difficulty at all in performing the CHAQ tasks (floor effect), making it difficult to discriminate distinct levels of function at the mild end of the disability continuum, which is seen in Figure 1. Validity. The fit of the DAS items to the disease activity construct is within acceptable levels (infit MnSq values 1.30). Vascular inflammations (redness, vessel dilation, and thrombosis) present clinically and are observed more frequently in this study than other disease indicators. Some musculoskeletal symptoms (difficulty swallowing and abnormal gait) are infrequent both clinically and in this sample; others (upper and lower proximal muscle, abdominal weakness, and nasal speech) fall in the middle range of frequency. The most frequently observed indicators of JDM represent a mild condition and may be observable with a relatively low level of disease activity, and the least frequently observed indicators represent a more severe condition or may only be observed in more severe cases. The fit statistics indicate that the DAS items all measure a single theoretical construct and, combined with the comparison of the empirical and clinically expected item ordering, provide evidence that the construct being measured is disease activity. The fit of CHAQ items to a single construct (disability) was less acceptable; 4 items misfit ( cut finger/toe nails, do household chores, run errands and shop, and open a new cereal box ), suggesting the existence of a second, perhaps psychomotor, construct in the item set. Similarly, the fit of the OT/PT ratings to a single construct (muscle strength) was also less acceptable; several items (hamstrings, gluteus medius muscles, adductors, and gastroc/soleus muscles) misfit, suggesting that manual muscle testing is less sensitive to lower extremity strength in this sample. If the purpose of this study had been to improve the measurement of disability or muscle strength, some of these items would be candidates for deletion. However, because the purpose was to compare the DAS measures with existing measures of JDM disease consequences, the existing measures were used as is. Additional evidence of validity is seen in the comparison of measures based on the DAS weakness items and OT/PT ratings of muscle strength, and a comparison of disease activity and disability. As expected, the relationship between measures of muscle strength and weakness is strong and negative (r 0.774) with more strength (as rated by therapists) being highly associated with less weakness (as rated by physicians). As expected, the relationship between measures of disease activity and disability is weak (r 0.20). The comparison of the DAS and CHAQ measures is illustrated in Figure 1. This plot shows a general trend toward agreement but a clustering of patients at the minimum extreme CHAQ measure. Figure 1 further shows that patients with minimum disability measures vary in disease activity as measured by DAS. These results confirm the floor effect in the CHAQ and suggest

5 Analysis of Disease Activity Score 11 that it contributes to the overall weak relationship between these 2 measures. Differential sensitivity. Although the ratings across items were internally consistent, differences in practitioner sensitivity to individual disease activity indicators were found. Across both pairs of physicians, 8 items were rated similarly, indicating that the physicians agreed on the items that record the presence of muscle weakness. The items include sit up with scapular clearance, upper extremity proximal muscles, lower extremity proximal muscles, Gower s sign, abnormal gait, and difficulty swallowing. The physicians also agreed on 2 items measuring skin abnormalities: nailfold erythema and nailbed telangiectasia. Finally, there is partial agreement (either physician A or physician C agrees with physician B) on 7 items: functional status, neck flexor weakness, nasal speech, eyelid erythema and vessel dilation, Gottron s papules, and other vasculitis. However, the ratings on 4 items were consistently different, indicating that the physicians disagree on the presence of those indicators of disease activity. Of these, physician B is more sensitive than either physician A or C on 2 skin indicators (eyelid thrombosis and palate dilation) and both physicians B and C are more sensitive than physician A on 2 other skin indicators (rash involvement and distribution). These differences suggest the need for more detailed descriptions of these indicators so that, regardless of training and experience, practitioners would be equally sensitive to the indicators presence. These results are confirmed by inter-rater reliabilities, as estimated by kappa coefficients. Using cutoffs of 0.40 and 0.20 to identify good and marginal agreement, respectively, 6 of the items for which coefficients could be estimated indicated good