The impact of time length to Boolean remission for tight disease activity control after acquisition in rheumatoid arthritis patients

Clinical importance of time length from initiation under treat-to-target (T2T) strategy to acquisition of clinical remission (TL) in treating patients with rheumatoid arthritis (RA) on disease activity control, daily activities, and quality of life maintenance was investigated. In patients who achieved Boolean remission once or more, relationship between TL and patients’ background data at initiation, and relationship between TL and mean simplified disease activity score (SDAI), Health Assessment Questionnaire Disability Index (HAQ-DI) score, pain score with visual analog scale (PS-VAS), Sharp/van der Heijde Score (SHS) and quality of life score (QOLS) at the first remission and thereafter were evaluated statistically. Patients were divided into two groups whether TL was within 6 months or longer (G ≤ 6 and G > 6). Change of the parameters and Boolean remission rate (BRR) after the first remission between the two groups were compared statistically. In 465 patients, TL correlated significantly with the SDAI score, the HAQ score, PS-VAS, SHS, and the QOLS after the remission. The SDAI score and the BRR after the remission were significantly better in the G ≤ 6 than in the G > 6. TL is an important key to guarantee good and stable clinical course in treating under T2T.


The impact of time length to Boolean remission for tight disease activity control after acquisition in rheumatoid arthritis patients
Ichiro Yoshii 1* , Tatsumi Chijiwa 2 & Naoya Sawada 3 Clinical importance of time length from initiation under treat-to-target (T2T) strategy to acquisition of clinical remission (TL) in treating patients with rheumatoid arthritis (RA) on disease activity control, daily activities, and quality of life maintenance was investigated.In patients who achieved Boolean remission once or more, relationship between TL and patients' background data at initiation, and relationship between TL and mean simplified disease activity score (SDAI), Health Assessment Questionnaire Disability Index (HAQ-DI) score, pain score with visual analog scale (PS-VAS), Sharp/ van der Heijde Score (SHS) and quality of life score (QOLS) at the first remission and thereafter were evaluated statistically.Patients were divided into two groups whether TL was within 6 months or longer (G ≤ 6 and G > 6).Change of the parameters and Boolean remission rate (BRR) after the first remission between the two groups were compared statistically.In 465 patients, TL correlated significantly with the SDAI score, the HAQ score, PS-VAS, SHS, and the QOLS after the remission.The SDAI score and the BRR after the remission were significantly better in the G ≤ 6 than in the G > 6. TL is an important key to guarantee good and stable clinical course in treating under T2T.The mean value of the SDAI score after the first Boolean remission to the last observation demonstrated a significant increase from the first Boolean remission in both groups and in the G > 6 group was significantly higher than that in the G ≤ 6 group.Similarly, the SDAI score, the HAQ-DI score, PS-VAS, and SHS after the first Boolean remission to the last observation in the G > 6 group were also significantly higher than those in the G ≤ 6 group, and the mean value of the QOLS in the G ≤ 6 group was significantly higher than that in the G > 6 group.The Boolean remission rate and SDAI remission rate after the first Boolean remission to the last observation were significantly higher in the G ≤ 6 group than those in the G > 6 group (Table 4).The change of the SDAI score from the first Boolean remission to after the remission was significantly lower in the G ≤ 6 group than that in the G > 6 group, whereas the changes in the HAQ-DI score, PS-VAS, SHS, and QOLS demonstrated no significant differences between the two groups (Table 5).

