Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever

AUTREV-01687; No of Pages 8 Autoimmunity Reviews xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Autoimmunity Reviews journal homepage: www.elsevier.com/locate/autrev

1

Review

4Q4

Sara Beça a,b,1, Ignasi Rodríguez-Pintó a,1, Marco A. Alba a,1, Ricard Cervera a, Gerard Espinosa a,⁎

5 6

a

7

a r t i c l e

8 9 10 11

Article history: Received 8 February 2015 Accepted 11 February 2015 Available online xxxx

12 13 14 15 16 17

Keywords: Fever Flare Infection Algorithm Systemic lupus erythematosus

i n f o

R O

Department of Autoimmune Diseases, Hospital Clínic, Barcelona, Catalonia, Spain Department of Internal Medicine, Hospital Pedro Hispano, Matosinhos, Portugal

a b s t r a c t

E

D

P

Objective: To develop and validate a predictive risk calculator algorithm that assesses the probability of flare versus infection in febrile patients with systemic lupus erythematosus (SLE). Methods: We evaluated SLE patients admitted because of fever in the Department of Autoimmune Diseases of our Hospital between January 2000 and February 2013. Included patients were those with final diagnosis of infection, SLE flare or both. Data on clinical manifestations, treatment and laboratory results were collected. Variables considered clinically relevant were used to build up all possible logistic regression models to differentiate flares from infections. Best predictive variables for SLE relapses based on their higher area under the receiver operating characteristic (ROC) curve (AUC) were selected to be included in the calculator. The algorithm was developed in a random sample of 60% the cohort and validated in the remaining 40%. Results: One hundred and thirty SLE patients presented 210 episodes of fever. Fever was attributed to SLE activity and to infection in 45% and 48% of the cases, respectively. Three independent variables for prediction of flares were consistently selected by multivariate analysis: days of fever, anti-dsDNA antibody titres and C-reactive protein levels. Combination of these variables resulted in an algorithm with calculated AUC of 0.92 (95% CI: 0.87 to 0.97). The AUC for the validation cohort was of 0.79 (95% CI: 0.68 to 0.91). Conclusion: The proposed flare risk predictive calculator could be a useful diagnostic tool for differentiation between flares and infections in febrile SLE patients. © 2015 Elsevier B.V. All rights reserved.

E

C

T

b

O

F

3

Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever☆

2Q2

34

Q3

R

1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . Patients and methods . . . . . . . . . . . . . . . . . . . . 2.1. Patients . . . . . . . . . . . . . . . . . . . . . . . 2.2. Definitions . . . . . . . . . . . . . . . . . . . . . 2.3. Laboratory data . . . . . . . . . . . . . . . . . . . 2.4. Statistical analysis . . . . . . . . . . . . . . . . . . 3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. General characteristics . . . . . . . . . . . . . . . . 3.2. Flares . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Infections . . . . . . . . . . . . . . . . . . . . . . 3.4. Infection and flare . . . . . . . . . . . . . . . . . . 3.5. Differences between infectious episodes and SLE flares . 3.6. Predictors of flares and development of the risk algorithm 4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . Take-home messages . . . . . . . . . . . . . . . . . . . . . . Conflict of interest statement . . . . . . . . . . . . . . . . . . .

N C O

41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

Contents

U

37 40 39

R

38 36 35

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

☆ MA Alba was supported by Consejo Nacional de Ciencia y Tecnología (CONACyT), Mexico and by the Agencia de Gestió d'Ajuts Universitaris i de Recerca (AGAUR, Generalitat de Catalunya). ⁎ Corresponding author. Tel.: +34 93 2275774; fax: +34 93 2271707. E-mail address: [email protected] (G. Espinosa). 1 S Beça, I Rodríguez-Pintó and MA Alba contributed equally to this study.

http://dx.doi.org/10.1016/j.autrev.2015.02.005 1568-9972/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

2

57 58 59 60

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

Appendix A.

Definitions of specific infections . . . . . A.1. Definitions of specific organ relapses Appendix B. Supplementary data. . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

0 0 0 0

61

2.1. Patients

98

A retrospective cohort study was conducted by reviewing the medical records of adult patients with SLE admitted because of fever at the Department of Autoimmune Diseases of Hospital Clínic, between January 2000 and February 2013. Patients with final diagnosis of a SLE flare and/or an infection were included. Drug-induced lupus patients and overlapping autoimmune syndromes were excluded. Fever episodes on a same patient were recorded independently. This study was approved by the local ethics committee, and was conducted in compliance with the protocol for Good Clinical Practices and Declaration of Helsinki principles. Using an electronic case report, data encompassing more than 140 variables were collected according to a standardized protocol. SLE was established when 4 of the 11 revised criteria classification of the American College of Rheumatology were met [23]. Data of previous organ involvement was recorded in addition to number and type of immunosuppressive drugs received in the six months preceding the fever episode. Also, prednisone (PDN) dose used in the last three months and during admission episode was retrieved. Disease activity was measured

87 88 89 90 91 92 93 94

99 100 101 102 103 104 105 106 107 Q5 108 109 110 111 112 113 114 115

C

85 86

E

83 84

R

81 82

R

79 80

O

77 78

C

75 76

N

73 74

U

71 72

120

F

97

69 70

2.2. Definitions

117 118 119

Fever was defined as an axillary temperature greater than 37.5 °C [25,26]. International consensus guidelines were used for the diagnosis of specific infectious disorders. When guidelines were not available, definition was similar to that described in previous studies [27]. The concept of flare was based on an international consensus for description of recurrences in SLE patients [28] and in the European League Against Rheumatism (EULAR) recommendations [29]. This was defined as a measurable increase in disease activity in one or more organ systems involving new or worse clinical signs and symptoms and/or laboratory measurements. The terms flare, relapse, recurrence or exacerbation was used indistinctly. Definition of relapses affecting specific organs was based on previous studies [30–37], including the British Isles Lupus Assessment Group (BILAG-2004 index) [38]. Of note, in order to assess the utility of anti-double-stranded DNA antibodies (dsDNA-Ab) and complement levels to differentiate between flares and infections, our definition of flare did not included an increased dsDNA-Ab titre or low complement levels per se. Definitions used for infections and relapses affecting specific organs are included in the Appendix A. Coexistence of flare and infection was considered only in a reduced number of patients. In these patients, promptly resolution of symptoms was observed after the start of both antimicrobial therapy and an increase in PDN dose.

121

2.3. Laboratory data

144

Levels of the following laboratory results performed in the first 3 days of admission were obtained: ESR (normal value b20 mm/h), CRP (b1 mg/dL), hemoglobin (Hb, 12–17 mg/L), ferritin (18–160 ng/mL), lactic dehydrogenase (250–450 UI/L), complement C3 (0.82–1.87 g/L) and C4 (0.11–0.44 g/L) levels, white blood cell (WBC, 4–11 × 106/L) with differential count and urinalysis. In addition, dsDNA-Ab levels obtained in the current admission or in the nearest date within three months were recorded (reference value b14.9 U/mL).

