Abstract
Introduction
Accurate assessment of disease activity in systemic lupus erythematosus (SLE) remains a persistent challenge. We aimed to validate the SLE disease activity score (SLE-DAS) in Chinese patients.
Methods
We analyzed the agreement between SLE-DAS and SLEDAI-2 K using our prospective STAR cohort. Construct validity was assessed through correlations with physician global assessment (PGA), as well as through known-group discrimination (PGA-cut off ≥ 1 for active). Factors associated with classification discordance and the association between baseline SLE-DAS and 12-month organ damage accrual were assessed using logistic regression.
Results
Among 507 SLE patients, SLE-DAS demonstrated strong correlations with PGA (r = 0.726, P < 0.001) and SLEDAI-2 K (r = 0.750, P < 0.001), with an overall agreement of 83.0% between SLE-DAS and SLEDAI-2 K. Both instruments demonstrated comparable discriminative power for active disease (SLE-DAS AUC = 0.832; SLEDAI-2 K AUC = 0.860). Among 211 patients with mild disease activity classified by SLEDAI-2 K, 61 were reclassified into higher categories by SLE-DAS. These patients exhibited higher PGA (0.8 ± 1.0 vs. 0.3 ± 0.7; P < 0.001) and glucocorticoid doses (25.7 ± 22.3 vs. 11.3 ± 15.5; P < 0.001) than those with concordant classifications. Lupus nephritis (45.9% vs. 18.0%; P < 0.001) and hematological involvement (32.8% vs. 18.0%; P = 0.020) were more frequent in them. We identified that nephritis (OR = 4.57, 95% CI 2.18–9.57, P < 0.001) and hematological involvement (OR = 2.76, 95% CI 1.27–6.02, P = 0.011) were associated with classification discordance. Baseline SLE-DAS was positively associated with SLE SLICC/ACR Damage Index (SDI) accrual at 12 months of follow-up (OR = 1.38, P = 0.038).
Conclusions
SLE-DAS demonstrates superior accuracy in assessing disease activity compared to SLEDAI-2 K, likely attributable to refinements in assessing nephritis and hematological manifestations. Baseline SLE-DAS also exhibited a positive association with short-term organ damage accrual, though long-term validation is required.
Keywords: Systemic lupus erythematosus, Disease activity, Systemic lupus erythematosus disease activity score, Systemic Lupus Erythematosus Disease Activity Index 2000, Physician global assessment
Key Summary Points
| Why carry out this study? | |
| Accurate assessment of disease activity in systemic lupus erythematosus (SLE) remains challenging. | |
| SLE-DAS is a recently developed tool with potentially improved sensitivity and accuracy. | |
| What was learned from the study? | |
| In a Chinese SLE cohort, SLE-DAS showed strong correlation and substantial agreement with SLEDAI-2K, while reclassifying a proportion of patients into higher disease activity categories. | |
| Nephritic and hematological parameters were key factors in the discrepancy between the two instruments. | |
| SLE-DAS demonstrated better performance than SLEDAI-2K in reflecting subsequent short-term organ damage accrual in an exploratory analysis, which should be validated in future prospective studies with extended follow-up and may deserve considering in clinical practice. |
Introduction
Systemic lupus erythematosus (SLE) is a chronic heterogeneous autoimmunity disease characterized by fluctuating disease activity across multiple organ systems. Accurate and reliable assessment of SLE disease activity is paramount for guiding therapeutic decisions, monitoring treatment response, predicting outcomes, and standardizing clinical research [1]. However, quantifying SLE disease activity remains a significant clinical challenge due to the diverse manifestations of SLE and the inherent limitations of existing measurement tools.
The SLE disease activity index 2000 (SLEDAI-2 K) has been widely adopted as a standard instrument for assessing SLE activity in both clinical practice and trials [2]. Its strengths lie in its relative simplicity and focus on clinically significant new or worsening manifestations within a defined time frame. Despite its utility, SLEDAI-2 K has recognized limitations. These include its binary scoring for most items (present/absent), which may lack sensitivity to subtle changes; potential underweighting of severe manifestations; and possible underestimation of activity in specific organ domains, such as pulmonary hemorrhage. These shortcomings can hinder the precise quantification of activity, particularly in patients with severe or persistent disease.
To address these gaps, the SLE disease activity score (SLE-DAS) was recently developed [3]. SLE-DAS incorporates continuous variables and assigns differential weights to specific manifestations, aiming to provide a more sensitive, granular, and potentially more accurate reflection of overall disease burden. Initial validation studies in diverse populations have demonstrated the promise of SLE-DAS, particularly among patients with low disease activity or severe manifestations such as thrombocytopenia and cardiopulmonary involvement [3–6]. SLE-DAS has also shown high accuracy in identifying disease flares [7, 8], and performs comparably to SLEDAI-2 K in capturing patient-reported outcomes [9]. Furthermore, emerging evidence supports its utility in evaluating belimumab efficacy and predicting hospitalization risk [10–12]. However, limitations have been noted, including inconsistencies with SLEDAI-2 K in highly active disease [13, 14], and controversy regarding the exclusion of active urinary sediments in renal activity assessment [14]. Overall, an optimal instrument assessing SLE activity has not yet been identified. The performance of SLE-DAS, particularly its correlation with physician global assessment (PGA) and its behavior across different disease activity levels and organ involvement patterns, requires rigorous evaluation in diverse ethnic cohorts, including Chinese patients.
Given the potential influence of ethnicity on SLE manifestations and severity, validation in Chinese patients is essential. Therefore, this study aims to validate the SLE-DAS in a well-characterized Chinese systemic lupus erythematosus (SLE) cohort and to identify factors associated with discordance in disease activity categorization between the SLE-DAS and SLEDAI-2 K.
Methods
Study Design and Population
This study analyzed baseline data from 507 SLE patients enrolled in the SLE TARget (STAR) cohort, a single-center, longitudinal observational cohort established by the Department of Rheumatology and Clinical Immunology at Peking University First Hospital in 2007. Demographic characteristics, disease duration, core disease activity measures, laboratory parameters, and treatment regimens were recorded at baseline. Patients were then prospectively followed, with visits scheduled every 1–3 months during periods of active disease and every 3–6 months during stable disease, with additional visits as clinically indicated. Age at disease onset was defined as the age at which patients first developed symptoms or laboratory abnormalities attributable to SLE. Detailed descriptions of this cohort have been published [15, 16]. All patients fulfilled the 1997 American College of Rheumatology (ACR) classification criteria, the 2012 Systemic Lupus International Collaborating Clinics (SLICC) classification criteria, or the 2019 European League Against Rheumatism (EULAR)/ACR classification criteria [17, 18]. The patients with the following conditions were excluded from enrollment: severe infectious diseases; other autoimmunity diseases excluding Sjögren’s syndrome and antiphospholipid syndrome; history of malignant tumor; pregnant and lactating women. This study was conducted based on baseline data. Patients were eligible if complete baseline data required for the assessment of SLE-DAS and SLEDAI-2 K were available. No restrictions were applied regarding disease duration, organ involvement, or treatment status at baseline, in order to reflect a real-world clinical population. Among all eligible patients, baseline disease activity data were missing in 23 cases. These patients were excluded from the analysis, resulting in a final study population of 507 patients. Given the small proportion of excluded cases and the fact that exclusion was based solely on data availability rather than clinical characteristics, the risk of selection bias related to missing baseline data was considered minimal. The ethics committee of Peking University First Hospital approved this study (project 2017[1284]), with informed consent obtained from all participants.
