There were patients referred by the primary care doctor to the Hospital, patients from the emergency department or referred by other hospital units. Data was collected between January 2013 and April 2014. The baseline characteristics, included sex, age, history of smoking and alcohol intake and medical comorbidities (osteoporosis, hypertension, ischemic heart disease, heart failure, infarct, chronic obstructive pulmonary disease, asthma, depression and diabetes). Clinical data of RA: 28 and 44 joint counts, general health and pain in visual analog scale (VAS) and morning stiffness. Patients completed the health assessment questionnaire disability index (HAQ-DI). Blood tests that included erythrocyte sedimentation rate (ESR), C reactive protein (CRP), hemoglobin levels and serology including rheumatoid factor (RF) and anti cyclic citrullinated peptide antibody (ACPA) level. 28 count disease activity scores (DAS28) index was calculated. Baseline hand and foot radiographic findings and previous and current treatments were also collected. All data was collected from outpatient medical records. Serum RF was measured by a quantitative immunoturbidimetric assay, and the positive results were defined as a value of more than 14YIU/ml. Laboratory test for ACPA level measurement has been introduced into HUGCDN since 2007, so some patients diagnosed RA before that time never took this blood test. The significant changes of radiographic findings included narrowing of joint space or erosions compatible with RA, which were reported in medical records by the rheumatologists or reported by radiologists. In the follow-up visits, 28 and 44 joint counts, VAS GH, VAS pain and morning stiffness, blood tests that included hemoglobin, ESR, CRP and annual hand and foot X-rays were parameters also collected. Incidental comorbidities, and current and new treatments were noted. Disease activity was assessed from DAS28 which was calculated using the original formula with 28 swollen joint count, 28 tender joint count, ESR and VAS GH. Clinical sustained remission was defined as DAS28<2.6 in the last two consecutive visits. Low, moderate, and high disease activity states were defined as DAS28≤3.2, >3.2 but ≤5.1 and >5.1, respectively. The improvement criteria used in this study was the EULAR response criteria. The good responders were patients who had a decline of DAS28 at least 1.2 units from baseline and low disease activity. Moderate responders were patients with a change ≥1.2 and moderate/high disease activity or patients with a change ≤1.2 and >0.6 and low/moderate disease activity, others were nonresponders.

Statistical analyses were performed using SPSS V.16.0 statistical software. Baseline characteristics with continuous data were reported as means±standard deviation (SD), and the categorical data was shown as frequencies and percentages. To test the differences between groups, the Mann–Whitney U test or the independent samples t test was used for continuous variables, and the X2 test for proportions. A stepwise multiple logistic regression model was used to find relevant independent prognostic variables. The variables included in the multivariate model were selected using univariate analysis (P<.05). The results were presented as the odds ratio (OR) with corresponding 95% confidence interval (95% CIs).

75% of our patients were women and the mean age at the analysis was 58 years. Almost 80% were RF (+) and 73% ACPA (+), these are similar characteristics to other large cohorts of RA like QEST-RA2 with more than 5000 patients, 79% female and a mean age of 57 years or CORRONA7 cohort with more than 10,000 patients, 76% women with a mean age of 53 years and 78% RF (+).

Regarding disease activity, there is a big variability depending on the research consulted because of the use of different criteria to define the activity of the disease. In our study, the medium value of DAS28 at basal visit was 4.7 and the VAS pain 63 points. In the database BARFOT that included 698 patients, the average of VAS pain was 46 points and the medium value of DAS28 was 5.2.3 In our research, 84% of the patients had DMARDs treatment and 22% DMARDb therapy; comparable data with other seria of patients like the French cohort ESPOIR with 532 patients, 90% with DMARDs therapy and 23% with biological therapy4; or TARACT cohort with 335 patients, where 89% had DMARDs and 10% DMARDb.

Talking about remission, in this study the prevalence of remission using DAS28 in the first visit was 5%, 11% achieve remission during the first year of treatment and 72% had remission at some point during the follow-up. A considerable number of research about prognostic factors of remission in RA show different results due to the heterogenicity of the patients studied, the different definitions of remission and the treatments used. It has been observed prevalences of remission from less than 10% to up to 40%,2,5 although the studies that find higher percentages of remission are those that include early RA. The remission criteria in RA are in continuous discussion because, depending on the tool we use, patients with the same activity have different scores. In fact, in a study of the QUEST-RA cohort,2 where they evaluated the prevalence of remission using different criteria, they found a great variability (8.6% using ACR and 19.6% using DAS28), concluding that using different criteria results in obtaining different percentages of remission in patients with the same disease activity.

Regarding sustained remission, the prevalence of sustained remission in our study was 46%. We did not find any relation between age and disease duration and clinical remission in RA; which has been published in other studies.6

1st suggestion → In the univariate analysis we did not find a relation between the age and the sustained remission. There were more patients achieving remission in the group of younger than 50 years old (58%) than in those patients older than 50 years old (49%) but this finding was not statistically significant. The duration of the disease was shorter in those patients that were in remission than in those who were not in remission but there were no statistically significant association. There were no statistically significant relation between sustained remission and disease duration categorized in less than 5 years, 5–10 years or more than 10 years.

