Skip Navigation


Rheumatology Advance Access originally published online on February 2, 2008
Rheumatology 2008 47(5):640-645; doi:10.1093/rheumatology/kem366
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
47/5/640    most recent
kem366v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Abrahamyan, L.
Right arrow Articles by Feldman, B. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Abrahamyan, L.
Right arrow Articles by Feldman, B. M.
Related Collections
Right arrow Vasculitis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Quality of randomized clinical trials in juvenile idiopathic arthritis

L. Abrahamyan1–3, S. R. Johnson1–3, J. Beyene1–4, P. S. Shah1,2,5 and B. M. Feldman1–5

1The Hospital for Sick Children, 2Department of Health Policy Management & Evaluation, 3Department of Child Health Evaluative Sciences, 4Department of Public Health Sciences and 5Department of Paediatrics, University of Toronto, Toronto, Ontario, Canada

Correspondence to: B. M. Feldman, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. E-mail: brian.feldman{at}sickkids.ca


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary data
 Acknowledgements
 References
 
Objectives. We evaluated the quality of randomized clinical trials (RCTs) of therapy for juvenile idiopathic arthritis (JIA) using an individual component approach and assessed temporal changes.

Methods. A systematic review of the literature was performed to identify all RCTs involving exclusively JIA patients. Two investigators independently assessed the identified articles for six quality indicators: generation of allocation sequence, allocation concealment, masking, intention-to-treat (ITT) analysis, dropout rates and clearly stated primary outcome.

Results. Fifty-two RCTs involving JIA patients were assessed. Generation of allocation sequence was unclear in 79% of the studies. Reporting of allocation concealment was adequate in only one-third of the studies. Masking was adequate in 73%, inadequate in 19% and unclear in 8% of the reports. ITT analysis was employed in 37% of the reports. Per-protocol analysis was used in 40% and in 23% the method was unclear. Most of the reports (67%) had dropout rates ≤20%. About half of the reports (n = 25) failed to show a significant effect of the experimental treatment. No significant associations were found between the study results and quality indicators. With the exception of adequate masking and dropout rate, all quality indicators showed a trend of improvement over the decades.

Conclusions. The quality of RCTs in JIA based on the selected indicators was poor. Although there were some positive changes over time, the reporting and methodological quality of trials should be improved. New, more powerful and acceptable RCT designs should be developed in this patient population.

KEY WORDS: Juvenile idiopathic arthritis, Randomized clinical trials, Systematic review, Quality


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary data
 Acknowledgements
 References
 
Juvenile idiopathic arthritis (JIA) is the current name used to describe several subsets of childhood arthritis; JIA includes the subsets previously called juvenile rheumatoid arthritis as well as other types of childhood arthritis. JIA is the most common rheumatological disease in children. The reported prevalence varies from 0.07 to 4.01 per 1000 children in different countries [1]. JIA has six categories, of which the most common are oligoarthritis (40–60% of patients), rheumatoid factor-negative polyarthritis (20–25% of patients) and systemic arthritis (10–20% of patients) [2]. The disease burden is considerable both in childhood and later in adulthood. Several studies have shown that more than one-third of the patients extend the disease into adulthood [3, 4] often with a significant level of disability [5], increased probability of arthroplasty [6], uveitis [7], osteoporosis [8], high treatment costs [9] and decreased quality of life [10].

A recent review of medical treatments of JIA summarized evidence from 34 randomized clinical trials (RCTs) conducted between 1966 and 2005 [2]. The authors concluded that despite the significant advances in the treatment of the disease, there is still a lack of evidence in treatment for several subtypes of JIA. The review outlined some of the design features of included RCTs (e.g. type of design, masking and length of treatment), but the quality of individual RCTs was not evaluated.

There are significant challenges in conducting RCTs in rare diseases such as JIA. In the study of rare diseases, a considerable amount of time may be needed to recruit adequate numbers of patients, which, in turn, can become a reason for discontinuing a study. In JIA, the disease progression can be complex and may require an individualized treatment approach. It can also take many years for a patient to develop an objective outcome such as disability. Moreover, in RCTs of paediatric diseases there may be ethical concerns such as an unwillingness to commit children to trials where they may receive a placebo [11]. Therefore, the challenges to adequately conducting RCTs in this population are considerable.