agreement, 6 indicated marginal agreement, and 4 indicated poor agreement. For all of the items identified as having good agreement using the results of the Rasch analysis, the kappa coefficients are at least marginal; however, for 2 of the items identified as not having good agreement in the Rasch analysis (eyelid thrombosis and erythema), the kappa coefficients are more than marginal. Rater effects. Differential rater sensitivity to the presence of these indicators appears to have had an effect on the overall estimate of disease activity for this sample. When raters are included in the Rasch analysis of the item responses, their effect is statistically significant ( 2 [1 degree of freedom] 6.7, P 0.03); that is, physician B rated the disease activity in some patients as higher than the other 2 physicians. For the vast majority of cases (approximately 80%), however, the estimated disease activity measures were essentially the same across physicians. This result was confirmed by a Pearson correlation of 0.79 between the 2 estimates of disease activity for each patient. In 10 cases, physician B rated disease activity higher than physicians A and C, and in 3 cases, physicians A and C rated disease activity higher than physician B. If the differential sensitivity of these physicians to the indicators had cancelled out, rater effects would have been of no consequence in the estimation of disease activity. However, because the differences in sensitivity were imbalanced (of the indicators on which the physicians disagreed, physician B was more sensitive on more items than were physicians A and C), the level of disease activity was estimated by physician B as higher for some patients. DISCUSSION The results of this study suggest that the rapid assessment of clinical indicators of JDM using the DAS may be useful in both assessing newly developed laboratory indicators of disease activity and evaluating a child s response to therapy. None of the evaluation instruments presented to date have included children under the age of 4, and none of them have scored cutaneous and muscle findings in addition to function (7 9). Most clinical assessment tools used to measure disease activity in idiopathic inflammatory myopathies are based on assessments of muscle strength (3). The DAS, however, combines the assessment of muscle strength and skin condition to arrive at a single estimate that is more sensitive to disease activity than measures of muscle strength alone. The advantage of this tool is the availability of skin for assessment regardless of the age of a child. Inclusion of skin assessments is important because of the early manifestation of these symptoms, which may precede evidence of muscle weakness. Cutaneous manifestations are the first definite symptom in 25% of JDM cases, occur in concert with muscle weakness in another 50%, and follow muscle weakness as the other definite symptom by 3 months in the remaining 25% of cases (3). Inclusion of these assessments over the course of JDM is also important as an indicator of disease activity. Severity of skin findings has been found to be directly associated with drop out of capillary loops in nailfold capillary studies (10) and to be an indicator of the hallmark of the disease, occlusion of small vessels (39). Because the assessment of muscle strength is problematic for children younger than 4 years of age, the DAS holds special promise for the measurement of disease activity in younger patients. Difficulty in accurately assessing disease activity in this population is not a minor problem. In a national study of newly diagnosed children with JDM, 18% were 4 years of age or younger at diagnosis (1). Thus, an instrument that can be used with younger patients is of great importance. Assessments of skin condition can be done at any age and can be combined with assessments of strength in selected muscles where compliance is not an issue. This study represents a step in the process of developing a set of indicators that appears to be applicable to children of all ages who have this condition. Inasmuch as our study tested a convenience sample of 2 sets of consecutive clinic visits, the number of younger children included is small; only 2 (3%) of the children were 4 years of age or younger. However, a review of the records of 185 children in the database at this facility identified 25 children with JDM who were 4 years of age or younger at diagnosis for whom the DAS was used on initial examination, prior to the initiation of this study. A calibration of their data showed more than acceptable reliability (person separation 1.95;

