Key points are not available for this paper at this time.
We decided to analyse the incidence and characteristics of infection in systemic lupus erythematosus (SLE) and determine the related risks factors. One-hundred and ten SLE patients and 220 controls were prospectively followed up over three years and all the infectious episodes were recorded. A case-control design was established to identify risk factors of infection. Thirty-nine SLE patients suffered at least one infection (36%) versus 53 controls (22%), RR = 1.63 (P < 0.05). The incidence of urinary infections, pneumonia and bacteraemia without known focus was significantly greater in SLE. E. coli was the chief isolated microorganism (21.3%). In the univariate analysis, nephritis, SLE activity, leucopenia, anti-dsDNA levels above 20 IU/mL, CH50 levels under 300 IU/mL, ever use of steroids, daily dose of prednisone higher than 10 mg and ever use of cyclophosphamide were significantly associated with infection. In the multivariate analysis, total serum complement levels below 300 UU/mL and a daily dose of prednisone above 20 mg during at least one month plus ever use of cyclophosphamide were found to be significant (P < 0.0001). We conclude that patients with SLE have an increased overall risk for infection and they are especially prone to develop urinary infection, pneumonia and bacteraemia without focus. Hypocomplementaemia represents an independent predictive factor for infection. It seems mandatory to closely follow up SLE patients with low complement levels and instruct them to report any suspicious sign of infection, especially in those receiving more than 20 mg/day of prednisone who have also been administered cyclophosphamide.
Bosch et al. (Fri,) studied this question.