RA
Except for SDAI at the first Boolean remission, all parameters such as SDAI, HAQ-DI, PS-VAS, and SHS at any moment from one to three years after in the G ≤ 6 group were significantly lower than those in the G > 6 group (p < 0.001), and QOLS in the G ≤ 6 group was significantly higher than that in the G > 6 group (p < 0.01).The change value of the SDAI score in the G ≤ 6 group was significantly lower than that in the G > 6 group at any moment from one to three years after the first Boolean remission (p < 0.001).The SDAI scores at one to three year after compared to that at the first Boolean remission were significantly higher in both groups (p < 0.001), whereas the PS-VAS after the first Boolean remission was significantly higher than that at the remission, and Table 1.Demographic and clinical characteristics of the patient and each group at baseline and time length of each period.G ≤ 6, a patient group who achieved Boolean remission within 6 months from the baseline; G > 6, a patient group who achieved Boolean remission longer than 6 months from the baseline; ACPA anticyclic citrullinated polypeptide antibodies, RF rheumatoid factor, MTX methotrexate, b/tsDMARD biologic/ targeted disease-modifying anti-rheumatic drug, GCS glucocorticoid steroid, TJC tenderness joint count, SJC swollen joint count, PGA patient's global assessment, EGA evaluator's global assessment, CRP C-reactive protein, DAS28 28-joints disease activity score, CDAI clinical disease activity index, SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde Score, QOLS quality of life score.Units: age, year old; disease duration, years; CRP, mg/L; PS-VAS, millimeter.The other parameters' units are numerical values.Except Female, ACPA positive, and RF positive, mean value (standard deviation), median, and range separated with comma are shown.In Female, ACPA positive, and RF positive, real numbers and their ratios in the parentheses are shown.In Female, ACPA positive, and RF positive, p-values were calculated using chi square test, while in the other parameters, Mann-Whitney U-test for the statistical comparison between the G ≤ 6 and the G > 6 was used.Statistically significant columns are shown in bold style.www.nature.com/scientificreports/ the p-values were < 0.05, < 0.05, and < 0.001 at one, two, and three years after, respectively.The QOLS three years after compared to that at the first remission was significantly lower in both groups, and the p-value was < 0.01 and < 0.05 in the G ≤ 6 group and G > 6 group, respectively (Fig. 2).
Relationship between Boolean remission rate and the time length, and other parameters.A scatter plot of the relationship between the overall Boolean remission rate and the length of time in each case is  www.nature.com/scientificreports/shown in Fig. 3.At a glance, the longer the interval from baseline to the first Boolean remission, the lower the Boolean remission rate.The most highly correlated approximation was the exponential equation, with a correlation coefficient of 0.5433, while the linear equation had a correlation coefficient of 0.4090, in which the two parameters were significantly highly correlated (p < 0.001).
The mean SDAI score, PS-VAS, and SHS demonstrated significantly greater values in group 3 (a patient group whose Boolean remission rate was less than 30%) than in group 2 (a patient group whose Boolean remission rate was less than 60% and 30% or more) and group 1 (a patient group whose Boolean remission rate was 60% or more), and demonstrated significantly greater values in the group 2 than in the group 1, whereas the mean time length and HAQ-DI score demonstrated significantly greater in the group 3 than in the group 2 and group 1 and the mean value of the QOLS demonstrated significant less in the group 3 than in the group 2 and group 1 (Fig. 4 and Table 6).Table 3. Correlation between the TL and clinical parameters at the First Boolean remission and between the TL and mean value of the parameters thereafter.TL time length until the first Boolean remission, SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde Score, QOLS quality of life score.Statistical analyses were performed using univariate linear regression analysis as the time span was defined as independent factor and the parameters were defined as dependent factors.Statistically significant columns are shown in bold style.