Data Collection
We collected baseline data, including demographics, disease duration of SLE, clinical manifestations, laboratory findings, and treatment at enrollment of the STAR cohort. Disease activity was assessed by a rheumatologist using the SLE Disease Activity Index 2000 (SLEDAI-2 K) and the Physician’s Global Assessment (PGA) (range 0–3.0) on the visiting day. SLE-DAS were calculated based on all the components obtained in routine prospective assessment.
The SLEDAI-2 K categories are defined as mild disease activity or remission (SLEDAI-2 K < 7), moderate activity (7 ≤ SLEDAI-2 K ≤ 12), and severe activity (SLEDAI-2 K > 12) [19]. We also adopted the previously validated disease activity categories based on SLE-DAS, score ≤ 7.64 as mild activity, score > 7.64 as moderate-to-severe disease activity [20]. Patients were considered as a discordant group if their disease activity stratification diverged when applying SLEDAI-2 K and SLE-DAS criteria. A PGA score ≥ 1 served as the rudimentary criterion for defining active SLE, based on the PGA criteria in the lupus low disease activity state [21].
Statistical Analysis
The data were analyzed using STATA16 and GraphPad Prism 15. Continuous variables were expressed as either mean (SD) or median [interquartile range (IQR)], depending on the distribution of the data. Categorical variables were expressed as frequencies and compared using the chi-squared test. The Spearman correlation was used to analyze the correlation between two disease activity scores. The Mann–Whitney U test was used to compare the differences between two non-normally distributed continuous variables. The disease activity states were categorized into two grades, mild activity/remission or moderate-to-severe disease activity, according to the disease activity categories of SLEDAI-2 K and SLE-DAS. The discordant group was defined as patients classified into different disease activity categories by SLEDAI-2 K and SLE-DAS, respectively. Agreement between SLEDAI-2 K and SLE-DAS was assessed through quadratically weighted kappa. Quadratic weighting was chosen because disease activity categories are ordinal and disagreements between non-adjacent categories represent greater clinical discordance than disagreements between adjacent categories. Correlation strength is classified as weak (r < 0.6), strong (0.6 ≤ r < 0.8), and very strong (r ≥ 0.8). Statistically significant was defined as P < 0.05. The area under a receiver operating characteristic (ROC) curve (AUC) was estimated for SLE-DAS and SLEDAI-2 K, to compare the sensitivity and specificity in identifying patients with active disease activity. For known-group discrimination, a PGA score ≥ 1 was used as a reference standard of active disease activity. Univariate and multivariate logistic regression analyses were used to determine factors associated with classification discordance. In the regression models, the dependent variable was discordant disease activity classification between SLEDAI-2 K and SLE-DAS. The independent variables included organ involvement, immunological profiles, and treatments. Multivariate model selection was based on backward stepwise elimination.
Results
Clinical and Demographic Features
This study assessed 507 patients with SLE, including 432 female patients and 75 male patients, with a median disease duration of 6 years and a median age of 36 (28–47) years. The main clinical and demographic characteristics are summarized in Table 1. Among these, the most frequent manifestation was rash (44.6%), followed by nephritis (43.4%) and hematological involvement (33.5%). There were 333 (65.7%) patients with positive anti-dsDNA and 301 (59.4%) had hypocomplementemia. Pharmacological interventions included glucocorticoids (92.3%), hydroxychloroquine (80.1%), and immunosuppressants (58.4%), either as monotherapy or in combination. The mean daily glucocorticoid dosage regimen was 23.8 mg. Figure 1 shows the distribution of disease activity grades of 507 patients classified by SLE-DAS and SLEDAI-2 K. The detailed disease activity classified by different instruments is listed in Table 2. The median scores of SLEDAI-2 K, SLE-DAS, and PGA were 8.0, 11.4, and 1.0, respectively.
Table 1.
Characteristics of 507 patients at enrollment
| Characteristics | n = 507 |
|---|---|
| Demographics | |
| Female patients, n (%) | 432 (85.2) |
| Age, median (IQR), years | 36 (28–47) |
| BMI, median (IQR), kg/m2 | 22.04 (20.00–25.33) |
| Disease duration, median (IQR), years | 6 (1–12) |
| SDI value 0, n (%) | 402 (79.3) |
| Organ involvement | |
| Rash, n (%) | 226 (44.6) |
| Nephritis, n (%) | 220 (43.4) |
| Hematological, n (%) | 170 (33.5) |
| Neuropsychiatric, n (%) | 49 (9.7) |
| Arthritis, n (%) | 44 (8.7) |
| Vasculitis, n (%) | 44 (8.7) |
| Cardiac/pulmonary, n (%) | 27 (5.3) |
| Myositis, n (%) | 9 (1.8) |
| Laboratory characteristics* | |
| White blood cells (× 109/l) | 5.60 (4.03–7.90) |
| Hemoglobin (g/l) | 117 (98–130) |
| Platelet (109/l) | 202 (148–254) |
| C-reactive protein (mg/l) | 1.90 (0.37–6.33) |
| UPCR (g/gcr) | 0.35 (0.10–1.60) |
| Serum creatinine (µmol/l) | 71.00 (61.20–87.40) |
| eGFR (ml/min/1.73 m2) | 71.00 (61.20–87.40) |
| Positive anti-dsDNA | 333 (65.7) |
| Complement C3 (g/l) | 0.60 (0.44–0.77) |
| Complement C4 (g/l) | 0.12 (0.07–0.18) |
| Immunoglobulin G (g/l) | 12.1 (8.48–17.00) |
| Immunoglobulin A (g/l) | 2.52 (1.74–3.45) |
| Immunoglobulin M (g/l) | 0.87 (0.51–1.33) |
| Treatments | |
| Glucocorticoids, n (%) | 468 (92.3) |
| Average dose of glucocorticoids, mg/day, mean (SD) | 23.8 (23.7) |
| Hydroxychloroquine, n (%) | 406 (80.1) |
| Immunosuppressants, n (%) | 296 (58.4) |
| Biologics, n (%) | 74 (14.6) |
BMI body mass index, IQR interquartile range, UPCR urine protein-to-creatinine ratio, eGFR estimated glomerular filtration rate, Anti-dsDNA anti-double-stranded DNA antibody, SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index
*Data presented are median (IQR) for non-normally distributed continuous variables and number (percentage) for categorical variables
Fig. 1.
UpSet plot showing the distribution of 507 SLE patients classified by disease activity using SLEDAI-2K and SLE-DAS. The largest intersection (n = 271) comprised patients with both SLEDAI-2K ≥ 7 and SLE-DAS > 7.64.
Table 2.