Male sex was identified as an independent prognostic factor of remission in the multivariate analysis; association that has been communicated before.6 Men had approximately twice the probability of achieving remission than women. Male sex has been related to remission in previous studies like the one with CORRONA cohort with more than 10,000 patients where they used as remission criteria CDAI<2.8.7 Also, in the study with the French cohort BARFOT, male sex has been identified as independent prognostic factor of remission (OR 1.5) using DAS28 in the last two consecutives visits as sustained remission definition.3 Finally, in the GISEA study with 1257 patients treated with DMARDb, it was found male sex as predictor factor of remission.8

Half of the patients in the Rheumatology Division HUGCDN were attended in the first six months after the onset of symptoms which had a relation with the probability of achieving remission either at any point during the follow-up or sustained remission. Early diagnosis in the first year was identified as an independent prognostic factor of maintained remission. So, this is one of the modifiable predictor factors to take into account in the management of a Rheumatology Service in order to achieve the therapeutic goals in RA. The evaluation of a specialist and the early diagnosis since the onset of symptoms will speed up the implementation of adequate treatment and will increase the patient's chances of remission. In this study, the data were variable because there were patients who were on follow up since 2000 up until 2013. Due to the improvement of derivation of patients from primary care and from other Hospital divisions, the delay has been shortened optimizing early diagnosis, although the results could be better. Early diagnosis has been related to remission in other studies such us TARACT,9 a retrospective research in which the definition of sustained remission was remission using DAS28 in two consecutive visits within three months.

In this study, it has been observed that the RA activity in the first visit will predict the prognosis of the disease. The medium DAS28 punctuation of our cohort of patients was 4.7 and more than 75% of the patients presented moderate to high RA activity at the beginning. DAS28 score was identify as an independent predictive factor of sustained remission in the multivariable analysis, and that has been observed in previous studies.6,10 Low baseline DAS28 was strongly predictive of the achievement of remission, which has been reported in other researches.3,11

A lower VAS pain score in the first visit was associated with remission during the first year of treatment, with remission in any point during the follow up and with maintained remission. Vázquez et al.11 also found a relation between low VAS pain at the beginning and sustained remission.

Functional disability will also predict the course of the disease. As seen in previous studies,6,8 the more disability at the beginning, the worse treatment response and the lower the probability of achieving remission. In our research, we found a correlation between a lower HAQ score and remission, correlation also observed in a French cohort of 191 AR patients10 and in the BARFOT cohort.3 Although, neither in our or in previous studies, HAQ score has been identified as an independent predictor factor.

In our study, we found an association between ESR<30 and sustained remission, that has been previously observed in other studies,11 but this association was not seen after the multivariate analysis. Regarding CRP or anemia, although patients in clinical remission had lower levels of CRP and less anemia, this data was not related to remission. Vázquez et al.11 found a relation between low levels of hemoglobin and a lower probability of remission and in a French cohort,10 they found a higher probability of remission in patients with levels of CRP<14.5mg/l. In two adalimumab studies,6,12 they found CRP<20mg/l as an independent factor of remission.

Serum levels of RF and ACPA have been identified as poor prognostic factors in RA, although, being ACPA or RF negative, is not clearly related to a better outcome. In our study we did not find a relation between RF or ACPA and remission. In previous researches, RF negativity has been associated with remission3,9–11 but not as an independent factor. Regarding ACPA, only one study has found association between its negativity and remission.3

Comorbidities in RA patients were not associated with remission in our study. The comorbidities profile on patients with RA is similar to the profile seen in other studies, where hypertension is the most prevalent comorbidity and the percentages of diabetes, depression and infarct are similar. We did not find a relation between Charlson comorbidity index and remission, association that has not been evaluated previously.

Although smoking is a poor prognostic factor in RA,13 in this study there was no relation found between tobacco and remission.

Regarding treatments, an adequate adherence to treatments was related to the probability of achieving remission at some point during the follow up. We also found a relation between the use of DMARDs in monotherapy and sustained remission. That is a fact that could be explained because, normally, patients with better response to treatments, have a less aggressive disease so they require less intensification of therapies. Therefore, a patient with a good response to methotrexate in monotherapy, with a good symptom follow up and low disease activity, probably will not need to add another DMARD. So, this patient profile will achieve remission in monotherapy.

2nd suggestion → The use of biologic DMARD in monotherapy was associated with persistent remission in the last year. But regarding this concept, we have to say that it was a small number of patients (15), and half of them were treated with tocilizumab, biologic agent that, as it is known, reduces in a significant way ESR levels and therefore leads to lower DAS28 scores.14 In general terms, biologic therapy was not associated with remission, probably due to a bias in prescription: patients receiving biologic DMARD are those patients that have not responded to conventional therapy with DMARDs and therefore have a more severe RA.

Another variable associated with remission in our patients was the good EULAR response to the first treatment. The good clinical response after the initial treatment has been identified as a remission predictor factor in some studies.9,12,15 In a study of the TARACT cohort with 366 patients, the good EULAR response during the first year (OR 3.1) was found as an independent predictor factor of sustained remission in the last two visits.9 In the RAPID1 study with Certolizumab pegol, with 703 patients with a disease duration of 6 years and a medium value of DAS28 of 7 (high activity), they observed that patients with a good response to treatment at weeks 4, 6 and 8, had a higher probability of achieving reduced disease activity during the first year. Achieving reduced disease activity or a decrease in the DAS28 score with the first treatment was associated with remission at some point during the follow up and with sustained remission in the last year.16 Not only the good EULAR response to the first treatment, but also achieving remission during the first year of treatment, was associated with sustained remission in our patients. This association has been observed in the ESPOIR study.4 In the study of the cohort CORRONA17 they observed that patients that were in remission during the first year, have a higher probability to continue in remission.