The Pediatric Rheumatology International Trials Organization criteria for the evaluation of response to therapy in juvenile systemic lupus erythematosusProspective validation of the disease activity core set.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 9, September 2005, pp 2854–2864 DOI 10.1002/art.21230 © 2005, American College of Rheumatology The Pediatric Rheumatology International Trials Organization Criteria for the Evaluation of Response to Therapy in Juvenile Systemic Lupus Erythematosus Prospective Validation of the Disease Activity Core Set Nicolino Ruperto,1 Angelo Ravelli,1 Ruben Cuttica,2 Graciela Espada,3 Seza Ozen,4 Oscar Porras,5 Flavio Sztajnbok,6 Fernanda Falcini,7 Ozgur Kasapcopur,8 Helen Venning,9 Blanca Bica,10 Rosa Merino,11 Cecilia Coto,12 Joan Ros,13 Gordana Susic,14 Marı́a Luz Gamir,15 Kirsten Minden,16 Yvonne See,17 Yosef Uziel,18 Masha Mukamel,19 Phil Riley,20 Francesco Zulian,21 Alma Nunzia Olivieri,22 Rolando Cimaz,23 Hermann Girschick,24 Ingrida Rumba,25 Silvio Cavuto,1 Angela Pistorio,26 Daniel J. Lovell,27 and Alberto Martini,28 for the Pediatric Rheumatology International Trials Organization (PRINTO) and the Pediatric Rheumatology Collaborative Study Group (PRCSG) Objective. To validate and promulgate a core set of outcome measures for the evaluation of response to treatment in patients with juvenile systemic lupus erythematosus (SLE). Methods. In 2001, a preliminary consensusderived core set of measures for evaluating the response to therapy in juvenile SLE was established. In the present study, the core set was validated through an evidence-based, large-scale data collection process that led to the enrollment of 557 patients from 39 different countries. Consecutive patients with active disease were assessed at baseline and after 6 months. The validation procedures included assessment of feasibility, responsiveness, discriminant and construct ability, agreement in the evaluation of response to therapy between physicians and parents, redundancy, internal consistency, and ability to predict a therapeutic response. Supported by a grant from the European Union (contract QLG1-CT-2000-00514), the IRCCS G. Gaslini, Genoa, Italy, and the NIH (grant RO3-AI-44046). 1 Nicolino Ruperto, MD, MPH, Angelo Ravelli, MD, Silvio Cavuto, PhD: IRCCS G. Gaslini, Pediatria II, Reumatologia, Genoa, 2 Italy; Ruben Cuttica, MD: Hospital General de Ninos Pedro de Elizalde, Buenos Aires, Argentina; 3Graciela Espada, MD: Hospital de Ninos Ricardo Gutierrez, Buenos Aires, Argentina; 4Seza Ozen, MD: Hacettepe University Children’s Hospital, Ankara, Turkey; 5Oscar Porras, MD, PhD: Hospital Nacional de Ninos, San Jose, Costa Rica; 6 Flavio Sztajnbok, MD: Hospital Universitario Pedro Ernesto, Rio de Janeiro, Brazil; 7Fernanda Falcini, MD: Ospedale A. Meyer, Florence, Italy; 8Ozgur Kasapcopur, MD: Cerrahpasa Tip Fakultesi, Istanbul, Turkey; 9Helen Venning, MD: Queen’s Medical Centre, Nottingham, UK; 10Blanca Bica, MD: Federal University de Rio de Janeiro, Rio de Janeiro, Brazil; 11Rosa Merino, MD: Hospital Universitario La Paz, Madrid, Spain; 12Cecilia Coto, MD: Hospital Pediatrico Universitario Docente Pedro Borras, Havana, Cuba; 13Joan Ros, MD: Hospital San Joan de Deu, Barcelona, Spain; 14Gordana Susic, MD: Institute of Rheumatology, Belgrade, Serbia and Montenegro; 15Marı́a Luz Gamir, MD: Hospital Ramon y Cajal, Madrid, Spain; 16Kirsten Minden, MD: Klinikum Berlin-Buch, Berlin, Germany; 17Yvonne See, MD: KK Women’s and Children’s Hospital, Singapore; 18Yosef Uziel, MD: Meir Medical Centre, Kfar Saba, Israel; 19Masha Mukamel, MD: Schneider Children’s Hospital, Petach-Tikvah, Israel; 20Phil Riley, MD: Great Ormond Street Hospital, London, UK; 21Francesco Zulian, MD: Clinica Pediatria, Padua, Italy; 22Alma Nunzia Olivieri, MD: Seconda Università degli Studi di Napoli, Naples, Italy; 23Rolando Cimaz, MD: Clinica Pediatrica de Marchi, Milan, Italy; 24Hermann Girschick, MD: University School of Medicine Children’s Hospital, Wurzburg, Germany; 25Ingrida Rumba, MD: University of Latria, Riga, Latvia; 26Angela Pistorio, MD, PhD: IRCCS G. Gaslini, Servizio di Epidemiologia e Biostatistica, Genoa, Italy; 27Daniel J. Lovell, MD, MPH: Children’s Hospital Medical Center, Cincinnati, Ohio; 28Alberto Martini, MD: IRCCS G. Gaslini, Pediatria II, Reumatologia and Università degli Studi, Genoa, Italy. Address correspondence and reprint requests to Nicolino Ruperto, MD, MPH, Pediatric Rheumatology International Trials Organization (PRINTO), IRCCS G. Gaslini, Università di Genova, Pediatria II–Reumatologia, Largo Gaslini, 5, 16147 Genoa, Italy. E-mail: email@example.com. Submitted for publication December 2, 2004; accepted in revised form May 16, 2005. 2854 RESPONSE TO THERAPY IN JUVENILE SLE Results. The following clinical measures were found to be feasible and to have good construct validity, discriminative ability, and internal consistency; furthermore, they were not redundant, proved responsive to clinically important changes in disease activity, and were associated strongly with treatment outcome and thus were included in the final core set: 1) physician’s global assessment of disease activity, 2) global disease activity measure, 3) 24-hour proteinuria, 4) parent’s global assessment of the patient’s overall well-being, and 5) health-related quality of life assessment. Conclusion. The members of PRINTO propose a core set of criteria for the evaluation of response to therapy that is scientifically and clinically relevant and statistically validated. The core set will help standardize the conduct and reporting of clinical trials and assist practitioners in deciding whether a patient with juvenile SLE has responded adequately to therapy. Systemic lupus erythematosus (SLE) is a multisystem inflammatory disease that is characterized by protean clinical manifestations, an unpredictable course, and a substantial risk of morbidity. Although recent improvements in care have markedly enhanced the survival of SLE patients who have the most severe and life-threatening manifestations, disease treatment remains largely empiric and unsatisfactory. One of the leading factors that has hampered a rational therapeutic approach to SLE is the lack of standardized and validated measures for assessing the response to therapy. This deficiency leads to an inability to accurately evaluate or compare the efficacy of drug therapies. To overcome this problem, the Outcome Measures in Rheumatology Clinical Trials (OMERACT) group, the American College of Rheumatology (ACR), and other investigators (1–3) recommended that outcome measures should be included in SLE