Discussion
RA is a chronic inflammatory disease that involves the joint structure, and this makes ADL difficult.Therefore, controlling inflammation in the early stage is recommended [2][3][4][5]28 , because early drug intervention prevents joint destruction due to persistent disease activity and the resulting damage to ADL by controlling disease activity with close monitoring of the objective and subjective disease activity 29 . Alhough progress in rheumatology has been remarkable, improvement in RA treatment is still the most important issue 30 .Inflammation control is also present in the overarching principle, which is first mentioned in the EULAR recommendation for managing RA [2][3][4][5] .This leads to the aim of the goal of clinical remission within 3-6 months [3][4][5] .All patients recruited in this study had been under the T2T strategy, which includes monitoring the patient's disease activity, ADL, and comprehensive condition, and considering treatment protocol upon shared decisionmaking with the patient in order to fulfill clinical remission.
Indeed, obtaining clinical remission can prevent joint destruction and impairment in ADL, and many studies have reported many facts that progression is prevented by obtaining clinical remission, particularly in early RA [31][32][33] .The importance of monitoring and recording disease activity is much more thorough in the digital era 34 .Disease activity improvement with optimal discriminatory ability was suggested after 6 months in a clinical trial 35 .A study described the benefit of tight disease control that achieved DAS28-CRP remission significantly earlier compared with conventional treatment for patients with early RA 27 .Tight disease control, namely treatment under the T2T strategy, can achieve earlier clinical remission, and it might lead to a more stable clinical course.However, to the best of our knowledge, no study has evaluated the effect of time to remission on the subsequent clinical course in cases with clinical remission using the T2T treatment strategy.
The high Boolean remission achievement rate in the study was surprising.However, using the real-world data, based on the T2T strategy, out of 685 patients during the relatively long follow-up period of > 3 years, 465 (67.9%) showed Boolean remission once or more, it is realistic, because patients were picked up from various background.That is different from clinical trial study background.Therefore, it can be considered that such a high rate in both of G ≤ 6 and G > 6 groups are realistic given that the Boolean remission achievement rate after the first Boolean remission to the last observation was 62.0% and 43.4%, respectively.However, it is also a fact that some patients could not unfortunately achieve clinical remission, or Boolean remission for some reason such as the reason of patient's personal characteristics, or for some refractory disease status.These patients were excluded from the study.
The reason why 6 months is the critical cut-off index was as follows.Prior to this study, we preliminarily analyzed our data for determining a cut-off index of the time length for the best SDAI course thereafter.The results showed 5.5 months was the best length as a COI as shown in Fig. 1, so we determined 6 months as a critical COI in this study.The AUC of 0.6 for ROC is far from reliable.In fact, COI had a sensitivity of 48.7% and a specificity of 70.1%.However, the p-value was less than 0.1%, and there was a clear correlation between higher SDAI remission rates and shorter time to Boolean remission.As shown in Fig. 4, the time to remission was clearly shorter when the Boolean remission rate was higher.Not only that, the higher Boolean remission rate was clearly advantageous for ADL, pain, joint destruction, and QOL.Furthermore, the higher remission rate was also reported to have a critical impact on the prevention of fragility fractures in RA patients 36 .
As shown in Table 3, not only disease activity but also the shorter time to Boolean remission has a clear positive impact on ADL, joint destruction, and QOL after remission.Furthermore, as shown in Table 4, when divided by 6 months, all clinical indicators after achieving remission show that the group of 6 months or less is better than the group of 6 months or more.These results suggest that separating 6 months from COI leads to stable and comprehensive clinical outcomes and support the rationale for setting the goal of achieving clinical remission at 3-6 months, as demonstrated by T2T.
This study aimed to answer the clinical question of whether a shorter time span to achieve clinical remission would be necessary to make a successful outcome after remission, and how short it is.The primary endpoint was the disease activity after remission between the group of patients who achieved Boolean remission within 6 months after the diagnosis of RA and those who required > 6 months.These results showed that the group who achieved it within 6 months showed significantly better disease activity compared with the group that required > 6 months.The secondary endpoints of the HAQ-DI score, PS-VAS, SHS, and QOLS also showed significantly superior results.However, above all, these parameters were significantly superior in the group that achieved remission within 6 months even at the baseline, and these differences were maintained throughout the treatment.
In the study, the primary endpoint was set as the SDAI score, but not the DAS28 score.The reason for this was a difference in strictness.The remission criteria by means of SDAI is more stringent than those of DAS28-CRP, www.nature.com/scientificreports/so as per previously published report quite different populations were recruited between the Boolean and the DAS28 remission, while similar populations were recruited between the Boolean and the SDAI remission 25,37 .QOL is also an important issue in treating RA because it directly correlates with work status 38 .Therefore, measuring QOL is important for monitoring disease status in RA treatment, as well as measuring physical function 39 .QOLS is calculated based on EQ-5D, which is a conversion formula applied to each disease in each country.QOLS is utilized in the study is used as a reference index when calculating cost-benefit for treating chronic low back pain 40 .To the best of our knowledge, this report is the first to be utilize QOLS to evaluate treatment outcomes in patients with RA, and it showed excellent reproducibility of QOL quantitative evaluation with convenient few questionnaire items and should be considered as a QOL index in patients with RA.
Disease duration and HAQ-DI as patient background factors were also significantly associated with the time length in univariate analysis, similar to Aletaha et al. study 33 , but PS-VAS, SHS, and QOLS were shown to be associated by multivariate analysis.These three parameters showed significant differences even at baseline, at the first Boolean remission, and after the first Boolean remission in comparison between the two groups, and the results strongly suggest that these factors correlated with the time length.However, these factors did not show any difference between the two groups regarding the change after the acquisition of Boolean remission, and the parameters that were affected by time length were SDAI remission rate, Boolean remission rate, and disease activity control after the achievement of Boolean remission.It is presumed that the significant difference in HAQ-DI and Boolean remission after the acquisition of at the first Boolean remission was because many cases with higher treatment resistance were included in the G > 6 group.
MTX dose and b/tsDMARD administration rate were significantly higher in the G > 6 group, despite these parameters demonstrated no significant difference between the two groups at baseline.This may be because the goal of Boolean remission resulted in the need for more intensive treatment compared with the G ≤ 6 group.However, the patient's drug adherence was not considered in the study.There is a wide variability of drug adherence in patients, which strongly influences clinical results 41 .Previous treatment including b/tsDMARD administration at baseline did not influence on the time length.Like these, treatment initiation before disease activity gets high may have no influence on the time length because no disease activity difference at baseline was demonstrated between the two groups.The treatment protocol in the study was commonly designed under the T2T strategy, so every patient recruited in the study accepted shared decision-making and had been treated in targeted clinical remission.It seems to be clear that patient-related outcomes (PRO) such as PS-VAS and QOLS, are rather important for obtaining shorter time length.These parameters and the SHS score throughout treatment from baseline to after the first Boolean remission acquisition demonstrated a significant correlation with the time length.These results suggested that a patient who has good PROs from the baseline is well responsible for treatment when tight disease control is targeted.
Even acquisition of Boolean remission, sustaining the remission is obviously important.The only temporary achievement of Boolean remission is inferior to sustaining Boolean remission.The results of the relationship between the Boolean remission rate and other parameters showed that every parameters in the lower remission rate group demonstrated more inferiority than those in the other groups, while the lower remission rate group also demonstrated significantly longer time length than the two higher remission rate groups as shown in Fig. 4.These results suggested that the time length reflected more sensitivity in clinical outcomes including disease activity control and PRO.
Overall, the validity of aiming for clinical remission within 3-6 months under the T2T strategy has been shown to be effective.The strategy leads to the maintenance of disease activity as well as disease control.This in turn appeared to correlate with the maintenance of ADL and prevention of altered QOL.However, it is conceivable that achieving Boolean remission leads to maintenance.These prospects have not yet been proven in the study and require further investigation.
But the population in the study was a bit trickier.The non-SDAI indices HAQ-DI, PS-VAD, SHS, and QOLS were significantly better in the G ≤ 6 group than in the G > 6 group, and this trend persisted after the remission.The parameters improved until the acquisition of Boolean remission and progressively deteriorated after acquisition (Table 5).These parameters after the first remission were significantly correlated with the time length, as shown in Table 3, and parameters other than the SDAI score already showed the same trend at baseline.One confounding factor was mean disease duration at baseline because G > 6 was significantly longer than G ≤ 6. Duration of disease was significantly correlated with all parameters except QOLS (Table 7).This suggested that patients with a longer history obtained a Boolean remission but had a relatively worse clinical course than those with a shorter history.www.nature.com/scientificreports/There are some limitations to the study.This was a single institutional study in which no ethnic and gender considerations were set and the influence of subjective comorbidities, such as depression, anxiety, fatigue, fibromyalgia, and other diseases were not considered 42 .However, this study addresses the effect of time length from the initiation of treatment to achieve Boolean remission for patients with RA.
In conclusion, we have investigated the impact of time length from baseline to achieve Boolean remission on disease activity control, ADL, and QOL maintenance with a cohort study.Results demonstrated that the shorter period to achieve remission, the tighter the disease activity is controlled, and the better ADL and QOL are maintained.Six months may be the key to guaranteeing a more stable clinical course after the acquisition of Boolean remission.