6 12 Bode et al separation reliability 0.79), suggesting that the DAS can be applied to this younger age range, even though compliance with muscle testing is variable, and CHAQ items may be difficult to interpret. Although the DAS shows great promise in the measurement of disease activity for children with JDM, the descriptions of individual disease indicators on which these physicians did not agree need to be clarified. Additional guidelines for physicians are now in place (International Study of Children with Juvenile Dermatomyositis), but future studies are needed to determine the success of these modifications. In developing these guidelines, the results of this study were carefully examined to identify the potential source of disagreement across these 3 physicians. Several explanations were thought to account for the disagreements. Familiarity with the patients and the range of disease variation may have made 1 physician more sensitive to their disease activity. Also, differential sensitivity related to a particular physician s experience with certain indicators (such as eyelid thrombosis and dilation of vessels at the junction of the soft and hard palate) may have been associated with varied sensitivity between physicians. Regardless of the specific explanation, if lack of agreement is related to the amount of experience a physician has with this condition, it is imperative that the descriptions of all the indicators be made explicit in terms of what to look for during the physical examination. That way, regardless of amount or kind of experience, rheumatologists should be able to detect JDM disease activity in a consistent and equivalent manner. In conclusion, we have presented a tool that can be used clinically to rate the disease activity in a child of any age with JDM, and have examined its psychometric characteristics and rater effects. The set of indicators in the DAS reliably measures a single construct of disease activity, and its relationship with other health assessments is as hypothesized. The individual indicators chosen to measure disease activity in children with JDM matches clinical expectations of the frequency with which these indicators are observed. It appears that the DAS is more appropriate for measuring disease activity than another widely used disability instrument that is recognized to be insensitive at low levels of JDM disease activity. It also appears that the physician ratings of muscle strength agree with other muscle assessments, specifically those by occupational and physical therapists. Although physicians using the DAS essentially agreed on the overall level of disease activity, they disagreed on the presence of a few individual indicators. This information is now being used to more completely describe indicators on which practitioners were differentially sensitive, to improve the instrument s interrater reliability in an international study of JDM (40). ACKNOWLEDGMENTS The authors wish to acknowledge the contribution of the therapists who performed the manual muscle testing, Kate Ricardi and Karen Erickson, and other colleagues at Childrens Memorial Hospital who reviewed the manuscript and provided many useful suggestions. REFERENCES 1. Mendez E, Lipton RB, Dyer A, Ramsey-Goldman R, Roftcher P, Bowyer S, et al. Incidence of JDM from the NIAMS Registry. Arthritis Rheum (Arthritis Care Res). In press. 2. Pachman LM, Mendez E, Lou H, Chung A, Davis V, Ramsey- Goldman R, et al. At diagnosis, African American children with juvenile dermatomyositis (JDM) may have a different presentation than Caucasian or Hispanic children: physician reported data from NIAMS JDM research registry [abstract]. Arthritis Rheum 1999;42 Suppl 9:S Pachman LM, Hayford JR, Chung A, Daugherty, Pallansch MA, Fink CW, et al. Juvenile dermatomyositis at diagnosis: clinical characteristics of 79 children. J Rheumatol 1998;25: Rider LG. Assessment of disease activity and its sequelae in children and adults with myositis. Curr Opin Rheumatol 1996;8: Pachman LM. Imperfect indication of disease activity in juvenile dermatomyositis. J Rheumatol 1995;22: Rider LG, Feldman BM, Perez MD, Rennebohm RM, Lindsley CB, Zemel LS, et al. Development of validated disease activity and damage indices for juvenile idiopathic inflammatory myopathies: I. Physician, parent and patients global assessments. Arthritis Rheum 1997;40: Lovell DJ, Lindsley CB, Rennebohm RM, Ballinger S, Bowyer SL, Feldman BM, et al. Development of validated disease activity and damage indices for the juvenile idiopathic inflammatory myopathies: II. The childhood myositis assessment scale (CMAS): a quantitative tool for the evaluation of muscle involvement. Arthritis Rheum 1998;42: Feldman BM, Ayling-Campos A, Luy L, Stevens D, Silverman ED, Laxer RM. Measuring disability in juvenile dermatomyositis: validity of the childhood health assessment questionnaire. J Rheumatol 1995;22: Huber AM, Hicks JE, Lachenbruch PA, Perez MD, Zemel LS, Rennebohm RM, et al. Validation of the childhood health assessment questionnaire in the juvenile idiopathic myopathies. Arthritis Rheum 