The quality of published RCTs in JIA has not been investigated. Thorough evaluation of the quality of individual studies will improve our understanding of the challenges in conducting methodologically sound RCTs and hopefully identify avenues for improvement. This will also serve to enhance subsequent systematic reviews and meta-analyses.

Our aim was to evaluate the quality of RCTs that examined all interventions (pharmacological and non-pharmacological) in JIA using an individual component approach. We hypothesized that the overall quality of RCTs in JIA was poor; however, there was an improvement in the quality of reported RCTs in JIA over the last three decades.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary data
 Acknowledgements
 References
 
Eligibility criteria and search strategy
We reviewed all RCTs of interventions (pharmacological and non-pharmacological) where the study population consisted exclusively of JIA patients irrespective of age and JIA subtype. We excluded quasi-randomized studies or studies that had a mixed population of JIA and other patients. If the same trial was published more than once with different follow-up periods or different outcome evaluations, we included only the study with the report of the principal analysis planned a priori. When the study had multiple phases (for example, began as an RCT and then was continued as open-label), only data from the RCT were extracted. We included articles published in the English language only. We excluded articles published only as an abstract.

We searched MEDLINE (1966 to December 2006), EMBASE (1980 to December 2006) and the Cochrane Central Register of Controlled Trials (CENTRAL) for all publications describing RCTs in JIA. In addition, the reference lists of identified trials were searched. Potentially relevant articles were first identified on the basis of their title, abstract and keywords by one of the authors (L.A.). Two authors (L.A. and S.R.J.) independently assessed the full text of selected articles and determined the set of trials for the final review. Medical Subject Headings (MeSH) terms for search included: juvenile rheumatoid arthritis; randomized controlled trial; controlled clinical trial; multicentre study; clinical trial; English; and human studies. The terms were adjusted according to database-specific terms.

Data extraction
We selected the important quality indicators to reflect areas where bias (selection bias, performance bias, measurement bias and attrition bias) may occur in RCTs [12]. A data collection form was developed based on the selected quality indicator components as recommended by Khan et al. [12] The form was pilot-tested in three RCTs involving patients who did not have JIA. Two independent reviewers (L.A. and S.R.J.) used the form to extract data from the selected JIA articles. Disagreements were solved by consensus and consulting the other authors (J.B., P.S.S. and B.M.F.).

We extracted the following data from the trials: first author's name, publication year, study design (parallel, cross-over), number of study arms, compared interventions, total sample size, mean age of patients, study duration (in weeks), allocation concealment (adequate, inadequate, unclear), generation of allocation sequence (adequate, inadequate, unclear), masking (adequate, inadequate, unclear, not applicable), intent-to-treat (ITT) analysis (full or modified ITT, per-protocol analysis, treatment-received analysis, unclear), dropout rate (exact rate; no dropout, ≤20%, >20%), methods used to handle missing data, clearly stated primary outcome(s) (yes, no), type of primary outcome(s) (single, composite, multiple) and description, participating centres (single, multicentre or multinational) and the final conclusion of the study. We also evaluated the study funding source (not sponsored by industry, industry sponsored, not reported). The study was considered as industry sponsored if the pharmaceutical manufacturer was acknowledged by the authors or if any of the authors was affiliated with a pharmaceutical company. The main quality indicators of interest were allocation concealment, generation of allocation sequence, masking, intention-to-treat (ITT) analysis, dropout rate and clearly stated primary outcome (Table 1). The definitions for the indicators were selected based on the Consolidated Standards of the Reporting of Trials (CONSORT) guideline and previous studies [13–15].