trials; however, until now these proposals have not yet been formally validated in prospective studies or clinical trials. Although children, adolescents, and adults with SLE share many signs and symptoms of disease, they differ in terms of the frequency and severity of disease activity and damage features (4–8), and treatment approaches should consider the distinctive characteristics of patients with juvenile SLE as well as their longer life expectancy. Therefore, it cannot be assumed a priori that the clinical measures developed for adults are suitable for children and adolescents. Furthermore, younger patients with juvenile SLE deserve to be assessed with specific instruments that take into account the disease and physical/mental age-related issues that 2855 are associated with growth and development. Therefore, all outcome measures developed for adults need to be subjected to a critical evidence-based evaluation of their measurement properties in children and adolescents. To standardize the evaluation of response to therapy in patients with juvenile SLE, the Pediatric Rheumatology International Trials Organization (PRINTO) (9), in collaboration with the Pediatric Rheumatology Collaborative Study Group (PRCSG), and with the support of the European Union and the US National Institutes of Health, undertook a multinational effort that was aimed to develop, validate, and promulgate a core set of measures and a definition of clinical improvement in patients with juvenile SLE, similar to that already done with the ACR Pediatric 30 criteria for juvenile rheumatoid arthritis (10–12). The results of the first part of the study, published previously (13), led to the definition of a preliminary consensus-based (14) core set of domains. Here, we report the results of the second phase of the project, which was aimed at formally validating the preliminary juvenile SLE core set for the evaluation of response to therapy through a prospective, large-scale data collection process among the members of the PRINTO/ PRCSG networks. Our objective was to further define and validate the preliminary juvenile SLE core set of response variables in order to document the response to therapy in patients with juvenile SLE. PATIENTS AND METHODS Study design. Enrollment began in June 2001 and ended in March 2004. The participating PRINTO/PRCSG members were asked to assess all variables in the preliminary core set, in all patients seen consecutively in their units who 1) had a diagnosis of SLE according to the 1997 revised ACR classification criteria (15,16), were younger than age 18 years, and were experiencing an active phase of disease, defined as either the need to start corticosteroid therapy and/or a new immunosuppressive medication or, in those receiving ongoing therapy, the need to receive a major increase in the dosage of corticosteroids and/or immunosuppressive drugs. Six months after the baseline evaluation, the core set variables were reassessed in each patient. We chose this study protocol and time frame to approximate what is usually done in a clinical trial, with a baseline assessment at the time when treatment is started and a final assessment at the end of the trial. Patients were excluded from the study if, at baseline, they were experiencing drug-induced or spontaneous clinical remission, were receiving a stable therapeutic regimen, or had a concomitant serious illness. In each center, written or verbal informed consent was obtained from a parent or legal guardian, according to the requirements of the local ethics committees. 2856 Assessment of preliminary core set variables. The following preliminary core set measures were assessed at baseline and 6 months later: 1) the physician’s global assessment of the patient’s overall disease activity on a 10-cm visual analog scale (VAS) (where 0 ⫽ no activity and 10 ⫽ maximum activity); 2) the parent’s global assessment of the patient’s overall well-being during the previous week on a 10-cm VAS (where 0 ⫽ very well and 10 ⫽ very poor); 3) anti–doublestranded DNA (anti-dsDNA) antibodies, determined by enzyme-linked immunosorbent assay (ELISA), immunofluorescence, or qualitative assessment as being positive or negative; 4) 24-hour proteinuria and the serum creatinine level; 5) global assessment of disease activity according to the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) (17) (the SLEDAI is a validated measure of SLE disease activity, also used in children , that contains 24 descriptors in 9 organ systems and is weighted to reflect the degree of activity during the last 10 days; the total SLEDAI score can range from 0 to 105); and 6) health-related quality of life assessment via the parent’s version of the Child Health Questionnaire (CHQ) (19,20). Briefly, the CHQ is a generic instrument designed to capture the physical, emotional, and social components of health status in children at least 5 years of age; it includes 15 subscales and 2 summary measures, the physical health score [PhS] and the psychosocial health score [PsS], standardized to have a mean of 50 and SD of 10; higher scores in the scales indicate better health-related quality of life. Additional clinical and laboratory assessments. As additional tools for assessing global disease activity, we used the Systemic Lupus Activity Measure (SLAM) (21) and the European Consensus Lupus Activity Measurement (ECLAM) (22,23). Both of these tools have been used in children (18,24). Briefly, the SLAM assesses symptoms that occurred during the preceding month; it includes 24 clinical manifestations and 8 laboratory parameters, and the maximum possible score is 90. The ECLAM measures changes in SLE disease activity, with respect to the last clinical observation (9 clinical manifestations and 3 laboratory assessments); the total score ranges from 0 to 10. Because the usefulness of the Health Assessment Questionnaire (HAQ) has been investigated in adults with SLE (25,26), we used the Childhood HAQ (C-HAQ) (20,27). Briefly, the C-HAQ measures a child’s functional ability in 8 activities of daily living. These 8 domains are then averaged to calculate the C-HAQ disability index, with scores ranging from 0 to 3 (where 0 ⫽ best and 3 ⫽ worst). As part of the C-HAQ, we also reported the parent’s global assessment of the child’s pain during the previous week on a 10-cm VAS (where 0 ⫽ no pain and 10 ⫽ very severe pain). The parent’s version of both the C-HAQ and the CHQ have been translated and validated in all of the languages of the participating countries (20). Additional laboratory tests were a complete blood cell count, the Westergren erythrocyte sedimentation