This study is the first to compare treatment of JIA in two different health care systems. Both these systems have universal health care. However, treatment modalities are reimbursed to varying degrees in the two systems, with medications being part of the publicly funded system in Germany but mostly covered by private insurance in Canada. The oligoarticular and seronegative polyarticular subtypes represented in this survey make up more than 50% of JIA cases in Canada and Germany[15, 16]. With close to two thirds of polled physicians answering, this survey constitutes a representative sample of the practicing physician population in paediatric rheumatology in their respective countries, and a large majority of respondents had both subspecialist training and significant experience in the field.
Working with vignettes is a sophisticated and valid method for measuring the quality of care and medical decision-making[17, 18]. Geographic variations in health care depending on the system have been shown in various studies[19, 20]. Compared to other findings, the approach to medical treatment of JIA observed in this study is rather uniform[8, 10, 21]. The largest variations were found in switching to additional DMARDs or biologics after failing the first DMARD in the treatment of seronegative polyarthritis (Table 2). Nonetheless, the majority of respondents followed published evidence and guidelines. This reflects the increasing numbers of international clinical trials with uniform standard of practice; also, standardisation of paediatric rheumatology training programs has shaped the approach to treatment. The efforts of the professional societies for paediatric rheumatology in their respective countries to standardise treatment approaches by guidelines and recommendations represent a continuation of this trend.
Drug availability may have played a major part in the observed differences in medication preferences between the two health care systems. German and Austrian paediatric rheumatologists were more willing to use DMARDs in refractory oligoarticular JIA, as well as initial treatment of polyarticular JIA. They were also more ready to switch to biologic agents after failure of a single DMARD in polyarticular JIA, and possibly at an earlier time. In Canada, funding agencies frequently demand the patient to fail more than one DMARD before approving treatment with biologic agents, despite lack of evidence for this approach. By contrast, both Germany and Austria have a state-governed system of mandatory health insurance which covers all licenced medications including biological without prior approval from the insurance companies (off-label uses are, however, exempt from this). Given that etanercept and adalimumab were both licensed for treatment of refractory JIA in Germany at the time of the survey, while leflunomide (as an alternative DMARD) was not, the observed differences are understandable.
Use of intra-articular steroids was remarkable, because only half of the physicians in both health care systems opted to treat an initial presentation of JIA this way. Intra-articular joint injections have been demonstrated to be of significant benefit in oligoarticular JIA, where as many as 70% of patients show no reactivation of disease for at least one year[23, 24]. The patient presented in the questionnaire was described as having four active joints (both knee and ankle joints), which could explain the reluctance of practitioners. However, practice variations even in JIA presenting as monoarthritis have been described previously and have been attributed to both views of effectiveness of intra-articular steroids and proficiency with the procedure.
A striking difference is presented in the role of physiotherapy in the treatment of JIA between the two health care systems (Table 1). German/Austrian physicians showed a clear preference for regular weekly physiotherapy in a controlled setting, while most Canadian physicians were content with recommending a home-exercise program for the patient. In an open study on 25 children with polyarticular JIA, an 8-week physical conditioning program led to significant improvement in joint symptoms. At least moderate adherence to an exercise program was associated with better function and less pain due to arthritis in one study on 175 Canadian children with JIA. Exercise training in a randomised controlled trial in 80 children with JIA resulted in self-reported improved function measured by CHAQ, but not other tests. However, a recent Cochrane review showed no statistically significant improvement of functional ability, quality of life or aerobic capacity by exercise programs. The small number of open studies on land- and water-based exercise have been reviewed, leading to the suggestion that participation in a physiotherapy program at least twice a week for at least 6 weeks may help to reduce disease symptoms. However, no study has directly compared efficacy of exercise programs versus controlled physiotherapy in JIA. A strong emphasis is placed on physiotherapy in the German/Austrian health care system especially; the German guidelines strongly state that ‘without adequate physiotherapy […], the treatment of patients with JIA is impossible’. This represents a fundamental difference from the approach to physiotherapy in Canada. However, it is unclear if this is also influenced by availability of the treatment modality; in Germany and Austria, health insurance usually covers physiotherapy in children, while in Canada, services are frequently privately funded or provided by non-government organisations such as the Arthritis Society.
This survey was limited by unequal sample and population sizes, with the German/Austrian respondent group being approximately three times larger. No unifying body comparable to the GKJR could be readily surveyed in Canada. As virtually all paediatric rheumatologists in Canada were members of CAPRI at the time of the survey, this was chosen as a substitute. The higher percentage of Canadian respondents working in academic centres is therefore not surprising. Differences in level of training and practice setting could also reflect the fact that paediatric rheumatology as a subspecialty has been introduced in most provinces of Germany and Austria only in the last decade and is not represented in all academic centres, while it has been recognized as a subspecialty in Canada since 1997. Nonetheless, both physician groups were comparable for age and experience. By design this survey could not address the fact that many children with JIA are treated by paediatricians and adult rheumatologists with no subspecialty training. Germany and Austria were considered having a sufficiently similar health care system and paediatric rheumatology tradition to group them together for purposes of statistical analysis. However, there are some differences especially in availability of certain drugs that might have influenced some choices. Numbers of Austrian paediatric rheumatologists were still too small for subgroup analysis. Respondents were also possibly influenced in their treatment choices by the subsequent changes in the patient vignettes; the number of polled physicians was insufficient to present multiple versions of this survey. Particular care was taken to present patients in as neutral a manner as possible.