View this table:
[in this window]
[in a new window]

  		TABLE 1. Quality indicators and definitions

 
Data analysis
Descriptive statistical analyses were performed using mean, S.D. or median for continuous data and frequencies and percentages for categorical data. Continuous data were compared using t-test and analysis of variance; categorical data were compared using chi-square test or Fisher's exact test as appropriate. Two-tailed P-values were reported. Characteristics of the studies and quality indicators were described and the final results of the studies were compared with the quality items and funding source. Finally, we evaluated the change of quality indicators over time in decades defined by year of publication. SPSS 11.0 statistical software (SPSS, Inc., Chicago, IL, USA) was used for data entry and analysis.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary data
 Acknowledgements
 References
 
Search results
After removing the duplicates, the applied search strategy identified 471 abstracts in total. The abstracts presenting non-randomized, cross-sectional comparative or cohort clinical studies were excluded (n = 417). Fifty-four full-text articles underwent full review. One study was excluded because it was not a truly randomized trial [16]. Another excluded study described only the pharmacokinetics of the drug [17]. The final review included 52 articles (Supplementary Appendix). Figure 1 presents the selection process of articles for the final review.


Figure 1
View larger version (24K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. Flow diagram for selection of articles for the review.

 
General description of studies
Table 2 presents general characteristics of the reviewed studies. There were eight cross-over studies. The remaining 44 RCTs used a parallel-group study design. Among them two studies randomized each patient's joints to different treatments. Another three utilized a randomized withdrawal design, where patients entered into the double-masked study only after they had responded to the treatment during an initial open-label run-in phase. In one study, patients entered into the RCT only if they did not respond to the treatment in the open-label phase.


View this table:
[in this window]
[in a new window]

  		TABLE 2. General characteristics of RCTs

 
Forty-two percent of the studies included placebo as a comparison group. The remaining studies compared experimental and contemporaneous ‘standard treatment’ methods. Most of the studies compared medical treatments for JIA (n = 36). The other RCTs evaluated treatment of JIA complications such as uveitis (n = 1), growth retardation (n = 3), a surgical intervention (n = 1), orthopaedic splints and shoes (n = 2), local analgesia for intra-articular injections (n = 1), folic acid supplementation (n = 1), behavioural interventions to increase calcium intake (n = 2), effect of daily supplementation with calcium (n = 1), interventions for medication adherence (n = 1), aquatic training (n = 1), massage therapy (n = 1) and combined hydrotherapy and physiotherapy (n = 1). About half (52%) of the studies were single-centre studies and 44% (n = 23) of the studies were industry sponsored. In seven studies, pharmaceutical companies provided either the standard or the experimental treatment but it was unclear whether industry funded the study.

Table 2 shows the median sample size of the studies by design. The median sample size was 30 for cross-over RCTs (n = 8) and 49 for parallel RCTs (n = 44). The median sample size of parallel RCTs in medication treatment of JIA (n = 30) was 71. Other general characteristics of RCTs of medical treatment of JIA were quite similar to the results from all studies.

The overall mean or median age of patients in the included RCTs was impossible to evaluate because of the poor and inconsistent reporting in some studies. The range of the reported means was from 5.5 to 15.9 yrs if excluding one study reporting all patients being 25 yrs old.

Quality indicators
The generation of the allocation sequence and presence/absence of allocation concealment were unclear in the majority of RCTs (Table 3).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Generation of allocation sequence, allocation concealment and masking

 
There were 35 double-masked studies. In seven studies, it was reportedly impossible to mask patients from the interventions (such as massage therapy, orthotic devices and dietary interventions). For such studies, we considered that masking was adequate if the outcome assessors were masked (n = 3).

ITT analysis was employed in 19 studies (37%); full ITT in 12 studies and modified ITT in 7. Twenty-one (40%) studies conducted per-protocol analysis, while in 12 studies (23%) the approach was unclear. Eighteen RCTs (32.7%) had dropout rates of more than 20% of the original sample while six studies (12%) had no dropout. Overall, the mean dropout rate was 15% (S.D. = 13; range from 0% to 52%). In two studies, it was impossible to estimate the exact dropout rate because of the inconsistent number of patients for different measured outcomes. No significant correlation was found between the indicators of study duration and exact dropout rate (Pearson's correlation coefficient was 0.273; P > 0.05). Among 19 studies that used ITT, only five mentioned the method of handling missing data.