rate (ESR), and assessment of serum C3 and C4. Standardization of laboratory tests. The test results were standardized based on the normal values provided by each local laboratory. All values were converted to the international standard (SI) unit system (28), according to the following formulas: xref ⫽ [(rangeref ⫻ xref)/100] ⫹ minref, where xref ⫽ value in the SI system, rangeref ⫽ normal range of the SI system, and minref ⫽ minimum value in the SI system; RUPERTO ET AL and xref ⫽ [(xi ⫺ minOMU)/rangeOMU] ⫻ 100, where xi ⫽ value in the original measurement unit (OMU), rangeOMU ⫽ normal range of the OMU, and minOMU ⫽ minimum value of the OMU. For example, a C3 level of 72 mg/dl (OMU normal values 80–120 mg/dl) was converted to 0.52 gm/liter (SI normal values 0.7–1.6 gm/liter). Validation procedures. Validation of the core set measures was conducted with the use of the OMERACT filter for outcome measures in rheumatology (29,30). The feasibility or practicality of the measures was determined by addressing the issues of brevity, simplicity, ease of scoring, and percentage of missing values. Face and content validity were based on results of the previous consensus conference (13). Responsiveness was examined by determining the ability of each variable to detect clinically important change between baseline and 6 months and was measured using the standardized response mean (SRM). The SRM was calculated as the absolute mean change in score divided by the SD of an individual’s change in score; 95% confidence intervals (95% CIs) were also provided (31,32). An absolute value of 0.2 is considered to represent a small effect, a value of 0.5 represents a moderate effect, and values of ⱖ0.8 represent a large effect (33,34). Discriminative ability was assessed by evaluating the capacity to distinguish patients who experienced improvement from those who did not. For this purpose, at the end of the 6-month treatment period physicians and parents were asked (independently of each other) to judge whether the patient’s disease had improved, was stable, or had worsened as compared with baseline. Physicians and parents were not allowed to review their baseline global assessments or the results of other core set variables. Patients who were judged as improving were compared with those who were judged as not improving (i.e., disease remained stable or worsened) by t-test or the Mann-Whitney U test, as appropriate. Moreover, the level of agreement between physicians and parents in the evaluation of response to therapy was assessed with the kappa statistic (35), using the threshold proposed by Landis and Koch (36). Convergent construct validity, which is a form of validation that seeks to examine whether the construct in question is related to other measures in a manner consistent with a priori prediction, was also investigated. As a surrogate measure, we chose the physician’s global assessment of the patient’s overall disease activity by Spearman’s rank correlation (where a value of ⬎0.7 was considered high, a value of 0.4–0.7 was moderate, and a value of ⬍0.4 was low). We predicted that correlation of the underlying construct of response to therapy with the surrogate gold standard measure would be in the moderate range, and thus would provide a different perspective and avoid redundancy (see below). The issue of colinearity (or redundancy) of variables was investigated by means of Pearson’s correlation coefficient; a coefficient of ⱖ0.7 was considered to represent evidence of colinearity. The internal consistency of the various scales was determined by Cronbach’s alpha (37), with the following cutoffs: ⬍0.6 ⫽ poor, 0.6–0.64 ⫽ slight, 0.65–0.69 ⫽ fair, 0.7–0.79 ⫽ moderate, 0.8–0.89 ⫽ substantial, and ⬎0.9 ⫽ almost perfect. We determined that a slight/fair Cronbach’s alpha would be sufficient to demonstrate the internal consistency of the core set. Finally, the association between the 5 core set mea- RESPONSE TO THERAPY IN JUVENILE SLE 2857 Table 1. Descriptive characteristics of the variables* Variable Final core set Physician’s global assessment of patient’s overall disease activity (0–10-cm scale)1 Parent’s global assessment of patient’s overall well-being (0–10-cm scale)1 Proteinuria, gm/24 hours1 ECLAM (range 0–10)1 CHQ physical health summary score (range 40–60)2 Additional measures SLEDAI (range 0–105)1 SLAM (range 0–90)1 CHQ psychosocial health summary score (range 40–60)2 C-HAQ disability index (0–3)1 Parent’s global assessment of child’s pain (0–10-cm scale)1 Anti-dsDNA, IU, by ELISA1 Anti-dsDNA titer, by immunofluorescence1 Anti-dsDNA, no. (%) positive1 Serum creatinine (normal 0.6–1.2 mg/dl)1 ESR (normal 0–30 mm/hour)1 C3 (normal 0.7–1.6 gm/liter)2 C4 (normal 0.2–0.4 gm/liter)2 Hemoglobin, gm/liter2 Sample size† Month 0 Month 6 % change SRM (95% CI) 529 5.8 (3.8, 7.9) 1 (0.3, 2.6) ⫺168 (640, ⫺100) 1.3 (1.2–1.4) 484 4.5 (1.6, 6.9) 0.8 (0, 2.4) ⫺100 (⫺261, 0) 0.8 (0.7–0.9) 397 533 432 0.3 (0.1, 1.2) 6 (4, 8) 41 (28, 49) 0.1 (0, 0.4) 2 (1, 3) 52 (43, 54) ⫺91 (⫺248, 0) ⫺100 (⫺250, ⫺67) 16 (1, 40) 0.4 (0.3–0.4) 1.3 (1.2–1.4) 0.8 (0.7–0.9) 531 533 432 504 485 17 (11, 23) 15 (10, 20) 45 (36, 52) 0.4 (0, 1.4) 3 (0.4, 6.5) 4 (2, 8) 4 (2, 7) 49 (43, 55) 0 (0, 0.3) 0.2 (0, 1.5) ⫺119 (⫺325, ⫺77) ⫺200 (⫺500, ⫺88) 7 (⫺3, 20) ⫺95 (⫺100, 0) ⫺100 (⫺241, 0) 1.1 (1–1.2) 1.3 (1.2–1.4) 0.5 (0.4–0.6) 0.6 (0.5–0.7) 0.7 (0.6–0.8) 228 84 64 514 487 486 477 520 140 (40, 300) 160 (40, 321) 34 (53) 0.7 (0.6, 0.8) 57 (32, 91) 0.5 (0.2, 0.8) 0.2 (0.2, 0.2) 10.9 (9.4, 12.1) ⫺100 (⫺347, ⫺20) ⫺100 (⫺700, 0) – 0 (⫺18, 14) ⫺135 (⫺342, ⫺28) 32 (2, 71) 13 (0, 28) 12 (2, 22) 0.3 (0.1–0.4) 0.4 (0.3–0.5) – 0.1 (0–0.2) 0.9 (0.9–1) 0.6 (0.5–0.7) 0.4 (0.3–0.5) 0.8 (0.7–0.9) 54 (10, 146) 80 (20, 160) 18 (28) 0.7 (0.6, 0.8) 21 (12, 35) 0.8 (0.6, 1.1) 0.2 (0.2, 0.3) 12.5 (11.5, 13.6) * Except where indicated otherwise, values are the median (interquartile range). 