145 146

2.4. Statistical analysis

153

Patients were categorized into three groups: SLE flares (group 1), infections (group 2) or both (group 3). First, main clinical and laboratory characteristics of the 3 groups are described. Continuous variables are presented as mean (SD) and categorical data as percentages. Association between selected covariates was analyzed using student's T test or ANOVA for quantitative variables. Fisher's exact test or χ2 test was used for categorical variables, depending on the validity of the underlying assumptions test for categorical data. Based on the data of groups 1 and 2, the identification of predictive variables for SLE relapses or infections was done using logistic regression analysis. Clinically relevant variables were considered for the multivariate approach. On the basis of previous literature [39–41], pre-defined potential predictors for differentiation of infections and flares were included (age at SLE diagnosis, previous history of lupus

154

O

2. Patients and methods

67 68

R O

96

65 66

P

95

Systemic lupus erythematosus (SLE) is a relapsing multisystem autoimmune disorder that may affect almost any organ [1]. Clinical presentation is broad and heterogeneous and non-specific constitutional symptoms are common in these patients [2]. Fever has been reported as part of SLE initial manifestations in 28–36% of patients and in 52– 60% during disease course [3,4]. Importantly, fever has been reported as one of the main causes leading to admission in this disease [5,6]. In SLE, fever can reflect an ongoing infection apart from being a manifestation of recurrences. Based on two retrospective series of hospitalized patients with this disorder, the estimated frequency of fever episodes that are of infectious origin or secondary to SLE activity is about 23–54% and 42–60%, respectively [7,8]. In clinical practice, differentiation between SLE flares and infections can be extremely difficult. On one side and despite current therapeutic regimes, relapses are still observed in 25–35% of lupus patients [9]. On the other side, immunosuppressive therapy used in moderate–severe cases increases the risk and severity of infections [10,11]. Infections are reported in 10–40% of SLE patients and are the main cause of death in 25–30% in large study cohorts [12,13]. To increase the complexity of this problem, systemic infections may also trigger SLE recurrences [14–16]. Based on these data, accurate discrimination of activity or infection in SLE patients presenting with a febrile episode is crucial, as treatment options are completely different. In this regard, several candidates have been evaluated as potential biomarkers for differentiation of SLE flares and infections: erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), procalcitonin (PCT), and the newer molecules pentraxin 3, soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) and neutrophil CD64+ [14,17–22]. The aim of this study was to develop and validate a predictive risk calculator algorithm that could assist daily clinical decisionmaking to differentiate flares from infections in SLE patients presenting with fever.

116

D

63 64

with the SLE Disease Activity Index (SLEDAI 2000) [24]. On the basis of a computer-generated randomized list of febrile episodes, cohort was divided into 2 sets. The first set (60% of all episodes) was used to develop the score. The other set of episodes (40%) was used to validate the score.

E

1. Introduction

T

62

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143

147 148 149 150 151 152

155 156 157 158 159 160 161 162 163 164 165 166 167

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

174 175 176 177 178 179 180 181 182 183 184 185 Q6 186

PðflareÞ ¼

Mucocutaneous Articular Hematologic Serositis Renal Muscular Cardiopulmonary Nervous system Articular & mucocutaneousa Articular & serositisa Articular & mucocutaneous & renala Articular & hematologic & renala

3.1. General characteristics

203 204

211

A total of 242 febrile episodes requiring admission to our Department were identified among 147 patients suffering from SLE. Thirtytwo episodes that occurred in 17 patients were excluded because fever was due to other causes instead of infection or SLE flare (drug related hypersensitivity, crystal induced arthritis, hemophagocytic syndrome, malignancy, deep venous thrombosis, Budd–Chiari syndrome or fever of unknown origin). Overall, 130 patients (83% women) presented 210 febrile episodes that were finally analyzed. Mean (SD) age at admission was 42 (15) years (range 18–82).

212

3.2. Flares

213 214

219 220

Clinical activity was responsible for 94/210 (45%) of episodes of fever. Flares with mucocutaneous involvement were the most frequent (40/94, 42.6%). Distribution of flares is showed in Table 1. Mean (SD) of PDN dose used to treat relapses was 34.7 (20.9) mg/day (0–90) which corresponds to an increase of 28 (21.8) mg/day (0–80) from doses used before flare. In 45/94 SLE flares more than one organ-system was involved (two in 27 and three or more in 18). Of interest, in the 6 months prior to current admission, 68% of these patients had a relapse.

221

3.3. Infections

222

An infectious disease was responsible for 101/210 (48%) febrile episodes. The most common diagnosis was urinary tract infection (UTI),

207 208 209 210

215 216 217 218

223

C

E R

R

N C O

205 206

U

199

227

Fifteen episodes (7%) presented with both disease activity and infection. The lower respiratory tract was the most frequent location for infection (CAP or bronchitis in 8 cases), while the most common targets of SLE activity were the hematological system (n = 6), kidneys and skin or oral mucosa (n = 5, each). Mean PDN dose for treating these patients was 37 (16) mg/day (15–60). Increase of PDN based on prerelapse dose was 23.2 (18.3) mg/day (0–56).

228

3.5. Differences between infectious episodes and SLE flares

235

Several significant differences were found in clinical and laboratory features when comparing febrile episodes corresponding to infections and SLE relapses (Table 3). Briefly, patients with SLE reactivation were younger and presented with a delayed history of symptoms and fever before the day of admission. As expected, low complements C3 and C4 and raised dsDNA-Ab levels were typical in flare episodes. In contrast, those suffering from infections were older, and had longer SLE evolution. These patients were exposed more frequently to PDN and immunosuppressive drugs and had higher levels of CRP and WBC.

236

3.6. Predictors of flares and development of the risk algorithm

245

Stepwise variable selection left lymphocytes and neutrophils count, CPR level, dsDNA-Ab titres, number of days with fever, number of days with symptoms, and chills as eligible variables. All possible logistic regression models mixing former quoted variables were created in order to select the best predictive model of SLE flare. Based on minimization of AIC and parsimony, the logistic regression model selected included three variables: number of days with fever, dsDNA-Ab titres, and CRP levels (Table 4). Collinearity was ruled out. The final equation was included in an interactive Excel spreadsheet (Appendix B). The ROC curve for the logistic model selected showed an AUC of 0.92 (95% CI: 0.87 to 0.97, Fig. 1). Assuming a frequency of flares of 48.2% in febrile SLE patients, a selected cut-off of 55% on the predicted risk algorithm would confer the following results: sensitivity of 80.5%, specificity of 88.6%, positive predictive value of 86.8% and negative predictive value of 83.0%. The model performed similarly in the 40% leftover episodes, with an AUC for the validation cohort of 0.79 (95% CI: 0.68 to 0.91).