Only half of the reports clearly specified the primary outcome (n = 26). Among the 26 studies with clearly stated outcomes, a single primary outcome was evaluated in 16 cases and multiple outcomes in 10 of them. Multiple outcomes mostly included general clinical, articular and/or laboratory assessments. Overall, 16 studies (31%) measured at least one composite outcome out of which 13 were clearly stated primary outcomes.

About 70% of all RCTs achieved an adequate rating in a maximum of three quality indicators out of the evaluated six; similar results of achieved numbers of quality indicators were observed for RCTs when we considered only medication treatment (Fig. 2).


Figure 2
View larger version (25K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 2. Quality ranking of RCTs in JIA.

 
Quality indicators and study results
In 25 (48%) studies, no statistically significant difference was found between the experimental and control treatments. The experimental treatment was superior in 25 RCTs. In one non-inferiority trial, the experimental treatment was found to be not inferior to standard treatment. In another study, the experimental intervention had less efficacy compared with the standard treatment. In three studies, patients in the standard treatment (acetylsalicylic acid was used in all cases) had significant adverse events.

We did not find statistically significant association between the study results and the source of funding (P > 0.05, chi-square test; raw data not shown). Similarly, there was no statistically significant difference in the reported quality indicators between positive and negative studies.

Temporal trends of the quality indicators
In order to show the temporal trends, RCTs were grouped in 10-yr intervals by their year of publication. The first two RCTs in JIA patients were conducted in 1977. The time periods and respective number of RCTs were: 1976–85 (n = 11), 1986–95 (n = 17) and 1996–2006 (n = 24). Compared with the first decade, the number of RCTs had almost doubled during the time period between 1996 and 2006. Table 4 shows the changes of quality indicators over the three decades.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Time trends of quality indicators

 
With the exception of adequate masking and dropout rate, all quality indicators showed a trend for improvement over the decades. Compared with the second decade, the number of RCTs reporting adequate allocation concealment increased almost 5-fold, and the number of RCTs with clearly stated primary outcomes increased almost two times during the third decade. Temporal increases in ITT analysis between the first and third decades, and second and third decades were also statistically significant.


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary data
 Acknowledgements
 References
 
Rare disease trials face a number of methodological challenges in the hope of determining the efficacy of an intervention. Thus, assurance of methodological quality is a critical and necessary endeavour to ensure that limited resources are effectively utilized and trial results are appropriately translated into clinical practice. Our study evaluated 52 RCTs for JIA patients beginning from 1977 to present; a number much lower than might be expected when compared with more common diseases in the adult population. We identified that most of the quality components that we evaluated were seldom met adequately.

The generation of random number sequences was the most rarely reported quality indicator; this was mostly persistent across the three time periods. The overall reporting of allocation concealment was low; however, in contrast to random number generation, it had a 5-fold increase after 1995. In different diseases, the proportion of RCTs with reported adequate randomization methods and adequate allocation concealment varies from 1% to 52% and from 2% to 39%, respectively [18]. While the overall rates of these quality indicators have been low, significant changes in reporting are likely due to the publication of the CONSORT statement in 1996 [13]. A study that compared the quality of RCTs in adult rheumatology between two time periods, 1987–88 and 1997–98, suggested that among the studied quality indicators only the reporting of allocation concealment increased significantly [19]. The improvement of quality over time could also be explained by the establishment of international networks such as the Pediatric Rheumatology Collaborative Study Group (PRCSG) in 1973 and the Pediatric Rheumatology International Trials Organization (PRINTO) in 1996, which aim to conduct high-quality clinical trials in children with rheumatic diseases.

Although the effect of adequate masking on the study's main results is still controversial [18], masking is still one of the current standards for RCTs. Adequate masking should be guaranteed especially when the outcomes are susceptible to measurement bias. In JIA, masking is especially important considering the use of subjective outcomes such as patient or physician global assessments using visual analogue scales. In our study, double masking was used in 67.3% of the studies, and no positive temporal change was observed. This result is inconsistent with some other similar studies in the field. A study that assessed RCTs of rheumatological disorders in adults identified a higher rate of double masking of studies both between 1987 and 1988 (73.7%) and between 1997 and 1998 (85.5%) [19]. In contrast, in systemic lupus erythematosus only 54.3% of RCTs were double-masked [20]. The rate of double-masked studies might depend on the type of compared interventions. If we remove those studies in which double masking was thought to be unfeasible (n = 7), our number of RCTs with double masking rises to 77.8%.