Methods
Treatment protocol.We have treated RA patients with RA since August 2010 under the T2T strategy in accordance with the EULAR recommendations in the institute where the only specialized clinic for rheumatic diseases in the community.All patients met the American College of Rheumatology/EULAR (ACR/EULAR) classification criteria 43 at the initiation of RA treatment in the institute.The mean time length from the first visit to diagnosis was 1 week, and all RA patients were treated immediately after the diagnosis under the T2T treating strategy in the institute.We set a time of initiation of treatment in the institute as a baseline.
Patients were consulted and treated every 2-3 months intervals since the diagnosis of RA.They were monitored for their tenderness joint count (TJC), swollen joint count (SJC), patient's global assessment (PGA), evaluator's global assessment (EGA), C-reactive protein (CRP), and disease activity indices, such as CDAI, SDAI, DAS28-CRP, and Boolean criteria at every visit to the institute.HAQ-DI score 44 , PS-VAS, and EuroQOL-5th dimension-5L (EQ-5D) were also monitored, and the QOLS calculated from EQ-5D 45 was determined at every visit from the time of diagnosis.As an index of joint damage caused by RA, SHS 46 was measured using X-ray pictures at the time of diagnosis and every other year, and within three months after the achievement of the first Boolean remission.SHS was confirmed by two physicians.The treatment protocol was considered every 3-6 months based on these clinical metrics and decided with a patient in sharing this information.
Patient and parameters selection, and regression analysis at baseline.In patients with RA who had been treated in our institute under the T2T strategy, we enrolled the patients who had achieved Boolean remission once or more, which is defined as all of four parameters such as TJC, SJC, PGA, and CRP ≤ 1 and were consecutively followed up for > 3 years, in the observational study.The time length from the first visit to the first Boolean remission was calculated.The relationship between the time length and each of the background parameters at baseline such as sex, age, disease duration of RA, SHS at the first visit, ACPA, rheumatoid factor (RF), TJC, SJC, PGA, EGA, CRP, SDAI, HAQ-DI, PS-VAS, SHS, and QOLS were evaluated statistically using univariate linear regression analysis, and then multivariate linear regression analysis was performed to evaluate the relationship between the time length and the parameters that demonstrated significant correlation in the univariate model.All data were collected retrospectively from the medical record.
Preliminary regression analysis and ROC study for determining appropriate interval from baseline to first Boolean remission.Before analyzing correlations, an association between time length from baseline to first Boolean remission and the mean SDAI score after the acquisition of Boolean remission was analyzed using linear regression analysis with a univariate model.After confirmation, the ROC study was conducted to determine the cut-off index (COI) for the length of time from baseline to first Boolean remission, demonstrating the most sensitivity and specificity values in SDAI remission rates as the mean value of SDAI after the first Boolean remission.