246 247

P

202

197 198

3.4. Infection and flare

D

3. Results

195 196

t1:4 t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15

E

201

193 194

T

200

191 192

t1:3

40 (42.6) 39 (41) 23 (24.5) 22 (23.4) 20 (21.3) 4 (4) 3 (3) 3 (3) 8/45 (17.7) 5/45 (11.1) 3/45 (6.6) 3/45 (6.6)

recorded in 26/101 cases (25.7%), followed by community-acquired 224 pneumonia (CAP) documented in 23 (22.8%) (Table 2). Cultures were 225 positive in 36 cases (Table 2). 226

1 1 þ e−ðβ1 xþβ2 xþβ3 xþβ4 xþβ5 xþβ6 xþβ7 xþβ8 xþβ9 xþÞ

Results are presented as β coefficient, P value (P), odds ratio (OR) and 95% confidence interval (CI). The 40% leftover cases not used to develop the logistic regression model were used to estimate its validity. We estimate the area under the ROC (receiver operating characteristic) curve (AUC) for the derivation and the validation cohort, treading-off true positives (sensitivity) and false negatives (1-specificity) at all possible thresholds. In order to increase the model feasibility, we created a user-friendly interactive calculator in a Microsoft Excel spreadsheet (Microsoft, Redmond, Washington, US), which can be downloaded at supplementary material available online (Appendix B). The analyses were performed using SPSS for Windows, version 20.0 and the !ROC macro [43] for SPSS and the pROC package for R [44].

n (%)

a Most frequent combinations observed in 45 episodes with 2 or more organs t1:16 involved.

188 189 190

t1:1 t1:2

F

172 173

Table 1 Distribution of organ involvement observed in flare episodes (n = 94).

O

170 171

nephritis, therapy with PDN, hydroxychloroquine or immunosuppressive drugs, hypocomplementemia, leukopenia and elevated dsDNA-Ab values). Other relevant clinical and laboratory data obtained after univariate analysis were also incorporated. A random sample of 60% of the whole cohort of episodes of fever was selected to develop the logistic model. Variables were first selected through a forward and backward stepwise method. Variables selected were used to calculate all possible logistic regression models. The model that minimized the Akaike information criterion (AIC) was selected for further analysis. To improve its goodness of fit, reproducibility and reliability, CRP and dsDNA-Ab were categorized into different sorts. CRP levels were classified as low (≤ 5 mg/dL), medium (5–10 mg/dL) or high (N 10 mg/dL). Anti-dsDNA-Ab were categorized into low (≤ 20 UI/mL), medium (20–50 UI/mL), high (50–100 UI/mL) or very high titres (N100 UI/mL). Categorization of CRP and dsDNA-Ab was done in similar fashion as in previous studies [20,42]. Logistic regression coefficients for this model were obtained. Thus, an equation was found to be able to estimate the probability of flares by including selected patient data with high accuracy.

R O

168 169

3

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

229 230 231 232 233 234

237 238 239 240 241 242 243 244

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262

4 t2:1 t2:2

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

Table 2 Distribution of specific infections and causal agents recorded in 101 SLE patients.

t2:3

Infection

Total (n = 101)

With positive culture (n = 36)

Microorganism

t2:4

Urinary tract infection

26 (25.7)

t2:5 t2:6

Community-acquired pneumonia Acute gastroenteritis

23 (22.8) 19 (18.8)

Urine (n = 17) Blood (n = 6) Sputum (n = 1) Blood (n = 4)

Escherichia coli (n = 14); Klebsiella pneumoniae (n = 3) Escherichia coli (n = 6) Staphylococcus aureus Salmonella typhi; Listeria monocytogenes; Coagulase-negative staphylococci; Escherichia coli

t2:7 t2:8

Acute bronchitis Cutaneous abscess

15 (14.9) 4 (4)

t2:9

Cellulitis

3 (3)

t2:10 t2:11 t2:12

Upper airway infection Herpes zoster Abdominal abscess

3 (3) 3 (3) 2 (2)

t2:13 t2:14

Meningitis Catheter-related infection

2 (2) 1 (1)

None Blood (n = 1) Pus (n = 2) Blood (n = 1) Pus from ulcer (n = 1) None None Blood (n = 1) Pus (n = 1) None Blood (n = 1)

280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308

C

E

278 279

R

276 277

R

274 275

O

272 273

C

270 271

N

268 269

U

F

O

R O

266 267

P

Here, we have created a simple risk algorithm calculator to assist clinicians in a frequently faced dilemma: try to differentiate flares from infections in SLE patients. To the best of our knowledge, this is the first study conducted with this purpose. Infections and active disease are the leading cause of death in patients with SLE [13,45]. As both conditions may present with similar clinical findings, implementation of ancillary tools that accurately identify them could be used to assist the selection of appropriate diagnostic procedures, guide treatment and improve the prognosis of affected patients [46,47]. In this sense, serum complements C3 and C4, antibodies against complement fraction C1q, WBC count, urinary determination of free light chains of immunoglobulins, ESR, or dsDNA-Ab levels have been evaluated as potential diagnostic biomarkers of flares and infections in lupus cases [14,20,21,48–50]. Unfortunately, disappointing results have been reported as all of these candidates lack discriminative accuracy when considered individually [20,48–50]. In contrast to the aforementioned laboratory tests, CRP seems to differentiate better between infectious conditions and SLE relapses [20–22, 51]. CRP levels rise significantly in SLE patients with active bacterial infections but elevation is mild or even absent in the setting of a relapse [21,22]. Previous reports concluded that elevated CRP is a strong predictor of bacterial infections in SLE patients with estimated sensitivity of 100% and specificity of 90% [20,51]. PCT is another interesting candidate for identifying bacterial infections in autoimmune diseases. However, conflicting results have been reported [52–54]. In a retrospective study evaluating infected and non-infected SLE patients, PCT showed no ability to differentiate cases with or without bacterial infections as levels remain within the normal range in both groups [52]. In contrast, serum PCT was found to increase significantly in patients with SLE and non-viral infections when compared with patients presenting with flares [53]. In addition, PCT has also been proposed for monitoring the response to antibiotic treatment, as serum levels decrease after defervescence [53] and correlate with the severity of infection [55]. When compared to CRP, PCT seems to perform better for identifying superimposed infections in the presence of underlying active SLE although CRP is superior in the case of patients in remission [55,56]. In addition, another study found that CRP was more sensitive and specific than PCT to differentiate bacterial infections from disease flares [51]. Based on the previous data, it seems clear that no single clinical feature or laboratory result can distinguish between flares and infections in this heterogeneous disease [57]. Therefore, our proposed algorithm is based on the combination of one clinical variable (days of fever) and two conventional serologic tests (CRP and dsDNA-Ab) that were shown to be independent risk factors for flares after multivariate