1 indicates that a higher score for that variable denotes worse disease activity; 2indicates that a lower score denotes worse disease activity. SRM ⫽ standardized response mean; 95% CI ⫽ 95% confidence interval; ECLAM ⫽ European Consensus Lupus Activity Measurement; CHQ ⫽ Child Health Questionnaire; SLEDAI ⫽ Systemic Lupus Erythematosus Disease Activity Index; SLAM ⫽ Systemic Lupus Activity Measure; C-HAQ ⫽ Childhood Health Assessment Questionnaire; anti-dsDNA ⫽ anti–double-stranded DNA; ELISA ⫽ enzyme-linked immunosorbent assay; ESR ⫽ erythrocyte sedimentation rate. † Number of patients for whom both the baseline and 6-month evaluations were available. sures and response to therapy as judged by the attending physician was evaluated through a multivariate logistic regression analysis, after having dichotomized the core set measures according to the best cutoffs obtained from the receiver operating characteristic (ROC) curve analysis (38). Determination of the best cutoff for each core set variable will help physicians decide whether a patient has improved based on the absolute change in that particular measure. Data were entered in an Access XP database and analyzed by 2 of the authors (NR and AP) with Excel XP software (Microsoft, Redmond, WA), XLSTAT-Pro 6.1.9 software (Addinsoft, Brooklyn, NY), Statistica 6.0 software (StatSoft, Tulsa, OK), and Stata version 7.0 software (Stata, College Station, TX). RESULTS Demographic characteristics. A total of 557 patients were enrolled from 118 centers in 39 countries, as follows: Argentina (n ⫽ 50), Australia (n ⫽ 1), Belgium (n ⫽ 2), Brazil (n ⫽ 64), Bulgaria (n ⫽ 15), Canada (n ⫽ 25), Chile (n ⫽ 8), Costa Rica (n ⫽ 10), Croatia (n ⫽ 5), Cuba (n ⫽ 7), The Czech Republic (n ⫽ 7), Denmark (n ⫽ 6), Finland (n ⫽ 4), France (n ⫽ 11), Germany (n ⫽ 28), Greece (n ⫽ 11), Hungary (n ⫽ 5), India (n ⫽ 6), Israel (n ⫽ 20), Italy (n ⫽ 91), Korea (n ⫽ 5), Latvia (n ⫽ 6), Mexico (n ⫽ 11), The Netherlands (n ⫽ 10), Norway (n ⫽ 2), Poland (n ⫽ 5), Portugal (n ⫽ 1), Saudi Arabia (n ⫽ 2), Serbia and Montenegro (n ⫽ 23), Singapore (n ⫽ 6), Slovakia (n ⫽ 12), Slovenia (n ⫽ 4), Spain (n ⫽ 26), Sweden (n ⫽ 9), Switzerland (n ⫽ 3), Tunisia (n ⫽ 1), Turkey (n ⫽ 20), the UK (n ⫽ 23), and the US (n ⫽ 12). Of the 557 patients enrolled, 24 patients (4%) did not complete the study; 14 of these patients were lost to followup, and the other 10 died (6 patients died of sepsis, 1 patient died of pancreatitis, 1 patient committed suicide, 1 patient had multiorgan failure, and 1 patient died of unknown causes). Of the 533 patients who completed both the baseline and 6-month assessments, 438 (82%) were female and 95 (18%) were male; the median age at diagnosis was 12.7 years (interquartile range [IQR] 11–14.7 years), and the median disease duration was 0.5 years (IQR 0.2–2.4 years). The median number of ACR classification criteria at the time of SLE diagnosis was 6 (IQR 5–7). Feasibility and responsiveness. Table 1 shows the characteristics of each clinical variable in the 533 2858 RUPERTO ET AL Figure 1. Ability of the variables (mean score changes) included in the core set to discriminate between patients who improved versus patients who did not improve according to the physician’s and the parent’s evaluation after 6 months of therapy. Data are presented as box plots, where the squares inside the boxes represent the mean, and the lines outside the boxes represent the 95% confidence interval. P values refer to the discriminant ability of the variables according to the physician’s evaluation and to the parent’s evaluation of response to therapy. MD ⫽ physician; ECLAM ⫽ European Consensus Lupus Activity Measurement; CHQ ⫽ Child Health Questionnaire. patients for whom the variable was measurable at both time points. The frequency of missing data was uniformly ⬍10%, demonstrating that all variables had excellent feasibility. At baseline, patients had, on average, a high level of disease activity, as shown by the median of the physician’s and parent’s global assessment and that of the SLEDAI. Of the 397 patients for whom 24-hour proteinuria values were available at both time points, 181 (46%) had renal disease at baseline (defined as 24-hour proteinuria of ⬎0.5 gm/day), and 78 patients RESPONSE TO THERAPY IN JUVENILE SLE Table 2. matrix* 2859 Construct validity for the variables included in the final core set, by Pearson’s correlation Variable 24-hour proteinuria ECLAM CHQ physical health summary score Parent’s global assessment of patient’s overall well-being Physician’s global assessment 24-hour CHQ physical health (0–10 cm) proteinuria ECLAM summary score 0.2 0.5 ⫺0.4 0.4 0.2 ⫺0.2 0.1 ⫺0.4 0.3 ⫺0.6 * Correlations for the absolute change in score (value at month 6 minus value at month 0) were performed and were expected to be in the moderate range (0.4–0.7). A Pearson’s coefficient of ⱖ0.7 was considered to represent evidence of redundancy. ECLAM ⫽ European Consensus Lupus Activity Measurement; CHQ ⫽ Child Health Questionnaire. (20%) had proteinuria of ⬎2 gm/day. Good responsiveness to clinical change (SRM ⱖ0.8) was demonstrated by the physician’s and parent’s global assessment, the 3 global disease activity tools (with the SLAM and the ECLAM being slightly superior to the SLEDAI), the physical health summary score of the health-related quality of life tool (the CHQ PhS), the ESR, and the hemoglobin concentration. Assessment of the C-HAQ score, the parent’s global assessment of the child’s pain, and the C3 level revealed intermediate responsiveness (SRM ⱖ0.6 and ⬍0.8), whereas 24-hour proteinuria, anti-dsDNA antibodies, the psychosocial health summary score of the health-related quality of life tool (the CHQ PsS), the serum creatinine level, and the C4 level proved to be poorly responsive (SRM ⬍0.6). Notably, the SRM for 24-hour proteinuria rose to 0.6 when only the 181 patients who had renal disease at baseline (⬎0.5 gm/day) were assessed and to 1.1 when only the 78 patients with proteinuria of ⬎2 gm/day were assessed. The poor performance of serum creatinine is probably attributable to the low number of patients (33 of 514 [6%]) who had an abnormal level at baseline (any value ⬎1.2 mg/dl). For this reason, only 24-hour proteinuria was retained as an indicator of kidney disease activity in subsequent analyses. The results obtained for anti-dsDNA antibodies should be regarded with caution, because the determination of these antibodies was performed with different technical methods, which included ELISA in 228 patients (43%) and indirect immunofluorescence in 84 patients (16%); in 64 patients (12%), only a qualitative evaluation (positive versus negative) was available. This heterogeneity precluded combining results obtained with different methods, and anti-dsDNA were therefore excluded from further consideration (see below). The white blood cell count with differential and the platelet count showed poor responsiveness (data not shown). Taken together, these results did not show a major advantage for any of the additional variables over the variables included in the preliminary core set (13). However, due to the superior responsiveness to clinically important change (and minor skewness) demonstrated by the ECLAM as compared with the SLEDAI, the ECLAM was selected for use instead of the SLEDAI; moreover, the ECLAM is the only one of the 3 indexes