In recent decades, the approach to the management of rheumatoid arthritis (RA) has changed the prognosis of this disease. A large body of evidence shows that early treatment (within the first 3 months) is more effective and is associated with higher remission rates, absence of radiological progression and less severe disability.1–5

The diagnosis and therapy vary widely in RA for a number of reasons, including patient delay in presenting to the primary care (PC) physician, delayed referral from PC to a specialist, and delay in being seen by a rheumatologist once the referral has been produced.6 According to the Spanish emAR study (Estudio sobre el manejo de la arteritis reumatoide [Study of the management of rheumatoid arthritis]) published in 2007, the lag time between symptom onset and the initiation of effective treatment on the part of a rheumatologist has gradually been reduced over the past few decades, but it is still far from ideal.7,8

In the context of the presentation of the master plan for musculoskeletal diseases by the health department of the autonomous government of Catalonia, in northeastern Spain, (2011–2015), a working group was created to develop improvements in the early diagnosis of RA in that region. An epidemiological survey was proposed to a number of rheumatology departments (the AUDIT study) to determine: (a) the diagnostic delay (“time to diagnosis”) and treatment delay (“time to treatment”) of RA patients to analyze the lag times observed from symptom onset to the first treatment with disease-modifying antirheumatic drugs (DMARDs); and (b) whether the application of certain health care mechanisms in these rheumatology departments had an impact on the possibilities of early diagnosis and treatment of the disease.

We conducted a cross-sectional study during 2011 and 2012, based on an epidemiological survey administered to 10 consecutive patients from 19 centers in Catalonia who had been diagnosed with RA in 2009 or 2010. Epidemiological data of the patients (age, sex, race, city or town of residence) were recorded, as were the number and types of DMARDs received by each. We analyzed the prevalence of rheumatoid factor (RF) and anti-cyclic citrullinated peptide antibodies (anti-CCP), and the presence of erosions at the time of diagnosis and the time of this writing. In addition, the time variables were analyzed in relation to the delay in patient care: (1) from the onset of the first symptom to the first visit to a rheumatologist; (2) from referral to the first visit to a rheumatologist; (3) from the onset of the first symptom to the definitive diagnosis; and (4) from the onset of the first symptom to the administration of the first DMARD. For each center, we indicated whether or not it offered any of 6 rheumatology care mechanisms or features: (1) dedicated early arthritis (EA) clinic; (2) dedicated RA clinic; (3) mechanisms for early appointment scheduling; (4) algorithms for referral from PC; (5) rheumatology unit in the PC center; and (6) counseling service in PC.

Statistical methods: Student's t test was used for the comparison of means. The chi-squared test and Spearman correlation were utilized to analyze the correlation between the different rheumatology care features. The relationship between lag time from symptom onset to initiation of the first DMARD and the different rheumatology care features was analyzed by linear regression. The probability of a lag time of less than 6 or 12 months was analyzed using multivariate logistic regression. Those features that were marginally significant in univariate analyses (P<.1) were included as explanatory variables in the regression models. As these variables were correlated, they were grouped as a single variable that indicated the number of rheumatology care features that had been used by the patients (0–3). Statistical significance was established at P<.05. The analyses were performed with the SPSS software package (version 18).

A total of 183 patients were included (132 women) with a mean age of 57.2±14 years (range: 22–87), with a disease history of 27.3±20 months. The disease began as palindromic rheumatism in 15.8%. Positivity for RF was detected in 73.2% and for anti-CCP in 68.6%. Only 6% (11 patients) had rheumatoid nodules; 32 patients (17.5%) had erosions on radiography at the time of diagnosis and 47 (25.7%) at the time of this writing. The DMARD most frequently administered was methotrexate, in 122 patients (73.5%), followed by leflunomide in 13%; 34 patients (18.7%) were being treated with biologic agents. In all, 71% (130 patients) had been referred from PC, 8.7% from the emergency department and the remaining 20%, from other medical services and the orthopedic department.

The time elapsed from the onset of the first symptom to the first visit to a rheumatologist was 10.2±12.7 months, to the time of diagnosis of RA, 11.3±13.2months, and to the first DMARD, 11.1±12.8 months. The time from referral to the first visit to a rheumatologist was 3.6±5.8months. In all, 34.4% of the patients had access to a dedicated EA clinic and 37.2% to a dedicated RA clinic; early appointment scheduling was applied in 66.1% and referral algorithms in 31.1%; 31.7% were seen in a PC rheumatology department and 61.7% in a counseling or consultation service.

The existence of the above features was associated with a shorter lag time between onset of symptoms and initiation of the first DMARD, although only access to an EA clinic was significant. This clinic was also associated with a shorter lag time from symptom onset to the first visit to a rheumatologist (7.5±10months vs 11.6±13.7months; P=.016) and to the diagnosis of RA (8.8±11.4months vs 12.6±13.7months; P=.046) (Table 1). The existence of dedicated RA clinics was associated with a shorter time from onset of the first symptom to the confirmation of the diagnosis of RA (P=.040). The availability of counseling in the PC center was significantly associated with a shorter delay from the onset of symptoms to the first visit to a rheumatologist and to diagnosis, but not to the first DMARD. The same was observed in PC centers that had a rheumatology department although, in this case, the association with a shorter lag time following referral was very significant (Table 2). The evaluation of the 6 possible rheumatology care features showed that only the existence of dedicated EA clinics was associated with a closer proximity to treatment of RA, with a shorter time from onset of the first symptom (months) to the initiation of the first DMARD (Table 3).