analysis. The SLE flare algorithm calculator performed well, with resulted AUC of 0.92 (95% CI: 0.87 to 0.97), suggesting that it may be a useful tool to differentiate flares from infections. As SLE is a disease with heterogeneous clinical scenarios, a simple flare risk calculator used to organize data and stratify patients could be of relevance. Similar efforts include the Global APS score (AUC 0.73) [58] and the aPL-S (AUC 0.75) [59], which estimate the risk of thrombosis in SLE and in patients with antiphospholipid antibodies, respectively. We notice that clinical and biochemical features of patients in the present study resemble those reported by others [5,13,41,60–62]. This is relevant because the usefulness of any risk algorithm system depends on the similarity between the population from which it was elaborated and the one in which it will be applied. Therefore, the reported frequency of infections and flares in lupus patients hospitalized because of fever is 30% to 54% and 42% to 60%, respectively, similar to the results of the present study [5,60,63]. As in our study, the most common sites of infection reported were the respiratory (12% to 54%) and urinary (8% to 36%) tracts with bacteria being the most frequent microorganisms identified [5,62]. Our study has several limitations. Due to its retrospective nature, the diagnostic evaluation protocol for the identification of infections and flares was not uniform. Most of the infections identified in this study were of bacterial and viral origin and therefore accuracy of the calculator for differentiation of flares versus infection due to other microorganisms needs further investigation. Moreover, our study included only hospitalized patients, which probably implies a more aggressive or extended disease. This fact may have affected our results, incurring in a not controlled selection bias. In addition, the flare probability given by the proposed model is based on the information that SLE activity is usually associated with lower CRP levels and raised dsDNA-Ab titres. However, CRP levels could be considerably elevated in organ-specific SLE activity, as in lung involvement or serositis [64–66], and reduce promptly with antimalarials or steroids [67]. Also, persistently elevated levels of dsDNA-Ab can be found in some patients with chronically mild SLE activity [68]. Therefore, these factors may alter the effectiveness of the proposed algorithm in certain subgroups of patients. More importantly, adjustment for the different categories that we proposed will be probably necessary as normal values for CRP and dsDNA-Ab may differ between centers. Finally, we did not evaluate PCT as a potential variable in the risk algorithm calculator, as this test is not widely used in our institution. We must remark that, as it occurs with all clinical scores and algorithms, our proposal does not substitute clinical judgment, but assists clinical decision-making at the bedside of these patients. Strength of this study includes the large number of patients derived from a single center, all classified as having SLE by strict fulfillment of criteria. In addition, conditions identified in this series reflect the common clinical

D

264 265

Enterobacter

E

4. Discussion

Coagulase-negative staphylococci Enterococcus faecalis

T

263

Streptococcus pneumoniae Pseudomonas aeruginosa; Streptococcus pyogenes Streptococcus agalactiae Staphylococcus aureus

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx Table 3 Comparison of main characteristics of SLE febrile episodes related to infections or relapses.

52 (51.5) 69 (68.3) 16 (15.8) 17 (16.8)

35 (37.2) 64 (68.1) 14 (14.9) 20 (21.3)

ns ns ns ns

85 (84.2) 14 (14)

65 (69.1) 11.1 (8.4)

0.017 ns

16.1 (13.6)

10.4 (9.9)

0.037

47 (46.5) 38 (37.6)

29 (30.9) 40 (42.6)

0.028 ns

7.2 (10.6) 5.9 (10.4) 38.5 (0.66) 91 (17) 1 (1) 7 (6.9) 7 (6.9) 38 (37.6) 25 (23.8) 16 (15.8) 23 (22.8) 19 (18.8) 4.6 (4.8) (0–18)

45.1 (159.5) 36.7 (168.2) 38.4 (0.70) 84 (15) 38 (40.4) 20 (21.3) 37 (39.4) 15 (16) 5 (5.3) 6 (6.4) 5 (5.3) 4 (4.3) 12.6 (7.1) (3–31)

b0.0001 b0.0001 ns 0.023 b0.0001 0.006 b0.0001 0.001 b0.0001 0.043 b0.0001 0.001 b0.0001

72.3 (32.8) 9.9 (8.9) 9.72 (6.41) 8.01 (6.11) 1.01 (0.86) 437.5 (311.5) 31 (30.7) 0.92 (0.31) 23 (22.8) 0.19 (0.22) 53 (52.5) 52.7 (64.3)

65.5 (34.1) 4.3 (5.0) 6.01 (4.12) 4.44 (3.69) 0.94 (0.69) 485.3 (254.7) 52 (55.3) 0.68 (0.32) 53 (56.4) 0.12 (0.14) 76 (80.9) 114.4 (91.0)

ns b0.0001 b0.0001 b0.0001 ns 0.04 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001

CRP = C-reactive protein; dsDNA-Ab = double stranded DNA antibodies; ESR =

t3:48 t3:49 t3:50 t3:51 t3:52 t3:53 t3:54 t3:55 t3:56 t3:57

erythrocyte sedimentation rate; LDH = lactate dehydrogenase; PDN = prednisone; WBC = White blood cell count. a Used regularly in the last 3 months. b Includes cyclophosphamide, azathioprine, methotrexate or mycophenolate mophetil. When compared as individual drugs, no differences were found. c Recorded on the day of admission. d SLE Disease Activity Index [24]. e b0.82. f b0.11. g N14.9.

t4:1 t4:2

Table 4 Logistic regression, final model (dependent variable, flares).

N C O

t3:47

U

R O

O

F

b0.0001 0.002

Fig. 1. Receiver operating characteristic (ROC) curves for the proposed algorithm for the derivation (blue line) and the validation cohort (red line).

P

Laboratory results ESR (mean [SD], mm/h) CRP (mean [SD], mg/dL) WBC (mean [SD] 106/L) Neutrophils (mean [SD], 106/L) Lymphocytes (mean [SD], 106/L) LDH (mean [SD], U/L) Decreased C3 levels e C3 (mean [SD], g/L) Decreased C4 levelsf C4 (mean [SD], g/L) Elevated dsDNA-Abg dsDNA-Ab (U/mL)

ns

D

t3:34 t3:35 t3:36 t3:37 t3:38 t3:39 t3:40 t3:41 t3:42 t3:43 t3:44 t3:45 t3:46

80/14 (85.1/14.9) 37 (12) 548 (375)

scenarios found in daily practice. Also, laboratory variables included in the score are widely used and did not require special equipment. In conclusion, we have proposed a new algorithm to assess the risk of flares and differentiating them from infections in febrile SLE patients. Replication and use of the calculator algorithm in larger populations would be useful to prove its reliability beyond this study.

