Infection in Hospitalized Patients With Systemic Lupus Erythematosus: Proposal of an Algorithm for Diagnosis.

BACKGROUND

Identification of infection in patients with systemic lupus erythematosus (SLE) is a major challenge in clinical practice.

OBJECTIVE

This medical records review study evaluated clinical markers, including the performance of C-reactive protein (CRP), neutrophil-lymphocyte ratio (NLR), and platelet-lymphocyte ratio (PLR) in the diagnosis of infection in SLE patients.

METHODS

One hundred four SLE patients hospitalized between 2014 and 2018 were allocated into 3 groups, namely, infection, infection and disease activity, and isolated disease activity. Groups were compared in relation to clinical and laboratory variables. Accuracy measures were calculated for CRP, NLR, and PLR.

RESULTS

C-reactive protein, NLR, and PLR differed between the groups with higher values observed in the infected group, intermediate values in the mixed group, and lower values in the group with isolated activity-CRP (56 vs 26 vs 15 mg/dL, p = 0.002), NLR (7.9 vs 4.0 vs 3.1, p = 0.005), and PLR (270 vs 227 vs 134, p = 0.025). Fever, tachypnea, and PLR were independently associated with infection. The cutoff points of the CRP of 20 mg/L, NLR of 3.5, and PLR of 151.4 presented values of sensitivity and specificity for the prediction of infection equal to 67% and 67%, 65% and 58%, and 71% and 53%, respectively. The developed algorithm showed a sensitivity of 86.6% and specificity of 81% for the diagnosis of infection.

CONCLUSIONS

The combined use of clinical and laboratory markers presented superior accuracy than their isolated use, suggesting a great potential for the application of the algorithm in clinical practice.