Does erythrocyte sedimentation rate reflect and discriminate flare from infection in systemic lupus erythematosus? Correlation with clinical and laboratory parameters of disease activity.

To examine disease activity parameters in patients with systemic lupus erythematosus (SLE) experiencing flare, infection, both, or neither condition, focusing on erythrocyte sedimentation rate (ESR). This study is a retrospective analysis of 371 consecutive inpatient SLE cases from 2006 to 2015. Cases were classified as flare (n = 147), infection (n = 48), both (n = 23), or neither (n = 135). ESR levels were correlated to C-reactive protein (CRP), ferritin, anti-dsDNA antibodies, complement C3 reduction, serositis, and erythrocyturia with proteinuria (Pearson's correlation). ESR levels were related to an age- and gender-adapted cut-off value (ESRp). We analyzed mean values of age, ESR, ESRp, CRP, ferritin and distribution of anti-dsDNA antibodies, C3 reduction, serositis, and erythrocyturia with proteinuria. Sensitivity and specificity were calculated via receiver operating characteristic or two-by-two table. Association of parameters with disease activity and infection was tested via two-sided chi square test. ESR correlated moderately with CRP in cases with flare and/or infection (r = 0.505-0.586). While ESR and CRP were normal in remission, mean values overlapped in cases with flare, infection, or both. ESRp was higher in flare than in infection (p = 0.048). ESR lost association to activity in infected cases, CRP to infection in flaring cases. ESRp, serositis, and anti-dsDNA antibodies were related to disease activity regardless of infections. Anti-dsDNA antibodies were most sensitive for detecting flares (74%), while serositis, proteinuria with erythrocyturia, anti-dsDNA antibodies, C3 reduction, and ESRp values ≥ 2 were most specific. ESR levels were raised by flares, infections, and age; adapting them to age and gender increased their diagnostic value. Obtaining several parameters remains necessary to differentiate flare from infection.