356

Take-home messages

363

E

t3:16 t3:17 t3:18 t3:19 t3:20 t3:21 t3:22 t3:23 t3:24 t3:25 t3:26 t3:27 t3:28 t3:29 t3:30 t3:31 t3:32 t3:33

Age at admission (mean [SD], years) Time from SLE diagnosis (mean [SD], weeks) Medical history (n, %) Previous lupus nephritis Previous flares Flares in the last 3 months Flares in the last 6 months Previous treatments (n, %) PDNa PDN at admission (mean [SD], mg/day) Mean (SD) PDN in the last 3 months (mg/day) Immunosuppressive drug a, b Hydroxychloroquinea Clinical manifestations (n, %) Days of symptoms (mean [SD]) Days of fever (mean [SD]) Temperature (mean [SD], °C) c Cardiac rate (mean [SD], per min) Arthritis Myalgias Cutaneous rash Cough Sputum production Vomit Diarrhea Abdominal pain SLEDAI (mean [SD], range) d

89/12 (88.1/11.9) 46 (16) 748 (443)

T

t3:5 t3:6 t3:7 t3:8 t3:9 t3:10 t3:11 t3:12 t3:13 t3:14 t3:15

P

C

Sex, female/male (n, %)

Flares (n = 94)

E

t3:4

Infections (n = 101)

R

t3:3

R

t3:1 Q1 t3:2

5

• In SLE patients, fever can be a sign of an ongoing infection or disease activity. • In daily clinical practice, differentiation between SLE flares and infections is usually difficult, as no single clinical or laboratory finding is reliable for this purpose. • The proposed risk algorithm for flares prediction performed well, and could be useful to differentiate between relapses and infections in febrile SLE patients. The impact of this finding needs to be assessed in large cohorts.

Variable

β coefficient

P

OR (95% CI)

t4:4 t4:5 t4:6 t4:7 t4:8 t4:9 t4:10 t4:11 t4:12

Days of fever CRP

0.25 0 (reference) 1.79 0.611 0 (reference) 0.55 2.65 1.82 −4.12

b0.001

1.28 (1.1–1.5) 1 6.01 (0.79–45.7) 1.84 (0.2–16.2) 1 1.74 (0.25–12.1) 14.15 (1.75–114.5) 6.14 (1.3–29.6) 0.016

t4:13

dsDNA-Ab

Intercept

0.576 0.013 0.024 b0.001

361 362

364 365 366 367 368 369 370 371 372 373

375

The authors declare no conflicts of interest.

0.083 0.58

359 360

374

Conflict of interest statement

t4:3

Low (≤5 mg/dL) Medium (5–10 mg/dL) High (N10 mg/dL) Low (≤20 UI/mL) Medium (20–50 UI/mL) High (50–100 UI/mL) Very high (N100 UI/mL)

357 358

CRP = C-reactive protein; dsDNA-Ab = double stranded DNA antibodies.

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 Q7 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437

F

O

R O

396 397

Appendix B. Supplementary data

A.1. Definitions of specific organ relapses –1) Articular: arthralgias with arthritis; –2) Mucocutaneous: malar rash, oral ulcers, discoid lesions, panniculitis, alopecia, digital infarcts and cutaneous vasculitis; –3) Serositis: pleuritis, pericarditis and pericardial or pleural effusion; –4) Hematological: drop to platelet count b 100,000/mm3, active positive Coombs hemolytic anemia with hemoglobin level b 10 mg/dL or lupus adenitis;

438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456

Supplementary data to this article can be found online at http://dx. 457 doi.org/10.1016/j.autrev.2015.02.005. 458 References

[1] Smith PP, Gordon C. Systemic lupus erythematosus: clinical presentations. Autoimmun Rev 2010;10:43–5. [2] Alarcon GS, Friedman AW, Straaton KV, Moulds JM, Lisse J, Bastian HM, et al. Systemic lupus erythematosus in three ethnic groups: III. A comparison of characteristics early in the natural history of the LUMINA cohort. LUpus in MInority populations: NAture vs. Nurture. Lupus 1999;8:197–209. [3] Pons-Estel BA, Catoggio LJ, Cardiel MH, Soriano ER, Gentiletti S, Villa AR, et al. The GLADEL multinational Latin American prospective inception cohort of 1,214 patients with systemic lupus erythematosus: ethnic and disease heterogeneity among “Hispanics”. Medicine (Baltimore) 2004;83:1–17. [4] Cervera R, Khamashta MA, Font J, Sebastiani GD, Gil A, Lavilla P, et al. Systemic lupus erythematosus: clinical and immunologic patterns of disease expression in a cohort of 1,000 patients. The European Working Party on Systemic Lupus Erythematosus. Medicine (Baltimore) 1993;72:113–24. [5] Zhou WJ, Yang CD. The causes and clinical significance of fever in systemic lupus erythematosus: a retrospective study of 487 hospitalised patients. Lupus 2009;18: 807–12. [6] Zimlichman E, Rothschild J, Shoenfeld Y, Zandman-Goddard G. Good prognosis for hospitalized SLE patients with non-related disease. Autoimmun Rev 2014;13: 1090–3. [7] Stahl NI, Klippel JH, Decker JL. Fever in systemic lupus erythematosus. Am J Med 1979;67:935–40. [8] Zhou WJ, Yang CD. The causes and clinical significance of fever in systemic lupus erythematosus: a retrospective study of 487 hospitalised patients. Lupus 2009;18: 807–12. [9] Barr SG, Zonana-Nacach A, Magder LS, Petri M. Patterns of disease activity in systemic lupus erythematosus. Arthritis Rheum 1999;42:2682–8. [10] Cervera R, Doria A, Amoura Z, Khamashta M, Schneider M, Guillemin F, et al. Patterns of systemic lupus erythematosus expression in Europe. Autoimmun Rev 2014;13: 621–9. [11] Seguro LP, Rosario C, Shoenfeld Y. Long-term complications of past glucocorticoid use. Autoimmun Rev 2013;12:629–32. [12] Abu-Shakra M, Urowitz MB, Gladman DD, Gough J. Mortality studies in systemic lupus erythematosus. Results from a single center. I. Causes of death. J Rheumatol 1995;22:1259–64. [13] Cervera R, Khamashta MA, Font J, Sebastiani GD, Gil A, Lavilla P, et al. Morbidity and mortality in systemic lupus erythematosus during a 10-year period: a comparison of early and late manifestations in a cohort of 1,000 patients. Medicine (Baltimore) 2003;82:299–308. [14] Sciascia S, Ceberio L, Garcia-Fernandez C, Roccatello D, Karim Y, Cuadrado MJ. Systemic lupus erythematosus and infections: clinical importance of conventional and upcoming biomarkers. Autoimmun Rev 2012;12:157–63. [15] Doria A, Canova M, Tonon M, Zen M, Rampudda E, Bassi N, et al. Infections as triggers and complications of systemic lupus erythematosus. Autoimmun Rev 2008;8:24–8. [16] Rigante D, Mazzoni MB, Esposito S. The cryptic interplay between systemic lupus erythematosus and infections. Autoimmun Rev 2014;13:96–102. [17] Allen E, Bakke AC, Purtzer MZ, Deodhar A. Neutrophil CD64 expression: distinguishing acute inflammatory autoimmune disease from systemic infections. Ann Rheum Dis 2002;61:522–5. [18] Pyo JY, Park JS, Park YB, Lee SK, Ha YJ, Lee SW. Delta neutrophil index as a marker for differential diagnosis between flare and infection in febrile systemic lupus erythematosus patients. Lupus 2013;22:1102–9.