Disease activity of 507 SLE patients
| Assessment tool and activity stratification | Number (%) |
|---|---|
| SLEDAI-2 K score | |
| SLEDAI-2 K < 7 | 211 (41.6) |
| 7 ≤ SLEDAI-2 K ≤ 12 | 165 (32.5) |
| SLEDAI-2 K > 12 | 131 (25.9) |
| SLE-DAS score | |
| SLE-DAS ≤ 7.64 | 175 (34.5) |
| 7.64 < SLE-DAS ≤ 9.9 | 47 (9.3) |
| SLE-DAS > 9.9 | 285 (56.2) |
| PGA | |
| PGA < 0.3 | 79 (15.6) |
| 0.3 ≤ PGA ≤ 1 | 226 (44.6) |
| PGA > 1 | 202 (39.8) |
SLE systemic lupus erythematosus, SLEDAI-2 K Systemic Lupus Erythematosus Disease Activity Index 2000, SLE-DAS SLE disease activity score, PGA physician global assessment
Comparability of SLE-DAS and SLEDAI-2 K Across Diverse Disease Activity Levels
We assessed the correlation and agreement between SLE-DAS and SLEDAI-2 K. A strong positive correlation was observed in the overall cohort of 507 patients (r = 0.750, P < 0.001) (Fig. 2). Based on SLEDAI-2 K criteria, patients were classified as remission/mild activity (n = 211, 41.6%) or moderate/severe activity (n = 296, 58.4%). Using SLE-DAS criteria, remission/mild activity was identified in 175 patients (34.5%) and moderate/severe activity in 332 patients (65.5%). Disease activity classification was consistent between the two instruments for 421 patients (83.0%). Cohen’s kappa showed moderate agreement (Cohen’s κ = 0.642, 95% CI 0.572–0.712; P < 0.001), suggesting comparable performance between the two instruments. However, the agreement between SLE-DAS and SLEDAI-2 K is reduced when the degree of disagreement across ordinal disease activity levels is taken into account (the quadratically weighted Cohen’s κ = 0.539, 95% CI 0.476–0.602; P < 0.001). This finding indicates that the two indices are generally concordant in overall classification, but may differ in their assessment of disease activity severity, particularly when larger category discrepancies are considered. We additionally provided a table detailing the distribution of concordant and discordant disease activity categories between the two scoring systems (Table 3). The numbers in bold represent the patients who were classified as being in the same category of disease activity by both SLEDAI-2 K and SLE-DAS.
Fig. 2.
Scatter plot depicting the correlation between SLE-DAS and SLEDAI-2K scores in 507 SLE patients. Spearman correlation analysis demonstrated a strong positive correlation between the two indices (R = 0.750, P 0.001)
Table 3.
Agreement between SLEDAI-2 K and SLE-DAS disease activity categories (n = 507)*
| SLE-DAS | |||||
|---|---|---|---|---|---|
| ≤ 7.64 | 7.64–9.9 | > 9.9 | Total | ||
| SLEDAI-2 K | < 7 | 150 (29.6%) | 19 (3.7%) | 42 (8.3%) | 211 (41.6%) |
| 7–12 | 13 (2.6%) | 21 (4.1%) | 131 (25.8%) | 165 (32.5%) | |
| > 12 | 12 (2.4%) | 7 (1.4%) | 112 (22.1%) | 131 (25.9%) | |
| Total | 175 (34.5%) | 47 (9.3%) | 285 (56.2%) | 507 | |
SLEDAI-2 K Systemic Lupus Erythematosus Disease Activity Index 2000, SLE-DAS SLE disease activity score
*The numbers in bold represent the patients who were classified as being in the same category of disease activity by both SLEDAI-2 K and SLE-DAS
Indeed, when stratified by SLEDAI-2 K-defined disease activity, the correlation between SLE-DAS and SLEDAI-2 K varied significantly. Strong correlation persisted in the remission/mild activity group (r = 0.765, P < 0.001), but diminished substantially in patients with moderate activity (r = 0.226, P = 0.003) and severe activity (r = 0.419, P < 0.001). Collectively, these findings indicate a limited association between SLE-DAS and SLEDAI-2 K across the full spectrum of disease activity.
Construct Validity
PGA scores and plasma complement (C3 and C4) levels are established indicators of SLE disease activity. To evaluate construct validity, we examined their correlations with SLE-DAS. SLE-DAS demonstrated a strong positive correlation with PGA (r = 0.726, P < 0.001), though slightly weaker than that between SLEDAI-2 K and PGA (r = 0.769, P < 0.001). However, these strong correlations were not consistently maintained across different disease activity subgroups (Fig. 3). In 211 patients with SLEDAI-2 K defined remission/mild disease activity, SLE-DAS showed a moderate correlation with PGA (r = 0.541, P < 0.001). This correlation decreased substantially in the moderate activity group (n = 164, r = 0.318, P < 0.001), but was stronger again in the severe activity group (n = 132, r = 0.643, P < 0.001). A similar fluctuation was observed for SLEDAI-2 K correlations with PGA (mild: r = 0.565; moderate: r = 0.224; severe: r = 0.566; all P < 0.001). These findings indicate that neither scoring system maintains robust correlation with PGA across the full spectrum of disease activity. Notably, SLE-DAS showed a slightly stronger correlation with PGA than SLEDAI-2 K, specifically in the severe activity group.
Fig. 3.
In 211 patients with SLEDAI-2K defined remission/mild disease activity, SLE-DAS showed a moderate correlation with PGA (r = 0.541, P 0.001). This correlation decreased substantially in the moderate activity group (n = 164, r = 0.318, P 0.001), but was stronger again in the severe activity group (n = 132, r = 0.643, P 0.001). A similar fluctuation was observed for SLEDAI-2K correlations with PGA (mild: r = 0.565; moderate: r = 0.224; severe: r = 0.566; all P 0.001)
Both SLEDAI-2 K and SLE-DAS showed weak negative correlation with C3 (SLEDAI-2 K: r = − 0.467, P < 0.001; SLE-DAS: r = − 0.405, P < 0.001) and C4 (SLEDAI-2 K: r = − 0.338, P < 0.001; SLE-DAS: r = − 0.297, P < 0.001). SLEDAI-2 K exhibited numerically stronger correlations with complement C3 and C4 levels than SLE-DAS.
For known-group discrimination, a PGA score ≥ 1 was used as reference standard of active disease activity to evaluate the instruments. ROC analysis revealed that SLE-DAS and SLEDAI-2 K had equally effective diagnostic performance for discriminating active disease (Fig. 4). The AUCs were 0.832 (95% CI 0.796–0.868, P < 0.001) for SLE-DAS and 0.860 (95% CI 0.829–0.892, P < 0.001) for SLEDAI-2 K.
Fig. 4.