that uses the entire range of possible scores (range 0–10; median score at baseline 5). Furthermore, since, of the 2 summary scales of the CHQ (PhS and PsS), only the PhS yielded significant results in previous analyses, we used only the CHQ PhS as a measure of health-related quality of life in all subsequent evaluations. Discriminant validity. Figure 1 shows the variables included in the final core set, which demonstrated significant ability in discriminating patients who were improved or not improved at 6 months based on the physician’s or parent’s assessment of the child’s response to therapy. Other variables that were able to show a statistically significant discriminant ability according to the physician’s and/or the parent’s evaluation were antidsDNA antibodies tested by ELISA, the SLEDAI, the SLAM, the ESR, the lymphocyte count, the hemoglobin concentration, several subscales of the CHQ, all subscales of the C-HAQ, and the parent’s global assessment of the child’s pain (data not shown). Notably, agreement between physicians and parents in the evaluation of response to therapy was substantial ( ⫽ 0.72 [95% CI 0.6–0.8]). Construct validity and redundancy. Table 2 shows Pearson’s correlation coefficients for the baselineto–6-month change in the final core set variables. This analysis was carried out to assess both the construct 2860 RUPERTO ET AL Table 3. Internal consistency of the variables in the core set* Variable Cronbach’s ␣ Physician’s global assessment of the patient’s overall disease activity 24-hour proteinuria ECLAM CHQ physical health summary score Parent’s global assessment of the patient’s overall well-being 0.52 0.63 0.53 0.76 0.52 * Values shown are Cronbach’s alpha when the individual variable is removed. The performance of the final core set, including all 5 variables, was 0.62. ECLAM ⫽ European Consensus Lupus Activity Measurement; CHQ ⫽ Child Health Questionnaire. validity, using as the reference measure the physician’s global assessment of the patient’s overall disease activity, and the colinearity (or redundancy) of clinical variables. As expected, the correlation with the physician’s global assessment of the patient’s overall disease activity was in the moderate range (r ⫽ ⫾0.4 to ⫾0.5) for all variables, except for 24-hour proteinuria, which showed poor correlation. These results demonstrate that the final 5 core set variables have good convergent construct validity. Finally, there was no redundancy between the core set variables (Pearson’s correlation coefficient ⬍0.7). Internal consistency. Cronbach’s alpha was used to assess the degree to which the final 5 core set variables “hold together” or, in other words, the degree to which they were measuring the same underlying construct. As shown in Table 3, assessment of the absolute change in scores for all 5 variables combined yielded a Cronbach’s alpha of 0.62, meaning, as expected, that there was slight internal consistency. When Table 4. Table 5. Domains and suggested variables included in the final core set for the evaluation of response to therapy in juvenile SLE* Domain Suggested variable(s) Physician’s global assessment of the patient’s overall disease activity Parent’s global assessment of the patient’s overall well-being Renal involvement Global juvenile SLE disease activity tool Health-related quality of life assessment 10-cm VAS 10-cm VAS 24-hour proteinuria ECLAM (or SLEDAI or SLAM) CHQ physical health summary score * SLE ⫽ systemic lupus erythematosus; VAS ⫽ visual analog scale; ECLAM ⫽ European Consensus Lupus Activity Measurement; SLEDAI ⫽ Systemic Lupus Erythematosus Disease Activity Index; SLAM ⫽ Systemic Lupus Activity Measure; CHQ ⫽ Child Health Questionnaire. we add anti-dsDNA antibodies (determined by ELISA or immunofluorescence), Cronbach’s alpha fell to 0.003, suggesting that the inclusion of this measure in the core set led to a disruption of its internal consistency. Association between changes in each of the 5 core set measures and overall outcome. The ability of the change in each core set variable to predict the response to therapy was analyzed in a multivariate analysis, which used as explanatory variables the baseline-to–6-month changes in each outcome measure and as the dependent variable the physician’s judgment of the treatment outcome at 6 months (improvement or no improvement). In the final model (Table 4), the physician’s global assessment of the patient’s overall disease activity and the ECLAM appeared to be the strongest predictors of response to therapy, whereas the predictive ability of the other 3 variables did not reach statistical significance. Logistic regression model to predict improvement according to the physician’s evaluation Variable OR (95% CI) P, likelihood ratio test Physician’s global assessment of patient’s overall disease activity ECLAM 24-hour proteinuria CHQ physical health summary score Parent’s global assessment of patient’s overall well-being 5.2 (2.6–10.5) ⬍0.0001 2.1 (1.0–4.3) 1.3 (0.6–2.5) 1.5 (0.7–3.3) 1.5 (0.7–3.2) 0.046 0.49 0.256 0.254 * Predictions were based on the absolute change in the variables included in the final core set. Variables were dichotomized according to the best cutoffs obtained from the receiver operating characteristics (ROC) curve analysis. The area under the ROC curve of the model was equal to 0.79. The best cutoffs (sensitivity and specificity) were as follows: for physician’s global assessment of the patient’s overall disease activity, ⱕ⫺2.1 (sensitivity 78.8%, specificity 69%); for the European Consensus Lupus Activity Measurement (ECLAM), ⱕ⫺3 (sensitivity 74.6%, specificity 67%); for 24-hour proteinuria, ⱕ⫺0.08 (sensitivity 59.4%, specificity 63.2%); for the Child Health Questionnaire (CHQ) physical health summary score, ⬎4.26 (sensitivity 66.2%, specificity 60%); for parent’s global assessment of the patient’s overall well-being, ⱕ⫺1.1 (sensitivity 69.3%, specificity 60.8%). OR ⫽ odds ratio; 95% CI ⫽ 95% confidence interval. RESPONSE TO THERAPY IN JUVENILE SLE Selection of the final core set. Taken together, the results of the validation analyses showed that the final core set for the evaluation of response to therapy in juvenile SLE has excellent psychometric properties. Table 5 presents the 5 domains and the related suggested variables used to measure each domain that is included in the final core set. DISCUSSION In this report, we present the final validated PRINTO core set of clinical and laboratory measures for the assessment of response to therapy in patients with juvenile SLE. To ensure an evidence-based selection process, the evaluative properties of