For multiple regression analysis, we excluded the features that had no statistically significant relationship to the lag time from symptom onset and the initiation of the first DMARD, and included the 3 that did. As these features were correlated, they were grouped in terms of the number of patients who had had access to 1, 2 or 3 of these features. The results, shown in Table 4, indicate that the patients with access to only 1 feature had the same lag time as the patients who had not had access to any of them. However, when the patients had had access to 2 or 3 of these features, the lag time was significantly shorter (P=.016) or showed a trend toward significance (P=.107), with reductions of between 6 and 5 months. Logistic regression analysis showed that the patients with access to 2 or 3 of these features had a 2.6 (95% confidence interval [CI], 1.3–5.6) times greater probability of having a lag time of less than 12 months (P=.01) and a 1.8 (95% CI, 1–3.4) times greater probability of having a lag time of less than 6 months (P=.07).

Successful early control of RA depends, first, on the patients, who must realize that they have the disease, assimilate the signs and symptoms and visit their family physicians. A second limitation corresponds to the delay in referral from their PC physicians to rheumatology specialists. It is in this second barrier that the delay produced can be excessive. There may be a variety of reasons, but the existence of mechanisms for urgent care and well-defined referral criteria are directly associated with more adequate care.9

In 2007, Clemente et al. analyzed the emAR Spanish cohort by reviewing the medical records of 865 patients, showing that the median time from the onset of symptoms to the first DMARD treatment was 14 months. However, they observed a reduction in the time to first DMARD over the past 2 decades (median 8 months in 1995–1999, the last period evaluated), due mostly to a reduction of the time to a visit to a specialist. Their findings contrast with those of our AUDIT study, in which we have reduced the time from referral to the visit to a rheumatologist, shortening the time to the initiation of the first DMARD to less than 1 month.7 Despite these favorable data, the lag time between the onset of the first symptom to the visit to a rheumatologist continues to be very long, slightly greater than that reported by other authors.9–15

There have been attempts to identify other explanations for differences in patient management, beyond the existing mechanisms. Kumar et al. observed that the delay in delivery of care to their cohort of 169 patients was mostly due to factors depending on the perception of the disease itself.10 Later, in 2012, an observational study in 10 representative centers in 8 European countries concluded that the delay in delivery of care to RA patients in Europe was multifactorial and unacceptably long.11 In 2002, Palm and Purinszky attempted to show sex-related differences to explain why the 59 women with EA were referred later, on average, than men, but found no objective data.12

In 2010, Van der Linden et al. evaluated the cause of delay in a cohort from Leiden, classifying the causes into 2 subgroups: those due to the patients themselves and those related to PC physicians. The factors that determined the delay in being assessed were mostly clinical, such as the insidious onset of the disease, normal acute-phase reactant levels, symmetry and small joint involvement.13 In our case, we demonstrated that closer proximity to PC physicians through counseling services favors direct interchange and referral; in contrast, the latter is delayed in centers that do not have mechanisms of urgent care.

The study of the French ESPOIR cohort, reported in 2010, evaluated the time to access to a rheumatologist in 814 patients with EA, comparing it with the European League Against Rheumatism (EULAR) recommendations, which indicate that it should be less than 6 weeks. The authors confirmed the difficulty in complying with the recommendations, and only 46.2% visited a rheumatologist within that period. As in our study and in another cohort from a metropolitan area in Canada, the differences observed were due to the barrier constituted by PC but, in contrast to the French cohort, the majority of our centers do not have features like arthritis units, a circumstance that adds to the lag time.14,15

The literature on inflammatory arthritis has placed little emphasis on factors that are less specific, but are of great importance in the overall management of arthritis, such as the time lost due to the patients’ failure to perceive the disease, the time it takes to be seen by a specialist and the lack of specific mechanisms of urgent or preferential care, like EA clinics. This situation was referred to in 2011 by the SERAP working group; however, our report demonstrates that 2 years later, the development of these clinics in Spain was still inadequate, at least in our geographic region.16

The AUDIT study is the first to analyze the importance of the existence of dedicated EA and RA clinics, and evaluates these features. Quality care in RA should be based on closer proximity to PC and on the creation of dedicated clinics, taking into account that, even in those centers in which they are already established, the lag time is still too long.

Nevertheless, our study has certain limitations. Firstly, although the sample is substantial, it is not a large cohort. Secondly, the sample was not randomized, the participating centers vary widely and the study did not evaluate sociodemographic or cultural factors. Thirdly, the great majority of the centers involved do not have EA clinics, a circumstance that results in a loss of homogeneity. However, we should point out that one of the strengths of the AUDIT study is that it represents diverse care models from different geographic regions, while it also provides a survey of real-world cases that are found in clinical practice. Finally, the close correlation observed between the presence of different rheumatology care features makes an exact interpretation of the multivariate analysis difficult, although the presence of EA clinics is identified as the most relevant factor in reducing the lag time from symptom onset to the first DMARD.

The AUDIT study confirms the suspicion that the rates of earlier diagnosis and treatment are higher in centers that have dedicated EA clinics or, in their absence, dedicated RA clinics. The times to care delivery and first DMARD treatment are shorter in those centers and result in better management of these inflammatory diseases. We consider that, in Spain, the implementation of improved referral criteria, the design of better defined algorithms and, above all, the creation of dedicated EA and RA clinics would result in earlier and more effective diagnosis and treatment of RA.

Project financed by Merck Sharp & Dohme de España.

All the authors have contributed to the same extent to the project, which was promoted and sponsored by the Societat Catalana de Reumatologia (SCR).

The authors declare they have no conflicts of interest.