ROC curve analysis demonstrated that both SLE-DAS and SLEDAI-2K exhibited comparable discriminatory capacity in identifying active disease status. Specifically, the AUC for SLE-DAS was 0.832 (95% CI 0.796–0.868, P 0.001), while SLEDAI-2K achieved an AUC of 0.860 (95% CI 0.829–0.892, P 0.001)
Organ-Specific Performance
Since SLE-DAS converts several items (notably arthritis, nephritis, and hematological involvements) into continuous variables to enhance sensitivity, we further evaluated the agreement between SLE-DAS and SLEDAI-2 K across specific organ involvement patterns (Table 4). Notably, agreement between the two scoring systems decreased markedly in patients without arthritis versus those with arthritis. Moreover, we observed a slightly lower concordance in nephritis patients (Cohen’s κ = 0.514, 95% CI 0.378–0.650, P < 0.001) versus non-nephritis patients (Cohen’s κ = 0.645, 95% CI 0.556–0.734, P < 0.001). Patients with hematological involvement (Cohen’s κ = 0.566, 95% CI 0.418–0.714, P < 0.001) also showed lower concordance between SLE-DAS and SLEDAI-2 K compared to those without hematological involvement (Cohen’s κ = 0.648, 95% CI 0.564–0.733, P < 0.001). Similar comparisons based on vasculitis, neuropsychiatric, and cardiopulmonary involvement were limited by small sample sizes. In these subgroups, complete classification discordance was observed, thereby precluding valid kappa coefficient calculation.
Table 4.
Agreement between SLE-DAS and SLEDAI-2 K in patients with different organ involvements
| Nephritis | Arthritis | Hematological involvement | Rash | |||||
|---|---|---|---|---|---|---|---|---|
| Yes (n = 220) | No (n = 287) | Yes (n = 44) | No (n = 463) | Yes (n = 170) | No (n = 337) | Yes (n = 226) | No (n = 281) | |
| Kappa | 0.514 | 0.645 | 0.808 | 0.625 | 0.566 | 0.648 | 0.689 | 0.587 |
| P value | < 0.001 | < 0.001 | < 0.001 | < 0.001 | < 0.001 | < 0.001 | < 0.001 | < 0.001 |
| Agreement | 84.1% | 82.2% | 95.5% | 81.9% | 84.1% | 82.5% | 88.3% | 79.0% |
Characteristics Associated with Classification Discordance
Eighty-six patients exhibited discordant disease activity classifications when assessed using SLEDAI-2 K and SLE-DAS. Among these cases, 61 (70.9%) patients originally categorized as mild activity/remission by SLEDAI-2 K, were reclassified as having moderate-to-severe disease activity by SLE-DAS, indicating a potential underestimation of disease activity by SLEDAI-2 K. In clinical practice, accurate determination of low disease activity is essential, since patients who are mistakenly categorized as having low disease activity but actually exhibit high disease activity require timely treatment adjustment. Thus, we evaluate discrepancies between SLE-DAS and SLEDAI-2 K in the 211 patients who had been classified as remission/mild disease activity according to SLEDAI-2 K. In this subgroup, we compared the clinical characteristics between patients with concordant (n = 150) and discordant (n = 61) classifications (Table 5). Compared to concordant patients, the discordant group more frequently exhibited lupus nephritis (LN) (45.9% vs. 18.0%, P < 0.001), hematological involvement (32.8% vs. 18.0%, P = 0.020), neuropsychiatric (13.1% vs. 0%, P < 0.001), and cardiac/pulmonary involvement (4.9% vs. 0%, P = 0.006). No significant differences were observed between the groups in terms of arthritis, rash, or immunological profiles. Notably, the median SLEDAI-2 K score (6 [IQR 4–8] vs. 4 [IQR 2–4]; P < 0.001), SLE-DAS (12.18 [IQR 9.57–19.12] vs. 2.08 [IQR 1.12–4.46]; P < 0.001), and PGA (0.8 ± 1.0 vs. 0.3 ± 0.7; P < 0.001) was significantly higher in the discordant group. Additionally, the discordant group was more likely to receive immunosuppressants (72.1% vs. 51.3%, P < 0.001), suggesting these patients had higher disease activity. Notably, among 61 patients with discordant assessments between the two scoring systems, all eight exhibiting neuropsychiatric involvement demonstrated peripheral neuropathy, and all three with cardiopulmonary involvement primarily presented with Libman–Sacks endocarditis and pulmonary hypertension. These clinical features represent newly incorporated assessment domains in SLE-DAS but were excluded from the SLEDAI-2 K framework.
Table 5.
Clinical characteristics of 211 SLE patients with mild disease activity classified by SLEDAI-2 K: Discordant versus concordant classification when assessed using SLE-DAS
| Characteristics | Discordant (n = 61) | Concordant (n = 150) | P value |
|---|---|---|---|
| Demographics | |||
| Female patients, n (%) | 50 (82.0%) | 140 (93.3%) | 0.013* |
| Age, median (IQR), years | 39 (32–49) | 38 (32–50) | 0.574 |
| Disease duration, median (IQR), years | 8.0 (2–13) | 7.5 (3–16) | 0.610 |
| Organ involvement | |||
| Rash, n (%) | 13 (21.3%) | 45 (30.0%) | 0.201 |
| Nephritis, n (%) | 28 (45.9%) | 27 (18.0%) | < 0.001*** |
| Haematological, n (%) | 20 (32.8%) | 27 (18.0%) | 0.020* |
| Arthritis, n (%) | 2 (3.3%) | 5 (3.3%) | 0.984 |
| Neuropsychiatric, n (%) | 8 (13.1%) | 0 (0%) | < 0.001*** |
| Cardiac/pulmonary, n (%) | 3 (4.9%) | 0 (0%) | 0.006** |
| Immunological profiles# | |||
| Positive anti-dsDNA | 29 (47.5%) | 76 (50.7%) | 0.681 |
| Complement C3 (g/l) | 0.70 (0.50–0.82) | 0.71 (0.60–0.86) | 0.145 |
| Complement C4 (g/l) | 0.14 (0.09–0.19) | 0.16 (0.11–0.20) | 0.338 |
| Immunoglobulin G (g/l) | 12.20 (8.16–18.20) | 12.95 (9.65–17.10) | 0.431 |
| Immunoglobulin A (g/l) | 2.35 (1.70–3.10) | 2.70 (1.88–3.82) | 0.068 |
| Immunoglobulin M (g/l) | 0.84 (0.47–1.29) | 0.89 (0.51–1.41) | 0.822 |
| C-reactive protein | 2.43(0.52–5.66) | 0.98 (0.15–3.66) | 0.007** |
| Disease activity | |||
| SLEDAI-2 K, median (IQR) | 6 (6–8) | 4 (2–4) | < 0.001*** |
| SLE-DAS, median (IQR) | 12.18 (9.57–19.12) | 2.08 (1.12–4.46) | < 0.001*** |
| PGA, median (IQR) | 0.8 (0.5–1.0) | 0.2 (0 – 0.5) | < 0.001*** |
| SDI value, mean (SD) | 0.5 (1.0) | 0.3 (0.7) | 0.489 |
| Treatments | |||
| Glucocorticoids, n (%) | 52 (85.3%) | 106 (70.7%) | 0.157 |
|
Average dose of glucocorticoid (mg/day), mean (SD) |
25.7 (22.3) | 11.3 (15.5) | < 0.001*** |
| Hydroxychloroquine, n (%) | 51 (83.6%) | 133 (88.7%) | 0.320 |
| Immunosuppressants, n (%) | 44 (72.1%) | 77 (51.3%) | 0.006** |
| Biologics, n (%) | 5 (8.2%) | 14 (9.3%) | 0.794 |
SLE systemic lupus erythematosus, IQR inter-quartile range, SLEDAI-2 K Systemic Lupus Erythematosus Disease Activity Index 2000, SLE-DAS SLE disease activity score, PGA physician global assessment
*P < 0.05, **P < 0.01, ***P < 0.001
#Data presented are median (IQR) for non-normally distributed continuous variables
To evaluate the association between specific independent variables with discordant disease activity classification by SLE-DAS among patients classified as having mild activity/remission based on SLEDAI-2 K, we conducted a multivariate logistic regression analysis (Table 6). The analysis found that the presence of nephritis (OR = 4.57, 95% CI 2.18—9.57, P < 0.001) and hematological involvement (OR = 2.76, 95% CI 1.27–6.02, P = 0.011) were independent factors correlated with classification discordance in patients classified as mild disease activity/remission by SLEDAI-2 K. Moreover, the use of immunosuppressants (OR = 2.29, 95% CI 1.07–4.92, P = 0.033) also predicts discordant classification.