The importance of clearly stated primary outcomes is very high in JIA and other systemic diseases, where outcomes are often subjective, and mortality is not a large concern. Depending on the evaluated treatment, the outcomes of interest can vary from study to study. In JIA, a validated measure of response to treatment was developed in 1997 [21]. The ACR Pediatric 30 score is a composite outcome consisting of six different measures, and is currently the most widely accepted method to estimate the response to treatment for JIA. Most probably, two factors, the publication of the CONSORT guideline and the development of a universal validated outcome, played a role in the significant increase in the number of RCTs in JIA with clearly stated outcomes. In the review of RCTs in systemic lupus erythematosus, adequately reported outcome was the only quality indicator that showed a significant improvement over time [20].

Although in this study the most frequently used approach was per-protocol analysis, a significant temporal increase of RCTs with ITT was observed. Application of ITT methods yields unbiased and valid estimates not compromising the balance between the groups achieved by randomization [22]. A study that evaluated ITT principles in superiority trials of structural outcomes in adult rheumatic diseases between 1994 and 2003 also found that most of the RCTs employed per-protocol analysis [14]. They reported that about 28.6% of the studies had >20% missing data. The method of handling missing data was reported in 23.5% of the studies and the most prevalent method was the last observation carried forward approach [14]. In our study, most of the RCTs had <20% dropout rate and there was no significant increase in the dropout rates over time. However, in our study, only five RCTs reported the method of handling missing data, four of which used the last observation carried forward approach and one used sensitivity analysis.

The main objective of our study was to evaluate the quality of RCTs in JIA, which was poor for most of the selected indicators. However, some of the observed characteristics of the studies highlighted the difficulties in conducting RCTs in this population. We found that the sample size was <50 in more than half of the included parallel RCTs. Some of the reviewed RCTs discussed explicitly the difficulties related to patient accrual. One of the studies enrolled only 25 patients over a 3-yr period from seven centres in the USA and Canada. Another study could not reach the needed sample size of 15 patients in the planned 18 months and was stopped early. Almost half of the studies failed to show a significant difference between the compared treatments—in some cases likely due to small sample sizes and low power. The results were neither associated with the levels of quality indicators nor with the source of funding. Lack of objective outcomes is another difficulty in this population. Although the ACR Pediatric 30 is a validated response measurement, it is a composite outcome and the clinical interpretation of the results can be challenging. Moreover, depending on the compared treatment methods (e.g. medication, massage therapy, behavioural interventions, surgery) the optimal outcome measurement may not be the ACR Pediatric 30. About half of the studies were either multicentre or multinational as a way to facilitate patient accrual. Alternative designs were also employed to make the studies more acceptable and more powerful. One-fifth of the RCTs in JIA used non-traditional designs such as cross-over and randomized withdrawal designs.

Study quality is relevant to the interpretation of study results. Several studies showed that the low quality studies are more likely to yield biased results [23, 24]. In our study, we did not see significant associations between the quality indicators and the study results. However, this type of conclusion depends also on the number of included studies (n = 52), which is relatively small compared with previous studies.

One of the limitations of our study was that only English language and full-text articles were included. We believe that this will not likely bias the results of the study because when the disease is rare and it is not easy to conduct an RCT, there should likely be more desire to report the results in English language journals. Moreover, the inclusion or exclusion of non-English language articles is felt by methodologists to be more of an issue when the systematic review is focused on treatment effectiveness rather than study quality [25]. Another limitation might be the reporting quality of the studies. It is possible that reporting quality is not always a good measure of the actual methodological quality of a study [26]. For example, we accepted masking as adequate if the authors named or described it as such. We categorized allocation concealment or method of random number generation as unclear if no information was available from the articles. This deficiency can be related to other factors related to reporting (e.g. journal word count restrictions) rather than the methodological quality of RCTs. In contrast, some of the components we have measured were directly related to methodological quality (such as ITT and dropout rate). However, the methodological and reporting quality of the studies are usually thought to be highly related [27]. Due to the constraints of feasibility we did not evaluate the methodological quality of RCTs by JIA subtype. Future investigators should be aware that the response to the treatment and the choice of the appropriate outcome can differ between the different subtypes. This is a further complicating issue for RCTs for this disease.