candidate variables were carefully assessed using a large prospective data collection procedure and a comprehensive validation process that closely mimicked the design of a clinical trial. The selected variables were shown to be feasible and to have good construct validity, discriminative ability, and internal consistency; furthermore, they were not redundant, they were responsive to clinically important change, and they were strongly associated with treatment outcome. Validation of the PRINTO core set is a fundamental step in the process of developing a definition of improvement in juvenile SLE (39). We did not attempt to create specific combinations of variables for the various types of organ involvement in juvenile SLE. Rather, we attempted to make the core set robust enough to cover all disease phenotypes, focusing on the central features of the physician’s subjective estimation of the level of disease activity, global disease activity scoring, parent’s global assessment of the patient’s overall well-being, and the health-related quality of life; only a specific domain aimed to assess renal disease, which represents one of the most important and frequent organ involvements in juvenile SLE, was included. Although it is uncertain whether use of agents with therapeutic benefit in juvenile SLE would lead to improvement in all core set measures, it is important that any therapy that is aimed at improving disease in one organ system does not worsen disease elsewhere. It should be kept in mind, however, that the recommended variables are not more than a minimal core set, and that investigators can measure as many other variables as they deem appropriate for the major hypothesis that is being tested. For example, serum creatinine, creatinine clearance, the protein:creatinine ratio, urine casts, red blood cells, and other variables would be considered for inclusion in clinical trials specifically focused on juvenile SLE nephritis (40); the same approach would apply to a 2861 study on central nervous system disease, which would require specific end points, such as findings on magnetic resonance imaging, electroencephalogram, singlephoton–emission computed tomography, and psychometric tests. Ideally, outcome criteria must be scientifically valid, clinically relevant, and feasible from a practical standpoint. However, identifying criteria for SLE that fulfill all of these requirements has been a difficult task (1–3,40,41). The OMERACT and the ACR groups recommended that any trial in SLE should include a disease activity measure (1–3). Thus far, however, there has been no consensus as to which measure is preferable. In our study, the decision to use the ECLAM was based on the indication of the statistical validation; notably, this index has been shown to be sensitive to change in another study of juvenile SLE (24). In our analysis, the responsiveness to change of the ECLAM was equal to that of the SLAM and was slightly superior to that of the SLEDAI. Interestingly, in the logistic regression model to predict improvement according to the physician’s evaluation (Table 4), we observed the same cutoff (less than or equal to ⫺3) for the ECLAM as that reported in a recent study by the ACR Ad Hoc Committee to define response criteria for SLE based on measures of overall disease activity (3). Serologic markers, including anti-dsDNA antibodies, play an important role in the assessment of disease activity in SLE, but they are imperfect indicators of major changes in disease activity when used alone; thus, most clinicians do not respond to isolated change in these tests by modifying therapy (40). Anti-dsDNA antibodies were included in the preliminary core set (13) because they were thought to be an important measure for the evaluation of response to therapy. The results of the validation analysis led to deletion of this parameter from the final core set, because it was found to be poorly responsive to change, and its inclusion into the core set of variables led to a marked decrease in internal consistency. Our findings should be regarded with caution, however, because anti-dsDNA antibodies were detected with 3 different methods across study centers. Use of standardized methods and the availability of a central laboratory are needed to address this issue more specifically. Concerning serum complement, this measure was not incorporated in the final core set because its responsiveness was only intermediate, and because of the fact that incomplete normalization due to inherited deficiencies may preclude its reliability in the definition of clinical response (42,43). Consensus conference attendees believed that 2862 24-hour proteinuria, which is a single-organ measure, is worth retaining in the criteria, because it is of foremost importance in the clinical assessment of patients with juvenile SLE. It was agreed, however, that in cases in which a patient has no signs of renal disease at the start of a trial, this feature can be omitted from the assessment of therapeutic response, and that only a 4-item core set should be used. Because proteinuria may reflect renal damage in the absence of renal activity, a careful assessment of activity parameters of juvenile SLE nephritis should be done to confirm that the patient has active renal disease. Furthermore, the fact that some drugs, such as angiotensin-converting enzyme inhibitors, may lead to a reduction in proteinuria should be taken into account (40). We believe this is the first study of juvenile SLE or other chronic rheumatic diseases in which the assessment of health-related quality of life has been incorporated in the definition of a therapeutic response. Indeed, in recent years, it has been increasingly recognized that health-related quality of life measures must be included in SLE trials because they address aspects of SLE and its impact that are not fully captured by other outcome end points (1,2,44,45). In general, the patient’s perspective in outcome assessment is becoming increasingly important, because it has been suggested that a patient’s perceptions regarding his or her juvenile SLE is frequently different from those of the clinician (46). It is essential that trials demonstrating improvement in a specific organ or in disease activity demonstrate no or minimal worsening in measures of quality of life. We previously observed that the physical health scale of the CHQ (but not the psychosocial health scale) had satisfactory psychometric properties and had the advantage of being available in several different languages (20). Greater impairment according to the ratings on the physical health domain of