Table 6.
Univariate and multivariate logistic regression analysis for factors associated with discordant disease activity classifications by SLE-DAS among patients classified as low disease activity based on SLEDAI-2 K
| Variable | Univariate | Multivariate | ||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| Age | 1.01 (0.99–1.03) | 0.383 | ||
| Disease duration | 0.99 (0.95–1.03) | 0.634 | ||
| Gender | 3.08 (1.23–7.69) | 0.016* | 2.61 (0.91–7.55) | 0.076 |
| Rash | 0.63 (0.31–1.28) | 0.202 | ||
| Nephritis | 3.87 (2.01–7.43) | < 0.001** | 4.57 (2.18–9.57) | < 0.001*** |
| Haematological | 2.22 (1.13–4.38) | 0.021* | 2.76 (1.27–6.02) | 0.011 |
| Arthritis | 0.98 (0.05–0.91) | 0.984 | ||
| Alopecia | 0.52 (0.19–1.44) | 0.208 | ||
| Decreased Complement C3 | 1.39 (0.73–2.65) | 0.320 | 0.94 (0.36–2.49) | 0.908 |
| Decreased Complement C4 | 1.67 (0.90–3.11) | 0.104 | 2.02 (0.81–5.09) | 0.134 |
| Positive ANA | 1.23 (0.32–4.72) | 0.759 | ||
| Positive Anti-dsDNA | 0.88 (0.49–1.60) | 0.681 | ||
| C-reactive protein | 1.00 (0.99–1.02) | 0.645 | ||
| SDI | 1.28 (0.92–1.78) | 0.138 | ||
| Use of hydroxychloroquine | 0.65 (0.28–1.52) | 0.321 | ||
| Use of glucocorticoids | 2.40 (1.09–5.28) | 0.030* | 1.72 (0.69–4.31) | 0.246 |
| Use of biologics | 0.87 (0.30–2.52) | 0.794 | ||
| Use of immunosuppressants | 2.45 (1.29–4.68) | 0.006** | 2.29 (1.07–4.92) | 0.033 |
OR odds ratio, 95% CI 95% confidence interval, Anti-dsDNA anti-double-stranded DNA antibody, SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index
*P < 0.05, **P < 0.01, ***P < 0.001
Associations Between Baseline Disease Activity Scores and 12-Month Organ Damage Accrual
To validate the SLE-DAS and SLEDAI-2 K scores against a long-term outcome, we conducted an exploratory analysis using Firth-corrected logistic regression models to examine the associations between these two disease activity scores and organ damage accrual at 12 months of follow-up (Tables 7, 8). Organ damage accrual was defined as an increase of ≥ 1 in SLE SLICC/ACR Damage Index (SDI). Firth-corrected logistic regression was used due to the limited number of damage accrual events observed at 12 months, as this penalized approach reduces small-sample bias associated with sparse data. Moreover, we restricted the number of covariates included in the models to minimize the risk of overfitting. Given the strong associations between baseline SDI, high-dose corticosteroid use (> 1 mg/kg/day), and disease duration with subsequent damage accrual, we included these key factors as adjustments in the models. Additionally, since the proportion of patients with nephritis was relatively high in our sample, we also included nephritis as a potential confounder.
Table 7.
Associations of baseline SLEDAI-2 K with increased organ damage accrual in 12 months
| Variable | Model A† | Model B‡ | Model C§ | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| SLEDAI-2 K | 0.82 (0.57–1.17) | 0.278 | 0.81 (0.57 -1.17) | 0.261 | 0.79 (0.53–1.18) | 0.252 |
| Baseline SDI | 1.40 (0.66–2.96) | 0.375 | 1.36 (0.65–2.89) | 0.416 | ||
| Disease duration | 1.02 (0.97–1.06) | 0.480 | ||||
| High-dose glucocorticoids | 1.21 (0.51–2.90) | 0.665 | ||||
| Lupus nephritis | 1.24 (0.62–2.49) | 0.541 | ||||
OR odds ratio, 95% CI 95% confidence interval, SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index, SLEDAI-2 K Systemic Lupus Erythematosus Disease Activity Index 2000
†Model A: unadjusted
‡Model B: adjusted for baseline SDI
§Model C: additionally adjusted for disease duration, high-dose glucocorticoid use, and lupus nephritis
Table 8.
Associations of baseline SLE-DAS with increased organ damage accrual in 12 months
| Variable | Model A† | Model B‡ | Model C§ | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| SLE-DAS | 1.34 (1.00–1.79) | 0.050 | 1.32 (0.99–1.77) | 0.058 | 1.38 (1.02–1.87) | 0.038 |
| Baseline SDI | 1.30 (0.61–2.75) | 0.498 | 1.22 (0.57–2.61) | 0.601 | ||
| Disease duration | 1.03 (0.98–1.07) | 0.239 | ||||
| High-dose glucocorticoids | 0.89 (0.37–2.11) | 0.789 | ||||
| Lupus nephritis | 0.94 (0.48–1.86) | 0.861 | ||||
OR odds ratio, 95% CI 95% confidence interval, SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index, SLE-DAS SLE disease activity score
†Model A: unadjusted
‡Model B: adjusted for baseline SDI
§Model C: additionally adjusted for disease duration, high-dose glucocorticoid use, and lupus nephritis
A total of 459 patients had available SDI data at the 12-month follow-up, after excluding 33 patients with follow-up shorter than 12 months and 15 patients were lost to follow-up. Although the mechanism underlying the missing data in 15 patients could not be definitively determined, the limited and non‑systematic nature supports the assumption of missing completely support the assumption of missing completely at random. These patients were therefore excluded from the analysis, with minimal expected impact on sample size or statistical power. Among 459 patients, 36 showed an increase in cumulative organ damage (SDI increase ≥ 1). SLEDAI-2 K was not associated with SDI accrual in the unadjusted model (Model A; OR = 0.82, 95% CI 0.573–1.173, P = 0.278), nor after adjustment for baseline SDI (Model B; OR = 0.81, 95% CI 0.567–1.167, P = 0.261) or further adjustment for disease duration, high-dose glucocorticoid use, and lupus nephritis (Model C; OR = 0.79, 95% CI 0.53–1.18, P = 0.252). In contrast, when the same modeling strategy was applied to SLE-DAS, a directionally stronger association with SDI accrual was observed. In the unadjusted model, baseline SLE-DAS exhibits a trend toward a positive association with SDI accrual at 12 months, with a borderline statistical significance (Model A; OR = 1.34, 95% CI 1.00–1.79, P = 0.050). After adjustment for baseline high-dose corticosteroid use, disease duration, and lupus nephritis, baseline SLE-DAS was significantly associated with SDI accrual at 12 months (Model C; OR = 1.38, 95% CI 1.02–1.87, P = 0.038). These findings suggest differential associations of the two disease activity indices with subsequent organ damage accrual, with SLE-DAS showing a more consistent association with 12-month organ damage accrual.