In summary, the overall quality of RCTs in JIA has been poor and there have been very few RCTs. As a result, there is little high-quality evidence for many current practices in JIA. There are positive changes in some of the indicators over time but more efforts should be made for further improvements of the design, conduct and reporting of the studies. In particular, studies should report the method of random number generation and allocation concealment, should make studies double masked whenever possible, conduct ITT analysis and ensure complete follow-up as far as possible. Considering the current difficulties in conducting RCTs in this population—and the likelihood that studies will continue to be relatively few—both patients and physicians will benefit if more powerful and acceptable study designs will be used in the future. Perhaps this can be achieved by the development and testing of new designs for intervention trials for rare diseases.

Formula


   Supplementary data
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary data
Acknowledgements
 References
 
Supplementary data are available at Rheumatology Online.


   Acknowledgements
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary data
 Acknowledgements
References
 
L.A. has been awarded a Canadian Arthritis Network (CAN) Graduate Student Award and Graduate Studentship from the Research Training Centre of the Hospital for Sick Children, Toronto, Canada. S.R.J. has been awarded an Abbott Scholar Award and CAN Fellowship. B.M.F. is supported by a Canada Research Chair in Childhood Arthritis.

Disclosure statement: The authors have declared no conflicts of interest.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary data
 Acknowledgements
 References
 