2000;43: Pachman LM, Mendez E, Lechman T, Sundberg J, Shamiyeh E, Dyer A. The rash of juvenile dermatomyositis (JDM) is associated with derangement of capillaries. Arthritis Rheum 2000;43 Suppl 9:S Rider LG, Miller FW. Laboratory evaluation of the inflammatory myopathies. Clin Diag Lab Immunol 1995;2: Eisenstein DM, O Gorman MR, Pachman LM. Correlations between change in disease activity and changes in peripheral blood lymphocyte subsets in patients with juvenile dermatomyositis. J Rheumatol 1997;24: O Gorman MRG, Corrochano V, Roleck J, Donovan M, Pachman LM. Flow cytometric analysis of the lymphocyte subsets in peripheral blood of children with untreated active juvenile dermatomyositis. Clin Diag Lab Immunol 1995;2: Pachman LM. Polymyositis and dermatomyositis in children. In: Maddison PJ, Isenberg DA, Woo P, Glass DN, editors. Oxford textbook of rheumatology, 2nd ed. Oxford, UK: Oxford University Press; p Pachman LM. Juvenile dermatomyositis: new clues to diagnosis and pathogenesis. Clin Exp Rheumatol 1994;12: Pachman LM. Pathophysiology and disease expression. Pediatr Clin North Am 1995;42: Pachman LM, Maduzia L, Chung A, Donovan M, Ramsey- Goldman M. Juvenile dermatomyositis (JDMS): disease activity scores are correlated with levels of neopterin in serum [abstract]. Arthritis Rheum 1995;38 Suppl 9:R Pachman LM, Sundberg MJ, Maduzia L, Daugherty C, Litt D. Sequential studies of nailfold capillary vessels (NFC) in 10 children with juvenile dermatomyositis (JDMS): correlation with disease activity score (DAS) but not von Willebrand factor antigen (vwf:ag) [abstract]. Arthritis Rheum 1995;38 Suppl 9:S Guzman J, Petty RE, Malleson PN. Monitoring disease activity in juvenile dermatomyositis: the role of von Willebrand factor and muscle enzymes. J Rheumatol 1994;21:

7 Analysis of Disease Activity Score Bohan A, Peter JB. Polymyositis and dermatomyositis (parts 1 and 2). N Engl J Med 1975;292:344 7, Fries JF, Spitz P, Kraines RG, Holman HR. Measurement of patient outcome in arthritis. Arthritis Rheum 1980;23: Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of health status in children with juvenile rheumatoid arthritis. Arthritis Rheum 1994;37: Kendall F, McCreary EK, Provance PG. Muscles: testing and function. Baltimore: Williams & Willkins; Resnick JS, Mammel M, Mundale MO, Kottke FJ. Muscle strength as an index of response to therapy in childhood dermatomyositis. Arch Phys Med Rehabil 1981;62: Rasch G. Probabilistic models for some intelligence and attainment tests. Copenhagen: Danish Institute for Educational Research; (Expanded edition, Chicago: The University of Chicago Press; 1980). 26. Bond T, Fox C. Applying the Rasch model: fundamental measurement in the human sciences. Mahwah, NJ: Lawrence Erlbaum Associates; Tsuiji T, Liu M, Sonoda S, Domen K, Chino N. The stroke impairment assessment set: its internal consistency and predictive validity. Arch Phys Med Rehabil 2000;81: Silverstein BJ, Fisher WP Jr, Kilgore KM, Harvey RF, Harley JP. Applying psychometric criteria to functional assessment in medical rehabilitation: I. Defining interval measures. Arch Phys Med Rehabil 1992;73: Grimby G, Andren E, Holmgren E, Wright B, Linacre JM, Sundh V. Structure of a combination of functional independence and instrumental activity items in community living persons: a study of individuals with cerebral palsy and spina bifida. Arch Phys Med Rehabil 1996;77: Granger C, Linn R. Biologic patterns of disability. J Outcome Measure 2000;4: Tesio L, Granger CV, Fiedler RC. A unidimensional pain/ disability measure for low-back pain syndromes. Pain 1997; 69: Wolfe F, Kong S. Rasch analysis of the Western Ontario Mac- Master Questionnaire (WOMAC) in 2205 patients with osteoarthritis, rheumatoid arthritis, and fibromyalgia. Ann Rheum Dis 1999;58: Wiren L, Whalley D, McKenna S, Wilhelmsen L. Application of a disease-specific, quality of life measure (QOL-ACHDA) in growth hormone-deficient adults and a random population sample in Sweden: validation of the measure by Rasch analysis. Clin Endocrin 2000;52: Doble S, Fisher AG. The dimensionality and validity of the Older Americans Resources and Services (OARS) Activities of Daily Living (ADL) Scale. J Outcome Measure 1998;2: Wright BD, Masters GN. Rating scale analysis: Rasch measurement. Chicago: MESA Press; Wright BD, Linacre JM. Winsteps: Rasch analysis computer program. Chicago: MESA Press; Linacre JM. Many-facet Rasch measurement. Chicago: MESA Press; Linacre JM, Wright BD. Facets: Rasch analysis computer program. Chicago: MESA Press; Banker BQ. Dermatomyositis of childhood. J Neuropathol Exp Neurolol 1975;34: The Pediatric Rheumatology International Trials Organization. International Consensus Conference to Define Core Sets of Outcome Measures for Juvenile Systemic Lupus Erythematosus and Juvenile Dermatomyositis. Preliminary data. Pavia Italy, April 2001.