the CHQ as compared with the psychosocial health domain was recently demonstrated in a large international cohort of patients with juvenile SLE (47). Further insights into the measurement properties of the health-related quality of life tools, including the potential impact of evaluations reported by children versus those reported by parents, will be obtained when these tools are used in the context of a clinical trial. Our study has certain limitations, which include the fact that it was not conducted in the context of a real clinical trial, and that the use of corticosteroids or immunosuppressive drugs as intervention therapy was not standardized based on recent recommendations (48) and might have led to changes in the level of disease activity much greater than those that would be expected RUPERTO ET AL in trials of novel immunosuppressive or biologic agents. The main strength of the study is the large amount of prospectively collected data, which ensured an evidencebased validation analysis. To our knowledge, this is the first time that clinical measures of juvenile SLE have been tested longitudinally for their statistical performance, individually and as a group. In conclusion, we have presented the validated PRINTO core set of outcome domains for the evaluation of response to therapy in juvenile SLE, which will constitute the basis for creating a definition of improvement to be used in randomized clinical trials. This will allow improved assessment of the efficacy of new therapeutic agents or regimens, with greater validity and comprehensiveness. ACKNOWLEDGMENTS We are indebted to Drs. Anna Tortorelli, Monica Tufillo, and Elisabetta Maggi for their help in data handling, their organization skills, and overall management of the project. We are also thankful to Dr. Luca Villa and Mr. Michele Pesce for their help in database development. We thank the following members of PRINTO who participated as investigators in the trial and whose enthusiastic efforts made this work possible: Maria Apaz, MD, Stella Garay, MD, Silvia Meiorin, MD, Ricardo Russo, MD (Argentina); Kevin Murray, MD (Australia); Rik Joos, MD (Belgium); Carlos Henrique M. Da Silva, MD, Virginia Ferriani, MD, Odete Hilario, MD, Claudia Machado, MD, Sheila Oliveira, MD, Iloite Scheibel, MD, Clovis Artur Silva, MD (Brazil); Dimitrina Mihaylova, MD (Bulgaria); Miroslav Harjacek, MD, Lana Tambic-Bukovac, MD (Croatia); Ximena Norambuena, MD, Arnoldo Quezada, MD (Chile); Santa Gomez Conde, MD (Cuba); Pavla Dolezalova, MD, Dana Nemcova, MD, Katerina Jarosova, MD (Czech Republic); Troels Herlin, MD, Susan Nielsen, MD (Denmark); Pekka Lahdenne, MD (Finland); Brigitte Bader-Meunier, MD, Michel Fischbach, MD, Sylvie Gandom-Laloum, MD, Chantal Job Deslandre, MD, Isabelle Kone Paut, MD, Anne-Marie Prieur, MD, Danièle Sommelet, MD (France); Jurgen Brunner, MD, Guenther Dannecker, MD, Stephan Ehl, MD, Ivan Foeldvari, MD, Gerd Ganser, MD, Gerd Horneff, MD, Hartmut Michels, MD, Johannes Roth, MD, Claudia Sengler, MD, Stephanie Schaueburg, MD (Germany); Christina Dracou, MD, Vasiliki Galanopoulou, MD, Florence KanakoudiTsakalidou, MD, Jenny Pratsidou-Gertsi, MD, Olga Voygioyka, MD (Greece); Zsolt Balogh, MD, Ilonka Orban, MD (Hungary); Amita Aggarwal, MD (India); Judith Barash, MD, Riva Brik, MD, Liora Harel, MD, Joseph Press, MD, Tsivia Tauber, MD (Israel); Maria Alessio, MD, Roberto Barcellona, MD, Fabrizia Corona, MD, Elisabetta Cortis, MD, Rosario Di Toro, MD, Franco Garofalo, MD, Loredana Lepore, MD, Alfredo Maria Lurati, MD, Sara Garozzo, MD, Valeria Gerloni, MD, Silvana Martino, MD, Lucia Trail, MD (Italy); Sang-Cheol Bae, MD, Chung Il Joung, MD (Korea); Valda Stanevicha, MD, Andrejs Scegolevs, MD (Latvia); Ruben RESPONSE TO THERAPY IN JUVENILE SLE Burgos-Vargas, MD, Carolina Duarte, MD, Raul Gutierrez Suarez, MD (Mexico); Lisetta Van Suijlekom Smit, MD, Rebecca Ten Cate, MD, Marion Van Rossum, MD, Nico Wulffraat, MD (The Netherlands); Berit Flato, MD (Norway); Anna Maria Romicka, MD, Malgorzata Wierzbowska, MD (Poland); Manuel Salgado, MD (Portugal); Sulaiman AlMayouf, MD (Saudi Arabia); Mirjana Jovanovic, MD, Aleksandra Minic, MD, Srdjan Pasic, MD, Jelena Vojinovic, MD (Serbia and Montenegro); Helena Koskova, MD, Dagmar Mozolova, MD, Jozef Rovensky, MD, Veronika Vargova, MD, Richard Vesely, MD (Slovakia); Tadej Avcin, MD (Slovenia); Immaculada Calvo, MD, Julia Garcia Consuegra, MD, Consuelo Modesto, MD, Maria Rosa Roldan Molina, MD (Spain); Anders Fasth, MD, Stefan Hagelberg, MD, Bo Magnusson, MD (Sweden); Michael Hofer, MD, Traudel Saurenmann, MD, Marie-Josephe Sauvain, MD (Switzerland); Rym Hajri Ben Ammar, MD (Tunisia); Aysin Bakkaloglu, MD, Huri Ozdogan, MD, Rezan Topaloglu, MD (Turkey); Virginia Brown, Helen Foster, MD, Madeleine Rooney, MD, Pat Woo, MD (UK); Edward H. Giannini, MSc, DrPH, Gloria Higgins, MD, Sampath Prahalad, MD, Marilyn Punaro, MD, Anne Reed, MD, Robert Rennebohm, MD, Peter Reuman, MD (US). REFERENCES 1. Strand V, Gladman D, Isenberg D, Petri M, Smolen J, Tugwell P. Outcome measures to be used in clinical trials in systemic lupus erythematosus. J Rheumatol 1999;26:490–7. 2. Smolen JS, Strand V, Cardiel M, Edworthy S, Furst D, Gladman D, et al. Randomized clinical trials and longitudinal observational studies in systemic lupus erythematosus: consensus on a preliminary core set of outcome domains. J Rheumatol 1999;26:504–7. 3. American College of Rheumatology Ad Hoc Committee on Systemic Lupus Erythematosus Response Criteria. The American College of Rheumatology response criteria for systemic lupus erythematosus clinical trials: measures of overall disease activity. Arthritis Rheum 2004;50:3418–26. 4. Tucker LB, Menon S, Schaller JG, Isenberg DA. Adult- and childhood-onset systemic lupus erythematosus: a comparison of onset, clinical features, serology, and outcome. Br J Rheumatol 1995;34:866–72. 5. Font J, Cervera R, Espinosa G, Pallares L, Ramos-Casals M, Jimenez S, et al. Systemic lupus erythematosus (SLE) in childhood: analysis of clinical and immunological findings in 34 patients and comparison with SLE characteristics in adults. Ann Rheum Dis 1998;57:456–9. 6. Carreno L, Lopez-Longo FJ, Monteagudo I, Rodriguez-Mahou M, Bascones M, Gonzalez CM, et al. Immunological and clinical differences between juvenile and adult onset of systemic lupus erythematosus. Lupus 1999;8:287–92. 7. Brunner HI, Silverman ED, To T, Bombardier C, Feldman BM. Risk factors for damage in childhood-onset systemic lupus erythematosus: cumulative disease activity and medication use predict disease damage. Arthritis Rheum 2002;46:436–44. 