  1. Manners PJ, Bower C. Worldwide prevalence of juvenile arthritis why does it vary so much? J Rheumatol (2002) 29:1520–30.[Abstract/Free Full Text]
  2. Hashkes PJ, Laxer RM. Medical treatment of juvenile idiopathic arthritis. JAMA (2005) 294:1671–84.[Abstract/Free Full Text]
  3. Zak M, Pedersen FK. Juvenile chronic arthritis into adulthood: a long-term follow-up study. Rheumatology (2000) 39:198–204.[Abstract/Free Full Text]
  4. Minden K, Niewerth M, Listing J, et al. Long-term outcome in patients with juvenile idiopathic arthritis. Arthritis Rheum (2002) 46:2392–401.[CrossRef][Web of Science][Medline]
  5. Packham JC, Hall MA. Long-term follow-up of 246 adults with juvenile idiopathic arthritis: functional outcome. Rheumatology (2002) 41:1428–35.[Abstract/Free Full Text]
  6. Oen K, Malleson PN, Cabral DA, Rosenberg AM, Petty RE, Cheang M. Disease course and outcome of juvenile rheumatoid arthritis in a multicenter cohort. J Rheumatol (2002) 29:1989–99.[Abstract/Free Full Text]
  7. Guillaume S, Prieur AM, Coste J, Job-Deslandre C. Long-term outcome and prognosis in oligoarticular-onset juvenile idiopathic arthritis. Arthritis Rheum (2000) 43:1858–65.[CrossRef][Web of Science][Medline]
  8. Oen K. Long-term outcomes and predictors of outcomes for patients with juvenile idiopathic arthritis. Best Pract Res Clin Rheumatol (2002) 16:347–60.[Medline]
  9. Minden K, Niewerth M, Listing J, Biedermann T, Schontube M, Zink A. Burden and cost of illness in patients with juvenile idiopathic arthritis. Ann Rheum Dis (2004) 63:836–42.[Abstract/Free Full Text]
  10. Foster HE, Marshall N, Myers A, Dunkley P, Griffiths ID. Outcome in adults with juvenile idiopathic arthritis: a quality of life study. Arthritis Rheum (2003) 48:767–75.[CrossRef][Web of Science][Medline]
  11. Lehman TJ. The future of pediatric rheumatology: many questions remain. Arthritis Rheum (2007) 56:2815–6.[CrossRef][Medline]
  12. Khan KS, Kunz R, Kleijnen J, Antes G. Systematic reviews to support evidence-based medicine: how to review and apply findings of healthcare research. Royal Society of Medicine Press, London (UK) (2003).
  13. Altman DG. Better reporting of randomised controlled trials: the CONSORT statement. Br Med J (1996) 313:570–1.[Free Full Text]
  14. Baron G, Boutron I, Giraudeau B, Ravaud P. Violation of the intent-to-treat principle and rate of missing data in superiority trials assessing structural outcomes in rheumatic diseases. Arthritis Rheum (2005) 52:1858–65.[CrossRef][Web of Science][Medline]
  15. Juni P, Altman DG, Egger M. Systematic reviews in health care: assessing the quality of controlled clinical trials. Br Med J (2001) 323:42–6.[Free Full Text]
  16. Zulian F, Martini G, Gobber D, Agosto C, Gigante C, Zacchello F. Comparison of intra-articular triamcinolone hexacetonide and triamcinolone acetonide in oligoarticular juvenile idiopathic arthritis. Rheumatology (2003) 42:1254–9.[Abstract/Free Full Text]
  17. Makela AL, Lempiainen M, Yrjana T, Haapasaari J. Benorylate in the treatment of children with juvenile rheumatoid arthritis. Scand J Rheumatol Suppl (1979) (25):1–7.
  18. Gluud LL. Bias in clinical intervention research. Am J Epidemiol (2006) 163:493–501.[Abstract/Free Full Text]
  19. Hill CL, LaValley MP, Felson DT. Secular changes in the quality of published randomized clinical trials in rheumatology. Arthritis Rheum (2002) 46:779–84.[CrossRef][Web of Science][Medline]
  20. Karassa FB, Tatsioni A, Ioannidis JP. Design, quality, and bias in randomized controlled trials of systemic lupus erythematosus. J Rheumatol (2003) 30:979–84.[Abstract/Free Full Text]
  21. Giannini EH, Ruperto N, Ravelli A, Lovell DJ, Felson DT, Martini A. Preliminary definition of improvement in juvenile arthritis. Arthritis Rheum (1997) 40:1202–9.[Web of Science][Medline]
  22. Heritier SR, Gebski VJ, Keech AC. Inclusion of patients in clinical trial analysis: the intention-to-treat principle. Med J Aust (2003) 179:438–40.[Web of Science][Medline]
  23. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA (1995) 273:408–12.[Abstract/Free Full Text]
  24. Moher D, Pham B, Jones A, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet (1998) 352:609–13.[CrossRef][Web of Science][Medline]
  25. Moher D, Pham B, Lawson ML, Klassen TP. The inclusion of reports of randomised trials published in languages other than English in systematic reviews. Health Technol Assess (2003) 7:1–90.[Medline]
  26. Soares HP, Daniels S, Kumar A, et al. Bad reporting does not mean bad methods for randomised trials: observational study of randomised controlled trials performed by the Radiation Therapy Oncology Group. Br Med J (2004) 328:22–4.[Abstract/Free Full Text]
  27. Huwiler-Muntener K, Juni P, Junker C, Egger M. Quality of reporting of randomized trials as a measure of methodologic quality. JAMA (2002) 287:2801–4.[Abstract/Free Full Text]
Submitted 20 July 2007; revised version accepted 7 December 2007.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Rheumatology (Oxford)Home page
J. Thornton, M. W. Beresford, and P. Clayton
Improving the evidence base for treatment of juvenile idiopathic arthritis: the challenge and opportunity facing the MCRN/ARC Paediatric Rheumatology Clinical Studies Group
Rheumatology, May 1, 2008; 47(5): 563 - 566.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
47/5/640    most recent
kem366v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Abrahamyan, L.
Right arrow Articles by Feldman, B. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Abrahamyan, L.
Right arrow Articles by Feldman, B. M.
Related Collections
Right arrow Vasculitis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?