Mortality outcomes in pediatric rheumatology in the US.

OBJECTIVE

To describe mortality rates, causes of death, and potential mortality risk factors in pediatric rheumatic diseases in the US.

METHODS

We used the Indianapolis Pediatric Rheumatology Disease Registry, which includes 49,023 patients from 62 centers who were newly diagnosed between 1992 and 2001. Identifiers were matched with the Social Security Death Index censored for March 2005. Deaths were confirmed by death certificates, referring physicians, and medical records. Causes of death were derived by chart review or from the death certificate. Standardized mortality ratios (SMRs) and 95% confidence intervals (95% CIs) were determined.

RESULTS

After excluding patients with malignancy, 110 deaths among 48,885 patients (0.23%) were confirmed. Patients had been followed up for a mean +/- SD of 7.9 +/- 2.7 years. The SMR of the entire cohort was significantly decreased (0.65 [95% CI 0.53-0.78]), with differences in patients followed up for > or =9 years. The SMR was significantly greater for systemic lupus erythematosus (3.06 [95% CI 1.78-4.90]) and dermatomyositis (2.64 [95% CI 0.86-6.17]) but not for systemic juvenile rheumatoid arthritis (1.8 [95% CI 0.66-3.92]). The SMR was significantly decreased in pain syndromes (0.41 [95% CI 0.21-0.72]). Causes of death were related to the rheumatic diagnosis (including complications) in 39 patients (35%), treatment complications in 11 (10%), non-natural causes in 25 (23%), background disease in 23 (21%), and were unknown in 12 patients (11%). Rheumatic diagnoses, age at diagnosis, sex, and early use of systemic steroids and methotrexate were significantly associated with the risk of death.

CONCLUSION

Our findings indicate that the overall mortality rate for pediatric rheumatic diseases was not increased. Even for the diseases and conditions associated with increased mortality, mortality rates were significantly lower than those reported in previous studies.