D

394 395

E

392 393

T

390 391

C

388 389

E

387

R

385 386

R

383 384

O

381 382

C

379 380

– Community-acquired pneumonia (CAP) was the term applied to the combination of cough, sputum production, dyspnea and audible crackles on physical examination with supporting evidence of a demonstrable infiltrate on chest X-ray or computed tomography (CT) [69]. Microbiological data recorded (when available) consisted in sputum Gram stain, blood or sputum cultures and urinary antigen determination for pneumococcus or Legionella. – Acute bronchitis was considered in the presence of cough lasting more than five days but less than three weeks, which could include sputum production and/or dyspnea with no evidence of pneumonia or other related complications (empyema, pleural effusion) observed in chest X-ray or CT [70,71]. – Acute pharyngitis, common cold and acute rhinosinusitis were all considered under the term upper tract respiratory infections. Symptoms include combination of malaise, sore throat, cough, headache, nasal congestion, hoarseness, sinus tenderness, nasal discharge, maxillary discomfort, and/or facial pain. Sinusitis was confirmed by plain radiographs or CT [72,73]. – Acute gastroenteritis (acute diarrhea) was defined as a decrease in consistency and an increase in frequency of bowel movements (≥3 stools per day) that resolved in b14 days. Accompanying symptoms could include nausea, vomiting, or abdominal cramps. No stool cultures were necessary for diagnosis [74]. – Abdominal abscess was considered a collection of purulent material confirmed by ultrasound, CT or magnetic resonance imaging (MRI), expressed clinically with symptoms of gastrointestinal dysfunction such as anorexia, nausea, vomiting, bloating, and/or obstipation [75]. – The term urinary tract infection (UTI) encompasses acute cystitis and uncomplicated pyelonephritis. This was recorded when dysuria, frequency or urgency (accompanied or not by suprapubic pain) was associated with pyuria or hematuria in urinalysis (either by microscopy or by dipstick). Urinalysis in the absence of urine culture was sufficient for diagnosis of uncomplicated cystitis if symptoms were consistent with it [76,77]. – Meningitis was considered when headache and meningismus accompanied by nausea, photophobia or cognitive alterations were associated with raised white blood cells (WBCs) and protein levels in cerebrospinal fluid (CSF). Designation of aseptic meningitis was done when CSF cultures for bacteria and fungus remain sterile. – Cellulitis and erysipelas. These were defined as skin erythema, oedema and warmth that involve the deeper dermis and subcutaneous fat (cellulitis) or the upper dermis and superficial lymphatics (erysipela). Blood cultures, needle aspirations, or punch biopsies were not necessary for diagnosis but data were retrieved when available [78]. – Cutaneous abscess was defined as a collection of pus within the dermis and deeper skin tissues [78]. – Catheter-related bloodstream infection was considered when a patient with previous need of an indwelling catheter presented with symptoms of sepsis or complications as endocarditis or suppurative thrombophlebitis and positive blood cultures in the absence of other identifiable sources of infection [79].

N

377 378

–5) Renal: a) an increase in 24-hour urine protein levels to N 1 g if previous value was b0.5 g, to N3 g if the baseline value was N 0.5–1 g, or to more than twice the baseline value if the baseline value was N1 g, or b) an increase in serum creatinine level of N25% accompanied by proteinuria, hematuria (N 10 RBCs/hpf), and/or red blood cell (RBC) casts, or c) increase in hematuria by 2 grades compared with the last value, associated with new RBC cast or an increase in 24-hour urinary protein levels [80–83]; –6) Muscular: inflammatory myositis; –7) Cardiopulmonary: lupus pneumonitis, myocarditis, interstitial lung disease, pulmonary hypertension, alveolar hemorrhage or shrinking lung syndrome [80]; –8) Gastrointestinal: lupus-related autoimmune hepatitis, proteinlosing enteropathy, colitis or pancreatitis secondary to SLE vasculitis; –9) Neuropsychiatric: one of the 19 described neuropsychiatric lupus syndromes [84].

Appendix A. Definitions of specific infections

U

376

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

P

6

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

[53]

[54]

[55]

[56]

[57] [58]

[59]

D

[60]

C

E

R

R

N C O

F

[52]

O

[51]

R O

[50]