8. Ravelli A, Duarte-Salazar C, Buratti S, Reiff A, Bernstein B, Maldonado-Velazquez MR, et al. Assessment of damage in juvenile-onset systemic lupus erythematosus: a multicenter cohort study. Arthritis Rheum 2003;49:501–7. 9. Ruperto N, Martini A. International research networks in pediatric rheumatology: the PRINTO perspective. Curr Opin Rheumatol 2004;16:566–70. 10. Giannini EH, Ruperto N, Ravelli A, Lovell DJ, Felson DT, 2863 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Martini A. Preliminary definition of improvement in juvenile arthritis. Arthritis Rheum 1997;40:1202–9. Ruperto N, Ravelli A, Falcini F, Lepore L, de Sanctis R, Zulian F, et al. Performance of the preliminary definition of improvement in juvenile chronic arthritis patients treated with methotrexate. Ann Rheum Dis 1998;57:38–41. ACR formally adopts improvement criteria for juvenile arthritis. ACR News 2002;21:3. Ruperto N, Ravelli A, Murray KJ, Lovell DJ, Andersson-Gare B, Feldman BM, et al. Preliminary core sets of measures for disease activity and damage assessment in juvenile systemic lupus erythematosus and juvenile dermatomyositis. Rheumatology (Oxford) 2003;42:1452–9. Delbecq AL, van de Ven AH, Gustafson DH. Group techniques for program planning: a guide to nominal group and Delphi processes. 1st ed. Glenview (IL): Scott, Foresman and Company; 1975. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1271–7. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997;40:1725. Bombardier C, Gladman DD, Urowitz MB, Caron D, Chang CH, and the Committee on Prognosis Studies in SLE. Derivation of the SLEDAI: a disease activity index for lupus patients. Arthritis Rheum 1992;35:630–40. Brunner HI, Feldman BM, Bombardier C, Silverman ED. Sensitivity of the Systemic Lupus Erythematosus Disease Activity Index, British Isles Lupus Assessment Group Index, and Systemic Lupus Activity Measure in the evaluation of clinical change in childhood-onset systemic lupus erythematosus. Arthritis Rheum 1999;42:1354–60. Landgraf JM, Abetz L, Ware JE. The CHQ user’s manual. 1st ed. Boston: The Health Institute, New England Medical Center; 1996. Ruperto N, Ravelli A, Pistorio A, Malattia C, Cavuto S, GadoWest L, et al. Cross-cultural adaptation and psychometric evaluation of the Childhood Health Assessment Questionnaire (CHAQ) and the Child Health Questionnaire (CHQ) in 32 countries: review of the general methodology. Clin Exp Rheumatol 2001;19:S1–9. Liang MH, Socher SA, Larson MG, Schur PH. Reliability and validity of six systems for the clinical assessment of disease activity in systemic lupus erythematosus. Arthritis Rheum 1989;32: 1107–18. Vitali C, Bencivelli W, Isenberg DA, Smolen JS, Snaith ML, Sciuto M, et al. Disease activity in systemic lupus erythematosus: report of the Consensus Study Group of the European Workshop for Rheumatology Research. II. Identification of the variables indicative of disease activity and their use in the development of an activity score. Clin Exp Rheumatol 1992;10:541–7. Vitali C, Bencivelli W, Mosca M, Carrai P, Sereni M, Bombardieri S. Development of a clinical chart to compute different disease activity indices for systemic lupus erythematosus. J Rheumatol 1999;26:498–501. Brunner HI, Silverman ED, Bombardier C, Feldman BM. European consensus lupus activity measurement is sensitive to change in disease activity in childhood-onset systemic lupus erythematosus. Arthritis Rheum 2003;49:335–41. Milligan SE, Hom DL, Ballou SP, Persse LJ, Svilar GM, Coulton CJ. An assessment of the Health Assessment Questionnaire functional ability index among women with systemic lupus erythematosus. J Rheumatol 1993;20:972–6. Gladman DD, Urowitz MB, Ong A, Gough J, MacKinnon A. A comparison of five health status instruments in patients with systemic lupus erythematosus (SLE). Lupus 1996;5:190–5. Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of 2864 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. health status in children with juvenile rheumatoid arthritis. Arthritis Rheum 1994;37:1761–9. Laposata M. SI unit conversion guide. Boston: NEJM books; 1992. Boers M, Brooks P, Strand CV, Tugwell P. The OMERACT filter for outcome measures in rheumatology. J Rheumatol 1998;25: 198–9. Bellamy N. Clinimetric concepts in outcome assessment: the OMERACT filter. J Rheumatol 1999;26:948–50. Liang MH, Fossel AH, Larson MG. Comparisons of five health status instruments for orthopedic evaluation. Med Care 1990;28: 632–42. Mosteller F, Tukey JW. Data analysis and regression. Reading (MA): Addison Wesley; 1977. Liang MH, Larson MG, Cullen KE, Schwartz JA. Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research. Arthritis Rheum 1985;28:542–7. Liang MH. Evaluating measurement responsiveness. J Rheumatol 1995;22:1191–2. Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960;20:37–46. Landis JR, Koch GC. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74. Cronbach LJ. Coefficient alfa and the internal structure of tests. Psychometrika 1951;16:297–334. Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978;8:283–98. Ruperto N, Ravelli A, Oliveira S, Alessio M, Mihaylova D, Pasic S, et al. The PRINTO criteria for the evaluation of response to therapy in juvenile systemic lupus erythematosus. II. Development RUPERTO ET AL 40. 41. 42. 43. 44. 45. 46. 47. 48. of the preliminary definition of improvement. Submitted for publication. Boumpas DT, Balow JE. Outcome criteria for lupus nephritis trials: a critical overview. Lupus 1998;7:622–9. Schiffenbauer J, Hahn B, Weisman MH, Simon LS. Biomarkers, surrogate markers, and design of clinical trials of new therapies for systemic lupus erythematosus. Arthritis Rheum 2004;50:2415–22. Walport MJ. Complement: first of two parts. N Engl J Med 2001;344:1058–66. Walport MJ. Complement: second of two parts. N Engl J Med 2001;344:1140–4. Carr A, Hewlett S, Hughes R, Mitchell H, Ryan S, Carr M, et al. Rheumatology outcomes: the patient’s perspective. J Rheumatol 2003;30:880–3. Brunner HI, Giannini EH. Health-related quality of life in children with rheumatic diseases. Curr Opin Rheumatol 2003;15: 602–12. Isenberg D, Ramsey-Goldman R. Assessing patients with lupus: towards a drug responder index. Rheumatology (Oxford) 1999;38: 1045–9. Ruperto N, Buratti S, Duarte-Salazar C, Pistorio A, Reiff A, Bernstein B, et al. Health-related quality of life in juvenile-onset systemic lupus erythematosus and its relationship to disease activity and damage. Arthritis Rheum 2004;51:458–64. Ad Hoc Working Group on Steroid-Sparing Criteria in Lupus. Criteria for steroid-sparing ability of interventions in systemic lupus erythematosus: report of a consensus meeting. Arthritis Rheum 2004;50:3427–31.