Regression analysis at the first Boolean remission.
The relationship between the time length and each of the mean values of the SDAI score, the HAQ-DI score, PS-VAS, SHS, and QOLS at the first Boolean remission and thereafter was then evaluated using univariate linear regression analysis considering each parameter as a dependent factor and the time length as an independent factor.Furtherly, odds ratios of the parameters at the baseline, which demonstrated a significant correlation with the time length to achieve the first Boolean remission, were evaluated using binary logistic regression analysis compared with the time length to attain Boolean remission separated in 6 months.

Comparison analysis.
Patients were subsequently divided into the G ≤ 6 and G > 6 groups based on the time length for the achievement of first Boolean remission within two groups: G ≤ 6, a patient group who attained Boolean remission within 6 months from the first visit; G > 6, a patient group who attained Boolean remission more than 6 months from the first visit.The two groups were compared with regard to the SDAI score, the HAQ-DI score, PS-VAS, SHS, and QOLS at the first visit and at the time of first Boolean remission, and the values of these parameters at 1-3 years and the mean values of these parameters after the first Boolean remission were assessed using the Mann-Whitney U test.Repeated measures of ANOVA were used for statistical procedures to evaluate the change of these parameters between the moments.Methotrexate (MTX), biologic/targeted diseasemodifying anti-rheumatic drug (b/tsDMARD), and glucocorticoid steroid (GCS) administration rate at the first visit were also compared between the two groups using Mann-Whitney U-test.Moreover, changes in these parameters from the first Boolean remission to thereafter between the two groups were also compared using the Mann-Whitney U test.Rates of treatment with mean doses of b/tsDMARD, MTX, and GCS administration rate and mean dose of administration at the first Boolean remission and thereafter between the two groups were also compared using the Mann-Whitney U and chi-square tests.The mean Boolean remission rate after the first

Figure 1 .
Figure 1.ROC curve of time length for attaining SDAI remission equivalent (SDAI ≤ 3.3) as a mean value after acquisition first Boolean remission of the patients analyzed in the study.Area under the curve was 0.606.

Table 5 .
Absolute value and change of parameters for each group at each moment.Standard deviations in parentheses are shown.In change at interval rows, *, **, and *** represented statistical significance of p < 0.05, p < 0.01, and p < 0.001 respectively, for which were evaluated time to time change using ANOVA with repeated measures.In p-value rows, *, **, and *** represented statistical significance of p < 0.05, p < 0.01, and p < 0.001 respectively, for which were evaluated differences between G ≤ 6 and G > 6 group for each moment at baseline, Boolean remission, and after Boolean remission, respectively.SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score with visual analog scale, SHS Sharp/van der Heijde score, QOLS quality of life score.