8 14 Bode et al APPENDIX A: Juvenile Dermatomyositis and Disease Activity Score Patient Identification Number (Patient ID) Visit Date (d/m/y) Visit number DISEASE ACTIVITY SCORE (DAS) SCORE FUNCTIONAL STATUS (choose 1 category and report value in the score column; range for this section is0to3): Normal function, able to attend school, keeps up with friends 0 Mild limitations, tires after walking several blocks, general fatigue 1 Moderate limitations, requires assistance with stair-climbing, activity of daily living 2 Severe limits, wheelchair-bound, unable to attend school, climb stairs, etc. 3 WEAKNESS (Score 1 point for each area of weakness noted; circle all that apply and sum all YES and then report total value in the score column; range score for this section 0 is to 8): No Yes a) Neck flexor weakness 0 1 b) Difficulty clearing scapula (abdominal 0 1 weakness): can do sit-up with arms: 1) out 2) crossed 3) behind head 4) 1 3 cleared) c) Upper proximal muscle weakness 0 1 d) Lower proximal muscle weakness 0 1 e) Gower s sign (assisted/unassisted) 0 1 f) Abnormal gait 0 1 g) Difficulty swallowing 0 1 h) Nasal speech 0 1 SKIN INVOLVEMENT TYPE (choose one category and report value in the score column; range for this section is 0 to 4): Absent or resolved completely 0 Atrophic changes only 1 Erythema-mild 2 Erythema-moderate 3 Erythema-severe 4 SKIN INVOLVEMENT DISTRIBUTION (choose one category and report value in the score column; range for this section is 0 to 3): None 0 Focal (including area of joint-related skin) 1 Diffuse (including extensor surfaces of limbs shawl area) 2 Generalized (including trunk involvement) 3 VASCULITIS If NONE of the categories below are present then score 0 point for this section, if ANY of the categories below are present, then score 1 point for this section; range is 0 to 1: Absent Present a) Eyelid erythema 0 1 b) Eyelid vessel dilation 0 1 c) Eyelid thrombosis 0 1 d) Nailfold erythema 0 1 e) Nail bed telangiectasia 0 1 f) Palate dilation 0 1 g) Other 0 1 GOTTRON S PAPULES: If NO papules are present then score is 0 point, if any papules are present (mild, moderate or severe) then score is 1 point; range for this section is 0 to 1: Absent 0 Mild 1 Moderate 2 Severe 3 SUM ALL THE VALUES IN THE SCORE COLUMN ON THE RIGHT SIDE TO OBTAIN THE DAS TOTAL SCORE (TOTAL RANGE MINIMUM 0 MAXIMUM 20) Print Name of Investigator City and Date (d/m/y)

9 Analysis of Disease Activity Score 15 APPENDIX A: Continued DISEASE ACTIVITY SCORE INSTRUCTIONS A) NECK FLEXOR STRENGTH 1. The examiner s hand is placed on the child s forehead 2. The child is asked to lift his or her head from the examining table 3. The examiner s (who places the effort in the appropriate age range) evaluates the strength of the effort B) SIT UP (DIFFICULTY CLEARING SCAPULA) The child is asked to perform a sit up alone, without any counterbalance. Lack of age appropriate performance is counted as weak for the scoring. The gradations are as follows: 1. The child rolls to one side and needs to push off an elbow to sit up 2. The child clears 1 3 or less of the scapula 3. Sit-ups are performed with arms extended, crossed, and behind head. C) UPPER ARM STRENGTH 1. The child sits on the examining table with the arms extended to the side at shoulder height and the examiner places the hands between the shoulder and the elbow and attempts to lower arms. 2. The child holds the arm in front of body with elbows flexed, and the examiner attempts to extend the arm. D) LOWER ARM STRENGTH 1. Child lies on back on the table and raises leg so that the foot is at least 18 from mattress. 2. The examiner places hand midway between the hip and the knee and attempts to push the leg down on the mattress. E) GOWER S SIGN The child sits on the floor (making sure that the examining gown is not in the way) and is asked to hold his or her hands at ear level and get up. Any unsteady maneuver or touching the body or the room s equipment (floor included) is noted and considered assisted. F) ABNORMAL GAIT Any deviation from normal gait is recorded as 1 G) DIFFICULTY SWALLOWING The child is asked if he or she had difficulty swallowing a cracker, and the answer is recorded. H) NASAL SPEECH The child is asked to say the alphabet up to the letter E, and to say The letter E three times. The quality of speech is noted assigning 1 to any deviation from normal. DISEASE ACTIVITY SCORE (DAS) SCORING To calculate the total DAS score for a JDM patient it is necessary to add points for 2 components: Muscle weakness Skin involvement Muscle weakness 1. Functional status 0 3 points possible 2. Weakness 0 or 1 point for each of the 8 categories; 8 total Possible points for muscle weakness: 0 11 Skin involvement 3. Type 0 4 points 4. Distribution 0 3 points 5. Vasculitis 0 or 1 point if any are present 6. Gottron s papules 0 or 1 point if any are present Possible points for skin involvement: 0 9 To obtain the total DAS score it is now sufficient to add the 2 components: muscle weakness (range 0 11) and skin involvement (range 0 9). The possible DAS total score range is therefore 0 20.