Discussion
Accurate assessment of disease activity is essential in SLE management, particularly for the timely identification of active disease and adjustment of treatment strategies. This study validated the SLE-DAS, a novel instrument for quantifying SLE disease activity, in a cohort of 507 Chinese patients. The results demonstrated that SLE-DAS showed strong correlation and high agreement with the widely used SLEDAI-2 K, supporting its validity in evaluating disease activity. Notably, most discordant categorizations occurred among the patients classified by SLEDAI-2 K as having remission or mild disease activity. SLE-DAS assigned higher disease activity categories to these patients compared to SLEDAI-2 K. These patients exhibited significantly higher PGA scores and glucocorticoid doses compared to those with concordant mild/remission classifications, supporting the superior accuracy of SLE-DAS in reflecting actual disease status. Multivariate logistic regression identified that nephritis and hematological involvement as factors associated with classification discordance between SLE-DAS and SLEDAI-2 K. Moreover, baseline SLE-DAS demonstrated a directionally positive association with SDI accrual at 12-month follow-up.
For over decades, SLEDAI-2 K has been the predominant tool for assessing SLE disease activity, remaining widely employed in clinical trials and high-quality registries, despite recognized limitations in detecting subtle changes and capturing certain severe manifestations. SLE-DAS represents an innovative approach to quantifying disease activity, with preliminary validation in small populations from Mexico, India, and China [4, 13, 14, 22]. Consistent with these studies, we observed a strong correlation between SLE-DAS and SLEDAI-2 K in the entire 507 patients. Additionally, previous reports [13, 14] and our own findings indicated that the correlation between SLE-DAS and SLEDAI-2 K weakened significantly in patients with severe disease activity. However, Rodríguez-González et al. found the correlation coefficient between these two tools remained almost unchanged when analyzing severe cases [22]. They attributed this discordance primarily to discrepancies in the clinical characteristics across different studies. As for the agreement of SLE-DAS and SLEDAI-2 K, earlier studies had inconsistent kappa values, ranging from a near-perfect agreement (κ = 0.933) [5] to relatively poor agreement (κ = 0.489) [4]. Our analysis demonstrated moderate agreement (κ = 0.642) falling within this range. In all, while SLE-DAS and SLEDAI-2 K demonstrate high correlation and agreement in mild disease, this concordance diminishes in severe disease. Both scoring systems exhibit better consistency in populations with low disease activity, plausibly explained by the initial development and validation of SLE-DAS predominantly based on such a population [3].
Since SLEDAI-2 K is not a perfect reference standard, its correlation with SLE-DAS alone is insufficient to fully validate the new instrument. Comprehensive SLE disease activity assessment in practice integrates clinical symptoms, laboratory findings, PGA, and patient-reported outcomes. Thus, we further evaluated correlations between SLE-DAS and established activity indicators, such as PGA and plasma complement levels. As observed, SLE-DAS showed a strong positive correlation with PGA, which was slightly lower than the correlation observed for SLEDAI-2 K. Moreover, both instruments demonstrated significant, albeit low, negative correlation with C3 and C4 levels, consistent with prior research [23], thereby supporting the comparable performance of SLE-DAS and SLEDAI-2 K. Importantly, the correlation of both scores with PGA varied considerably across disease activity strata. SLE-DAS maintained a slightly stronger correlation with PGA than SLEDAI-2 K in severe disease, suggesting potentially improved robustness in the high-activity population. In summary, while SLE-DAS showed limited concordance with SLEDAI-2 K in high-activity patients, its stronger correlation with PGA indicates superior performance relative to SLEDAI-2 K.
As an innovative attempt for quantifying disease activity, SLE-DAS enables improved assessment by continuous measurements of key clinical parameters (rash, arthritis, nephritis, and hematological involvement). Our organ-specific analysis revealed that the agreement between SLE-DAS and SLEDAI-2 K differed based on the presence or absence of specific clinical manifestations. For instance, the two scoring systems exhibited a lower degree of concordance in patients with arthritis than in those without arthritis. These suggested that responsiveness to specific clinical manifestations, such as arthritis, across different scoring systems may be associated with discordant classification.
Notably, in our study of 507 patients with SLE, the majority of cases (83.0%) were concordantly classified by SLE-DAS and SLEDAI-2 K, suggesting an excellent performance for SLE-DAS in quantifying SLE activity. In 61 of the 86 discordant cases, SLE-DAS classified patients as having moderate-to-severe disease activity, while SLEDAI-2 K categorized these patients as having low disease activity. These results indicate that the discordance between the two instruments was more pronounced in patients with low disease activity, where SLEDAI-2 K appeared to underestimate the disease activity. We subsequently examined the 211 patients classified by SLEDAI-2 K as remission/low disease activity, who were either concordant or discordant when assessed using SLE-DAS. The discordant group exhibited higher scores in SLE-DAS, SLEDAI-2 K, and PGA, along with increased C-reactive protein and more frequent use of immunosuppressants compared to the concordant group, suggesting more severe disease activity. The higher frequency of nephritis and hematological manifestations in the discordant group indicated that scoring discrepancies may be attributable to refinements in these components. Multivariate logistic regression analysis further demonstrated that the nephritis and hematological involvement in patients with initially low SLEDAI-2 K scores were predictive of discordant classification upon assessment with the SLE-DAS. Similarly, a recent retrospective study of 324 patients identified cutaneous and hematologic manifestations as predictors of classification discordance between SLE-DAS and SLEDAI-2 K [6]. All of these results align with the design of SLE-DAS, which incorporates refined assessments for renal, cutaneous, and hematologic manifestations, although some of these modifications remain debated. For instance, Mathew A et al. [14] found that SLE-DAS may fail to reflect active LN without significant proteinuria but with active urinary sediments, as only proteinuria contributes to the nephritis score in SLE-DAS. Indeed, while isolated urinary sediments may indicate renal activity, they often correlate poorly with severe histopathological changes and long-term renal outcomes compared to proteinuria. Exclusion of these variables may not necessarily impair the performance of SLE-DAS. Conversely, quantitative assessment of proteinuria in SLE-DAS, the more clinically significant parameter, may improve the accuracy of renal activity in SLE.