disease activity in patients with systemic lupus erythematosus. Arthritis Rheum 1985;28:904–13. Linnik MD, Hu JZ, Heilbrunn KR, Strand V, Hurley FL, Joh T. Relationship between anti-double-stranded DNA antibodies and exacerbation of renal disease in patients with systemic lupus erythematosus. Arthritis Rheum 2005;52:1129–37. Mastroianni-Kirsztajn G, Nishida SK, Pereira AB. Are urinary levels of free light chains of immunoglobulins useful markers for differentiating between systemic lupus erythematosus and infection? Nephron Clin Pract 2008;110:c258–63. Kim HA, Jeon JY, An JM, Koh BR, Suh CH. C-reactive protein is a more sensitive and specific marker for diagnosing bacterial infections in systemic lupus erythematosus compared to S100A8/A9 and procalcitonin. J Rheumatol 2012;39:728–34. Lanoix JP, Bourgeois AM, Schmidt J, Desblache J, Salle V, Smail A, et al. Serum procalcitonin does not differentiate between infection and disease flare in patients with systemic lupus erythematosus. Lupus 2011;20:125–30. Shin KC, Lee YJ, Kang SW, Baek HJ, Lee EB, Kim HA, et al. Serum procalcitonin measurement for detection of intercurrent infection in febrile patients with SLE. Ann Rheum Dis 2001;60:988–9. Quintana G, Medina YF, Rojas C, Fernandez A, Restrepo JF, Rondon F, et al. The use of procalcitonin determinations in evaluation of systemic lupus erythematosus. J Clin Rheumatol 2008;14:138–42. Yu J, Xu B, Huang Y, Zhao J, Wang S, Wang H, et al. Serum procalcitonin and C-reactive protein for differentiating bacterial infection from disease activity in patients with systemic lupus erythematosus. Mod Rheumatol 2013. Bador KM, Intan S, Hussin S, Gafor AH. Serum procalcitonin has negative predictive value for bacterial infection in active systemic lupus erythematosus. Lupus 2012;21: 1172–7. Agmon-Levin N, Mosca M, Petri M, Shoenfeld Y. Systemic lupus erythematosus one disease or many? Autoimmun Rev 2012;11:593–5. Sciascia S, Sanna G, Murru V, Roccatello D, Khamashta MA, Bertolaccini ML. GAPSS: the Global Anti-Phospholipid Syndrome Score. Rheumatology (Oxford) 2013;52: 1397–403. Otomo K, Atsumi T, Amengual O, Fujieda Y, Kato M, Oku K, et al. Efficacy of the antiphospholipid score for the diagnosis of antiphospholipid syndrome and its predictive value for thrombotic events. Arthritis Rheum 2012;64:504–12. Stahl NI, Klippel JH, Decker JL. Fever in systemic lupus erythematosus. Am J Med 1979;67:935–40. Hidalgo-Tenorio C, Jimenez-Alonso J, de Dios Luna J, Tallada M, Martinez-Brocal A, Sabio JM. Urinary tract infections and lupus erythematosus. Ann Rheum Dis 2004; 63:431–7. Cuchacovich R, Gedalia A. Pathophysiology and clinical spectrum of infections in systemic lupus erythematosus. Rheum Dis Clin North Am 2009;35:75–93. Jallouli M, Frigui M, Marzouk S, Maaloul I, Kaddour N, Bahloul Z. Infectious complications in systemic lupus erythematosus: a series of 146 patients. Rev Med Interne 2008;29:626–31. Mochizuki T, Aotsuka S, Satoh T. Clinical and laboratory features of lupus patients with complicating pulmonary disease. Respir Med 1999;93:95–101. Moutsopoulos HM, Mavridis AK, Acritidis NC, Avgerinos PC. High C-reactive protein response in lupus polyarthritis. Clin Exp Rheumatol 1983;1:53–5. Lee SS, Singh S, Link K, Petri M. High-sensitivity C-reactive protein as an associate of clinical subsets and organ damage in systemic lupus erythematosus. Semin Arthritis Rheum 2008;38:41–54. Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest 2003; 111:1805–12. Birmingham DJ, Irshaid F, Nagaraja HN, Zou X, Tsao BP, Wu H, et al. The complex nature of serum C3 and C4 as biomarkers of lupus renal flare. Lupus 2010;19:1272–80.

P

[49]

E

[61]

T

[19] Jackish J, Somers E, McCune J. Comparison of ESR (erythrocyte sedimentation rate) and CRP (C-reactive protein) in lupus patients presenting with fever. Abstract. American College of Rheumatology Annual Scientific Meeting; 2006. p. L22 [Washington DC]. [20] Firooz N, Albert DA, Wallace DJ, Ishimori M, Berel D, Weisman MH. High-sensitivity C-reactive protein and erythrocyte sedimentation rate in systemic lupus erythematosus. Lupus 2011;20:588–97. [21] Pereira Da Silva JA, Elkon KB, Hughes GR, Dyck RF, Pepys MB. C-reactive protein levels in systemic lupus erythematosus: a classification criterion? Arthritis Rheum 1980;23:770–1. [22] Bertouch JV, Roberts-Thompson PJ, Feng PH, Bradley J. C-reactive protein and serological indices of disease activity in systemic lupus erythematosus. Ann Rheum Dis 1983;42:655–8. [23] Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997;40:1725. [24] Gladman DD, Ibanez D, Urowitz MB. Systemic lupus erythematosus disease activity index 2000. J Rheumatol 2002;29:288–91. [25] Mackowiak PA, Wasserman SS, Levine MM. A critical appraisal of 98.6 degrees F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich. JAMA 1992;268:1578–80. [26] Niven DJ, Laupland KB, Tabah A, Vesin A, Rello J, Koulenti D, et al. Diagnosis and management of temperature abnormality in ICUs: a EUROBACT investigators' survey. Crit Care 2013;17:R289. [27] Anton JM, Castro P, Espinosa G, Marcos M, Gandia M, Merchan R, et al. Mortality and long term survival prognostic factors of patients with systemic autoimmune diseases admitted to an intensive care unit: a retrospective study. Clin Exp Rheumatol 2012;30:338–44. [28] Ruperto N, Hanrahan LM, Alarcon GS, Belmont HM, Brey RL, Brunetta P, et al. International consensus for a definition of disease flare in lupus. Lupus 2011;20: 453–62. [29] Gordon C, Bertsias G, Ioannidis JP, Boletis J, Bombardieri S, Cervera R, et al. EULAR points to consider for conducting clinical trials in systemic lupus erythematosus. Ann Rheum Dis 2009;68:470–6. [30] Zhu TY, Tam LS, Lee VW, Lee KK, Li EK. The impact of flare on disease costs of patients with systemic lupus erythematosus. Arthritis Rheum 2009;61:1159–67. [31] Petri MA, van Vollenhoven RF, Buyon J, Levy RA, Navarra SV, Cervera R, et al. Baseline predictors of systemic lupus erythematosus flares: data from the combined placebo groups in the phase III belimumab trials. Arthritis Rheum 65:2143-53. [32] Petri M, Buyon J, Kim M. Classification and definition of major flares in SLE clinical trials. Lupus 1999;8:685–91. [33] Houssiau FA, Vasconcelos C, D'Cruz D, Sebastiani GD, Garrido Ed Ede R, Danieli MG, et al. Immunosuppressive therapy in lupus nephritis: the Euro-Lupus Nephritis Trial, a randomized trial of low-dose versus high-dose intravenous cyclophosphamide. Arthritis Rheum 2002;46:2121–31. [34] Sinclair A, Appel G, Dooley MA, Ginzler E, Isenberg D, Jayne D, et al. Mycophenolate mofetil as induction and maintenance therapy for lupus nephritis: rationale and protocol for the randomized, controlled Aspreva Lupus Management Study (ALMS). Lupus 2007;16:972–80. [35] Moroni G, Quaglini S, Gallelli B, Banfi G, Messa P, Ponticelli C. The long-term outcome of 93 patients with proliferative lupus nephritis. Nephrol Dial Transplant 2007;22:2531–9. [36] Petri M, Genovese M, Engle E, Hochberg M. Definition, incidence, and clinical description of flare in systemic lupus erythematosus. A prospective cohort study. Arthritis Rheum 1991;34:937–44. [37] The American College of Rheumatology nomenclature and case definitions for neuropsychiatric lupus syndromes. Arthritis Rheum 1999;42:599–608. [38] Isenberg DA, Rahman A, Allen E, Farewell V, Akil M, Bruce IN, et al. BILAG 2004. Development and initial validation of an updated version of the British Isles Lupus Assessment Group's disease activity index for patients with systemic lupus erythematosus. Rheumatology (Oxford) 2005;44:902–6. [39] Danza A, Ruiz-Irastorza G. Infection risk in systemic lupus erythematosus patients: susceptibility factors and preventive strategies. Lupus 2013;22:1286–94. [40] Ruiz-Irastorza G, Olivares N, Ruiz-Arruza I, Martinez-Berriotxoa A, Egurbide MV, Aguirre C. Predictors of major infections in systemic lupus erythematosus. Arthritis Res Ther 2009;11:R109. [41] Ines L, Duarte C, Silva RS, Teixeira AS, Fonseca FP, da Silva JA. Identification of clinical predictors of flare in systemic lupus erythematosus patients: a 24-month prospective cohort study. Rheumatology (Oxford) 2014;53:85–9. [42] Gonzalez-Pulido C, Croca S, Abrol E, Isenberg DA. Long-term activity index after renal failure in a cohort of 32 patients with lupus nephritis. Clin Exp Rheumatol 2014;32:301–7. [43] Domenech J, Granero R. Macro !ROC for SPSS Statistics. Roc Analysis [computer program]. V2010.02.04. In: Editor, editor. Book Macro !ROC for SPSS Statistics. Roc Analysis [computer program]. V2010.02.04. Bellaterra: Universitat Autònoma de Barcelona; 2010 [pp.]. [44] Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, et al. pROC: an opensource package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 2011;12:77. [45] Borchers AT, Keen CL, Shoenfeld Y, Gershwin ME. Surviving the butterfly and the wolf: mortality trends in systemic lupus erythematosus. Autoimmun Rev 2004;3: 423–53. [46] Calixto OJ, Franco JS, Anaya JM. Lupus mimickers. Autoimmun Rev 2014;13:865–72. [47] Doria A, Gatto M, Zen M, Iaccarino L, Punzi L. Optimizing outcome in SLE: treatingto-target and definition of treatment goals. Autoimmun Rev 2014;13:770–7. [48] Valentijn RM, van Overhagen H, Hazevoet HM, Hermans J, Cats A, Daha MR, et al. The value of complement and immune complex determinations in monitoring

