Mortality Outcomes in Pediatric Rheumatology Conditions

Purpose: To describe the mortality rate and causes of death in pediatric rheumatology conditions in the U.S.
Methods: We used the Pediatric Rheumatology Disease Registry (PRDR), including 48,858 analyzable patients, newly diagnosed between 1992-2001 from 61 U.S. centers. Only initials and birth dates identified PRDR patients. Identifiers were matched with the Social Security Death Index censored for March 2005. 749 potential deceased were matched. Death certificates were obtained from the National Death Index and using standard rules were reviewed independently by 3 physicians blinded to the original PRDR diagnoses. 395 deaths were definitely, probably or possibly associated with rheumatic conditions. Physicians from the PRDR centers were contacted and 127 deaths (0.26% of entire cohort) were confirmed using medical records. The one-sample log-rank test was used to compare cohort mortality to the age, race, gender and calendar year adjusted U.S. population. Associations between mortality and demographic and clinical variables was assessed with Chi-square, Fisher’s exact or t-tests, as appropriate. Causes of death were derived from death certificates.
Results: The death rate of the entire cohort after 7.8 ± 2.7 yrs of follow-up (range 0-14.6) was significantly less than the adjusted U.S. population (P=0.004) with significantly less mortality seen in patients followed for >10 yrs. However the mortality rate was significantly greater in SLE [18/1383 (1.3%) deaths, mean time to death after diagnosis 5.6 ± 3.3 yrs, P<0.001], other connective tissue diseases [14/3325 (0.42%), 5.7 ± 4.0, P=0.042], hematologic diseases/malignancies [12/494 (2.4%), 2.5 ± 2.1, P<0.001] and systemic JRA [6/999 (0.6%), 5.5 ± 3.7, P=0.04]. The mean age at death was 16.0 ± 6.4 yrs and the mean follow up of those who died from diagnosis was 4.8 ± 3.4 yrs, significantly less than the whole cohort follow-up. 56% of the deceased were female, 44% male. 69% were Caucasian, 18% African American, 6% Hispanic and 6% other/mixed. Gender and race were not significant risk factors. Deceased patients were significantly older at diagnosis than survivors (11.3 ± 5.3 yrs vs. 10 ± 4.7, P<0.009). The main causes of death were disease-related or complications 32 (25%), infections 28 (22%), non-natural causes 23 (18%), background disease 12 (9%), other treatment complications (5%) and unknown/unclear 26 (20%).
Conclusions: In the first systematic study of mortality rates in pediatric rheumatology the mortality rate for new patients seen in the U.S. during the 1990s was not increased in relation to the general population. The mortality rate was significantly increased in SLE, other connective tissue diseases, malignancies and systemic JRA. Most deaths occurred relatively early after diagnosis and older patients at diagnosis had a higher mortality rate. Major mortality causes were from disease or disease complications, infections and non-natural causes.

 P.J. Hashkes, None; M.S. Lauer, None; B.M. Wright, None; S. Worley, None; P. Roettcher, None; S.L. Bowyer, None.