An important issue raised by our findings is the reason underlying the discordance between SLE-DAS and SLEDAI-2 K classifications. This discordance reflects fundamental differences in index design between the two instruments. SLEDAI-2 K is a long-established index that relies on dichotomous definitions of most clinical and laboratory items and assigns fixed weights to disease manifestations. While this design confers robustness and simplicity, it may limit sensitivity to gradations in disease severity, particularly in patients with partial improvement or persistent low-to-moderate activity. In contrast, SLE-DAS incorporates continuous variables and refined weighting schemes, which may allow a more nuanced capture of disease activity, especially in organ systems such as nephritis and hematological involvement. From this perspective, the observed discordance may reflect both a relative strength of SLE-DAS in detecting clinically relevant activity and an inherent limitation of SLEDAI-2 K in capturing subtle but meaningful changes in disease burden. Importantly, patients reclassified into higher activity categories by SLE-DAS exhibited higher physician global assessment scores, greater glucocorticoid exposure, and more frequent major organ involvement, supporting the clinical plausibility of SLE-DAS-based classification. The discordance may reflect both a relative strength of SLE-DAS and structural limitations of SLEDAI-2 K in certain clinical contexts. Further prospective studies incorporating longitudinal outcomes are needed to clarify whether discordance reflects improved sensitivity, reduced specificity, or a combination of both.
Importantly, SLE-DAS showed a consistent association with short-term organ damage accrual after multivariable adjustment, whereas no such association was observed for SLEDAI-2 K. This additional analysis may provide complementary evidence supporting the relationship between SLE-DAS and subsequent organ damage. However, the relatively short follow-up period and the limited number of damage events preclude firm conclusions regarding long-term outcomes. They should be interpreted with caution and do not establish definitive predictive validity. Future prospective studies with longer follow-up and larger numbers of damage accrual events are needed to determine the performance of SLE-DAS in predicting long-term irreversible organ damage in SLE.
Established SLE disease activity indices, including SLEDAI-2 K and SLE-DAS, are designed to assess intrinsic disease activity based on clinical and laboratory manifestations rather than treatment-related variables. In contrast, SLEDAI-2 K glucocorticoid index (SLEDAI-2 KG) incorporates weighted scoring for glucocorticoid dose, which may enable a more refined assessment of disease severity within the SLEDAI-2 K framework [24]. Thus, we compared SLE-DAS not only with SLEDAI-2 K but also with SLEDAI-2 KG. SLEDAI-2 KG was categorized using the same thresholds as SLEDAI-2 K for consistency. We observed the agreement between SLE-DAS and SLEDAI-2 KG (quadratically weighted Cohen’s κ = 0.652) and was higher than that between SLE-DAS and SLEDAI-2 K (quadratically weighted Cohen’s κ = 0.539), indicating a closer alignment between SLE-DAS and glucocorticoid-adjusted disease activity assessment. However, we consider that this finding should be interpreted with caution. There is ongoing debate as to whether current glucocorticoid dose should be included as a component of disease activity scores, as doing so may confound disease activity with physician prescribing behavior and potentially reduce comparability across patients and centers. In line with international recommendations, glucocorticoid exposure is considered a treatment modifier and a therapeutic outcome rather than a direct measure of disease activity.
Several limitations warrant consideration. First, the cross-sectional design may restrict the generalizability of our findings. Second, the ability to demonstrate superior discriminative capacity of SLE-DAS for cardiopulmonary involvement was limited by the moderate sample size and low prevalence of severe manifestations. Finally, the lack of a gold standard for disease activity assessment remains a fundamental challenge, as all current instruments have inherent limitations. Consequently, our validation relied primarily on comparisons with SLEDAI-2 K and PGA, which may not fully capture the performance of SLE-DAS. Future studies involving comparisons with the BILAG index are warranted.
Conclusions
In conclusion, our study demonstrates that SLE-DAS is a reliable instrument for assessing SLE disease activity. Refinements in the evaluation of nephritis and hematological involvement likely contribute to its superior performance compared to SLEDAI-2 K. SLE-DAS also showed a more consistent association with short-term organ damage accrual compared with SLEDAI-2 K. However, the low agreement observed between SLE-DAS and SLEDAI-2K in patients with severe disease emphasizes the need to improve SLE-DAS in the future. Moreover, further research is needed to evaluate the utility of SLE-DAS in capturing dynamic clinical changes, predicting long-term organ damage, and evaluating treatment response.
Acknowledgements
We express our sincere gratitude to Tung Wah Medical and Healthcare Co., Ltd. for their expertise and support in the development of the electronic database platform.
Medical Writing/Editorial Assistance
The authors received no medical writing or editorial assistance.
Author Contribution
Zhuoli Zhang conceived and critically revised the manuscript. Tingting Ding had full access to all of the data collection, analysis, interpretation, and drafted the manuscript. Yamin Li, Dai Gao, Lanlan Ji, and Xiaohui Zhang also contributed to data collection and analysis. All the authors approved the publication of the manuscript.
Funding
This work and the journal’s Rapid Service Fee were supported by the Key Program of the Major Research Plan of the National Natural Science Foundation of China (Grant No. 92474201) and the National High-Level Hospital Clinical Research Funding (Grant No. 2024HQ06).
Data Availability
The datasets supporting the findings of this study are available from the corresponding author on reasonable request.
Declarations
Conflict of Interest
Tingting Ding, Yamin Li, Dai Gao, Lanlan Ji, Xiaohui Zhang, and Zhuoli Zhang have declared no financial or commercial conflicts of interest.
Ethical Approval
The ethics committee of Peking University First Hospital approved this study (project 2017[1284]), with informed consent obtained from all participants.