U

512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 Q9 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597

7

[62]

[63]

[64] [65] [66]

[67] [68]

598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 Q10 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652

Appendix A References

653 Q11

[69] Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44(Suppl. 2):S27–72. [70] Snow V, Mottur-Pilson C, Gonzales R. Principles of appropriate antibiotic use for treatment of acute bronchitis in adults. Ann Intern Med 2001;134:518–20. [71] Gonzales R, Bartlett JG, Besser RE, Cooper RJ, Hickner JM, Hoffman JR, et al. Principles of appropriate antibiotic use for treatment of uncomplicated acute bronchitis: background. Ann Intern Med 2001;134:521–9. [72] Chow AW, Benninger MS, Brook I, Brozek JL, Goldstein EJ, Hicks LA, et al. IDSA clinical practice guideline for acute bacterial rhinosinusitis in children and adults. Clin Infect Dis 2012;54:e72-112. [73] Pelucchi C, Grigoryan L, Galeone C, Esposito S, Huovinen P, Little P, et al. Guideline for the management of acute sore throat. Clin Microbiol Infect 18 Suppl 1:1-28. [74] Guerrant RL, Van Gilder T, Steiner TS, Thielman NM, Slutsker L, Tauxe RV, et al. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001;32:331–51. [75] Solomkin JS, Mazuski JE, Bradley JS, Rodvold KA, Goldstein EJ, Baron EJ, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis 2010;50:133–64. [76] Gupta K, Hooton TM, Naber KG, Wullt B, Colgan R, Miller LG, et al. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clin Infect Dis 2011;52:e103–20.

654 655 656 657 658 659 660 661 662 663 664 665 666 Q12 667 668 669 670 671 672 673 674 675 676 677 678 679

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

8

680 681 682 683 684 685 686 687 688 689 690 691

S. Beça et al. / Autoimmunity Reviews xxx (2015) xxx–xxx

[77] Wilson ML, Gaido L. Laboratory diagnosis of urinary tract infections in adult patients. Clin Infect Dis 2004;38:1150–8. [78] Stevens DL, Bisno AL, Chambers HF, Everett ED, Dellinger P, Goldstein EJ, et al. Practice guidelines for the diagnosis and management of skin and soft-tissue infections. Clin Infect Dis 2005;41:1373–406. [79] O'Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO, Maki DG, et al. Guidelines for the prevention of intravascular catheter-related infections. Centers for Disease Control and Prevention. MMWR Recomm Rep 2002;51:1–29. [80] Houssiau FA, Vasconcelos C, D'Cruz D, Sebastiani GD, Garrido Ed Ede R, Danieli MG, et al. mmunosuppressive therapy in lupus nephritis: the Euro-Lupus Nephritis Trial, a randomized trial of low-dose versus high-dose intravenous cyclophosphamide. Arthritis Rheum 2002;46:2121–31.

[81] Sinclair A, Appel G, Dooley MA, Ginzler E, Isenberg D, Jayne D, et al. Mycophenolate mofetil as induction and maintenance therapy for lupus nephritis: rationale and protocol for the randomized, controlled Aspreva Lupus Management Study (ALMS). Lupus 2007;16:972–80. [82] Moroni G, Quaglini S, Gallelli B, Banfi G, Messa P, Ponticelli C. The long-term outcome of 93 patients with proliferative lupus nephritis. Nephrol Dial Transplant 2007;22:2531–9. [83] Gordon C, Jayne D, Pusey C, Adu D, Amoura Z, Aringer M, et al. European consensus statement on the terminology used in the management of lupus glomerulonephritis. Lupus 2009;18:257–63. [84] The American College of Rheumatology nomenclature and case definitions for neuropsychiatric lupus syndromes. Arthritis Rheum 1999;42:599–608.

U

N

C

O

R

R

E

C

T

E

D

P

R O

O

F

704

Please cite this article as: Beça S, et al, Development and validation of a risk calculator to differentiate flares from infections in systemic lupus erythematosus patients with fever, Autoimmun Rev (2015), http://dx.doi.org/10.1016/j.autrev.2015.02.005

692 693 694 695 696 697 698 699 700 701 702 703