Risk factors associated with mortality in systemic lupus erythematosus. A case-control study in a tertiary care center in Mexico City.

OBJECTIVE: To identify the mortality risk factors in a group of Mexican patients with SLE.
METHODS: A case-control autopsy study in a tertiary care center in Mexico, City. Patients with SLE who died during 1958 to 1994 with an autopsy study were selected as cases, and alive patients matched by age (+/- 3 years), decade of SLE onset, and disease duration (+/- 5 years) were defined as controls. Clinical charts were reviewed looking at clinical variables. SLE disease activity was evaluated with the MexSledai index, and SLE disease severity with the Severity Index. Variables were classified as present at any moment during the follow-up and 3 months before death in cases or cut-off date in controls. STATISTICAL ANALYSIS: matched univariate and multivariate analysis by multiple logistic regression were performed, and the results were presented as odds ratio and 95% confidence intervals (OR, 95%CI).
RESULTS: 76 matched pairs of patients were studied. Age, gender, and years offormal education were similar in the cases and controls. Variables associated with mortality three months before death were: lung involvement OR= 15.6, 95%CI (4.8-50.3), p<0.001; severe thrombocytopenia 9.6 (2.9-31.7), p<0.001; heart involvement 5.8 (2.6-13.0), p<0.001; and the severity index (cases 8.8 mu, 2.4 sigma vs controls 3.5, 2.0, respectively) 2.2 (1.5-3.4), p<0.001. Variables associated with mortality detected at any moment before death were kidney involvement 2.16 (1.09-4.29), p<0.02; the steroid therapeutic index 2.3 (1.2-4.5), p<0.001; number of previous admissions 2.4 (1.4-4.3), p<0.001; the MEX-SLEDAI index (cases 21.6 mu 6.3 sigma vs controls 12.6, 5.8), 1.2 (1.1-1.3), p<0.001; and the number of severe infections 14.4 (4.4-46.2), p<0.001. Protective variables were skin involvement 0.1 (0.3-0.6), p<0.001; daily dose of chloroquine (cases 3.9 mu, 24.1 sigma vs controls 39.4, 60.0 mg), p <0.0001 and the time from thefirst SLE symptom to the patient's demise or the cut-off date 0.7(0.6-0.9), p<0.001. Multiple logistic regression showed that the model which best explained mortality consisted of a severity index 2.6 (1.7-3.8), p<0.001; heart disease 6.5 (1.5-28.2), p=0.01, and steroid therapeutic index 3.3 (1.6-6.6), p=0.001.
CONCLUSIONS: An active SLE with multi-organic involvement, steroids and infections were associated with mortality in Mexican patients with lupus attended in a tertiary care center A protective effect of cutaneous disease and chloroquine use was observed.