References
- 1.Parra Sánchez AR, Voskuyl AE, van Vollenhoven RF. Treat-to-target in systemic lupus erythematosus: advancing towards its implementation. Nat Rev Rheumatol. 2022;18(3):146–57. 10.1038/s41584-021-00739-3. [DOI] [PubMed] [Google Scholar]
- 2.Gladman DD, Ibañez D, Urowitz MB. Systemic Lupus Erythematosus Disease Activity Index 2000. J Rheumatol. 2002;29(2):288–91. [PubMed] [Google Scholar]
- 3.Jesus D, Matos A, Henriques C, Zen M, Larosa M, Iaccarino L, et al. Derivation and validation of the SLE disease activity score (SLE-DAS): a new SLE continuous measure with high sensitivity for changes in disease activity. Ann Rheum Dis. 2019;78(3):365–71. 10.1136/annrheumdis-2018-214502. [DOI] [PubMed] [Google Scholar]
- 4.Ma J, Zhang L, Yan M, Liu Z, Xue L. Value of SLE-DAS in assessing disease activity in patients with systemic lupus erythematosus: a single-centre retrospective study. Lupus Sci Med. 2024. 10.1136/lupus-2024-001196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Assunção H, Jesus D, Larosa M, Henriques C, Matos A, Le Guern V, et al. Definition of low disease activity state based on the SLE-DAS: derivation and validation in a multicentre real-life cohort. Rheumatology (Oxford). 2022;61(8):3309–16. 10.1093/rheumatology/keab895. [DOI] [PubMed] [Google Scholar]
- 6.Heras-Recuero E, García-Fernández A, Blázquez-Sánchez T, Gómez-Moreno C, Ferraz-Amaro I, Llorca J, et al. Strong correlation between SLEDAI and SLE-DAS in the Spanish population: assessment of discordant patients. Semin Arthritis Rheum. 2025;73:152758. 10.1016/j.semarthrit.2025.152758. [DOI] [PubMed] [Google Scholar]
- 7.Saraiva L, Cunha RN, Jesus D, Gatto M, Zen M, Iaccarino L, et al. The SLE-DAS provides an accurate and feasible flare tool in the clinical setting: a validation study. Rheumatology (Oxford). 2024;63(4):1123–9. 10.1093/rheumatology/kead353. [DOI] [PubMed] [Google Scholar]
- 8.Cunha RN, Saraiva L, Jesus D, Doria A, da Silva JP, Inês LS. Predictors of flare in SLE patients fulfilling lupus low disease activity state: a cohort study of 292 patients with 36-month follow-up. Rheumatology (Oxford). 2023;62(11):3627–35. 10.1093/rheumatology/kead097. [DOI] [PubMed] [Google Scholar]
- 9.Onishi A, Tsuji H, Takase Y, Nakakubo Y, Iwasaki T, Kozuki T, et al. Comparisons of SLE-DAS and SLEDAI-2K and classification of disease activity based on the SLE-DAS with reference to patient-reported outcomes. Rheumatology (Oxford). 2023;62(12):3909–15. 10.1093/rheumatology/kead132. [DOI] [PubMed] [Google Scholar]
- 10.Ceccarelli F, Picciariello L, Natalucci F, Moretti V, Spinelli FR, Alessandri C, et al. Application of SLE-DAS to assess subcutaneous belimumab efficacy in a cohort of systemic lupus erythematosus patients. Clin Exp Rheumatol. 2024;42(7):1491–4. 10.55563/clinexprheumatol/l5976o. [DOI] [PubMed] [Google Scholar]
- 11.Hasegawa Y, Arinuma Y, Asakura H, Shindo R, Ino K, Kanayama Y, et al. Real-world efficacy of belimumab in achieving remission or low-disease activity in systemic lupus erythematosus: a retrospective study. Mod Rheumatol. 2024;34(4):732–40. 10.1093/mr/road078. [DOI] [PubMed] [Google Scholar]
- 12.Wang CL, Koo M, Hsu CW, Lu MC. Increased frequency of hospital admissions with active systemic lupus erythematosus disease activity defined by two different disease activity indices: a cohort study. Lupus. 2023;32(7):864–72. 10.1177/09612033231175268. [DOI] [PubMed] [Google Scholar]
- 13.Rodríguez-González MG, Valero-Gaona GA, Vargas-Aguirre T, Amezcua Guerra LM. Performance of the systemic lupus erythematosus disease activity score (SLE-DAS) in a Latin American population. Ann Rheum Dis. 2020;79(12):e158. 10.1136/annrheumdis-2019-216028. [DOI] [PubMed] [Google Scholar]
- 14.Mathew A, Chengappa KG, Shah S, Negi VS. SLE-DAS: ready for routine use? Ann Rheum Dis. 2020;79(9):e116. 10.1136/annrheumdis-2019-215704. [DOI] [PubMed] [Google Scholar]
- 15.Gao D, Hao Y, Mu L, Xie W, Fan Y, Ji L, et al. Frequencies and predictors of the Lupus Low Disease Activity State and remission in treatment-naïve patients with systemic lupus erythematosus. Rheumatology. 2020;59(11):3400–7. 10.1093/rheumatology/keaa120. [DOI] [PubMed] [Google Scholar]
- 16.Ji L, Gao D, Hao Y, Huang H, Wang Y, Deng X, et al. Low-dose glucocorticoids withdrawn in systemic lupus erythematosus: a desirable and attainable goal. Rheumatology. 2023;62(1):181–9. 10.1093/rheumatology/keac225. [DOI] [PubMed] [Google Scholar]
- 17.Aringer M, Costenbader K, Daikh D, Brinks R, Mosca M, Ramsey-Goldman R, et al. 2019 European League Against Rheumatism/American College of Rheumatology classification criteria for systemic lupus erythematosus. Ann Rheum Dis. 2019;78(9):1151–9. 10.1136/annrheumdis-2018-214819. [DOI] [PubMed] [Google Scholar]
- 18.Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1997;40(9):1725. 10.1002/art.1780400928. [DOI] [PubMed] [Google Scholar]
- 19.Fanouriakis A, Kostopoulou M, Andersen J, Aringer M, Arnaud L, Bae S-C, et al. EULAR recommendations for the management of systemic lupus erythematosus: 2023 update. Ann Rheum Dis. 2024;83(1):15–29. 10.1136/ard-2023-224762. [DOI] [PubMed] [Google Scholar]
- 20.Jesus D, Larosa M, Henriques C, Matos A, Zen M, Tomé P, et al. Systemic Lupus Erythematosus Disease Activity Score (SLE-DAS) enables accurate and user-friendly definitions of clinical remission and categories of disease activity. Ann Rheum Dis. 2021;80(12):1568–74. 10.1136/annrheumdis-2021-220363. [DOI] [PubMed] [Google Scholar]
- 21.Franklyn K, Lau CS, Navarra SV, Louthrenoo W, Lateef A, Hamijoyo L, et al. Definition and initial validation of a Lupus Low Disease Activity State (LLDAS). Ann Rheum Dis. 2016;75(9):1615–21. 10.1136/annrheumdis-2015-207726. [DOI] [PubMed] [Google Scholar]
- 22.Yang L, Gu B, Wang X, Ren Q, Shen M, Su D. Association of disease activity with depression and anxiety in systemic lupus erythematosus: a comparison of SLEDAI-2K and SLE-DAS. Rheumatology (Oxford). 2025;64(2):632–8. 10.1093/rheumatology/keae070. [DOI] [PubMed] [Google Scholar]
- 23.Lam F, Chan KL, To CH, Mok CC. Disease activity and quality of life in patients with systemic lupus erythematosus: a validation study using the SLE-DAS. Clin Exp Rheumatol. 2025. 10.55563/clinexprheumatol/1n8j8d. [DOI] [PubMed] [Google Scholar]
- 24.Touma Z, Gladman DD, Su J, Anderson N, Urowitz MB. A novel lupus activity index accounting for glucocorticoids: SLEDAI-2K glucocorticoid index. Rheumatology (Oxford). 2018;57(8):1370–6. 10.1093/rheumatology/key103. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets supporting the findings of this study are available from the corresponding author on reasonable request.