Figure 2 .
Figure 2. Time courses of the parameters from the first Boolean remission to 3 years after the remission comparing in the G ≤ 6 and in the G > 6. Error bars that show standard deviation in each group were shown at each moment.Except for the SDAI score at the first Boolean remission, mean values of all parameters at any moment were significantly lower in the G ≤ 6 group than those in the G > 6 group (p < 0.001), and the QOLS was significantly higher at any moment in the G ≤ 6 group the those in the G > 6 group at any moment (p < 0.01).Change of mean SDAI score was significantly lower in the G ≤ 6 than in the G > 6 (p < 0.001), and change of the other parameters such as HAQ-DI, PS-VAS, SHS, and QOLS demonstrated no significant difference between the two groups.Statical significances of time change at each moment after the first Boolean remission (BL) for each parameter in the each group compared to the values at the BL were symbolized in the figure (*p < 0.5; **p < 0.01; ***p < 0.001).

Figure 3 .
Figure 3. Relationship between total Boolean remission rate (Y-axis) and time length to the first Boolean remission (X-axis).Approximate equations were linear and exponential equations, and correlation coefficient was 0.4090 and 0.5467, respectively.

Figure 4 .
Figure 4. Comparison among Boolean remission rate groups.(1) a patient group whose Boolean remission rate was 60% or more; (2) a patient group whose Boolean remission rate was less than 60% and 30% or more; (3) a patient group whose Boolean remission rate was less than 30%.Error bars represent 95% confidence interval (95% CI).Every parameter in the group 3 demonstrated significantly worse results than those in the other two groups.

Table 2 .
Relationship between TL and parameters at the baseline.TL time length from the baseline to the first Boolean remission, ACPA anti-cyclic citrullinated polypeptide antibodies, RF rheumatoid factor, TJC tenderness joint count, SJC swollen joint count, PGA patient's global assessment, EGA evaluator's global assessment, CRP C-reactive protein, DAS28 28-joints disease activity score, CDAI clinical disease activity index, SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde Score, QOLS quality of life score.Statistical analyses were performed using univariate linear regression analysis as the time span was defined as dependent factor and the parameters were defined as independent factors, and then multivariate linear regression analysis was performed with the parameters that demonstrated significant correlation with the univariate model.Statistically significant columns are shown in bold style.

Table 4 .
Comparison of parameters at the first Boolean remission and mean values of parameter thereafter between the G ≤ 6 and the G > 6. G ≤ 6, a patient group who achieved Boolean remission within 6 months from the baseline; G > 6, a patient group who achieved Boolean remission longer than 6 months from the baseline; MTX methotrexate, GCS glucocorticoid steroid, b/tsDMARD biologic or targeted synthetic disease-modifying anti-rheumatic drug, SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde Score, QOLS quality of life score.Units: time span, months; MTX dosage, mg per week; mean GCS dosage, mg per day; PS-VAS, millimeter.The other parameters' units are numerical values.In time span at the first Boolean remission columns, the time spans from the baseline to the first Boolean remission for each group are expressed.In time span after the first Boolean remission columns, the mean time spans from the first Boolean remission to the last observation for each group are expressed.In time span, SDAI, HAQ-DI, PS-VAS, and QOLS, mean values and standard deviations separated with comma are shown.In mean MTX dosage and mean GCS dosage, mean dosages in patients who was administered with each drug were shown.p-values in time span, mean MTX dosage, mean GCS dosage, SDAI, HAQ-DI, PS-VAS, SHS, and QOLS, are calculated using Mann-Whitney U-test.The other statistics are calculated using chi square test.Statistically significant columns are shown in bold style.Vol:.(1234567890)Scientific Reports | (2023) 13:13908 | https://doi.org/10.1038/s41598-023-39711-4www.nature.com/scientificreports/

Table 6 .
Comparison between each pair of the groups.Statistical procedures: Student T-test.In columns, mean value and 95% confidence interval in parentheses are shown.Statistical procedure: Student T-test.Columns show initial mean, 95% confidence interval in parentheses.All parameters are calculated as means.TL time from baseline to first Boolean remission, SDAI simplified disease activity index score, HAQ-DI Health Assessment Questionnaire Disability Index score, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde score, QOLS quality of life score calculated from EuroQol 5th-dimension score.

Table 7 .
Correlation between parameters and disease duration at baseline.95% CI 95% confidence interval, SDAI simplified disease activity index, HAQ-DI Health Assessment Questionnaire Disability Index, PS-VAS pain score using visual analog scale, SHS Sharp/van der Heijde Score, QOLS quality of life score.