0711: Improvement in Excess Mortality in Patients with Rheumatoid Arthritis over the Last Two Decades: A Danish Population-based Matched Cohort Study

Bolette Soussi¹, Kirsten Duch¹, René Cordtz¹, Christian Bork², Salome Kristensen¹, Erik Schmidt², Jesper Lindhardsen³ and Lene Dreyer¹, ¹Center of Rheumatic Research Aalborg, Department of Rheumatology, Aalborg University Hospital, Aalborg, Denmark, ²Department of Cardiology, Aalborg University Hospital, Aalborg, Denmark, ³Lupus and Vasculitis Clinic, Center for Rheumatology and Spine Diseases, Rigshospitalet, Copenhagen, Denmark

Background/Purpose: Results from previous studies investigating mortality trends in patients with rheumatoid arthritis (RA) have been inconsistent, most likely due to varying cohort sizes and characteristics. We therefore sought to estimate the temporal trend in 5-year all-cause mortality in a large nationwide unselected cohort of RA patients.

Methods: Danish nationwide population-based matched cohort study. Patients with RA diagnosed between 1996 and the end of 2015 were identified by record linkage using Danish health registers and followed until the end of 2020; thus, allowing 5 full years of follow-up across all calendar groups.

RA was defined as a first-time registration of a M05 and M06 (except M06.1) ICD-10 code and a redeemed prescription of a conventional synthetic disease-modifying anti-rheumatic drug in the following year. Patients with incident RA were matched 1:5 on year of birth and sex with non-RA individuals from the general population by exposure density sampling. Patients and controls were considered at risk from the diagnosis/matching index date to the date of 5 years follow-up, death, emigration, or, for matched controls, to the date of a RA diagnosis, whichever occurred first. Matched controls from the period between 1996 to 2000 were used as reference group.

Pseudo-observations for the cumulative incidence proportion (CIP) were calculated assuming independent censoring. Generalized linear regression with robust variance estimator was used to estimate the risk difference (RD) and relative risk (RR) for age-adjusted all-cause mortality for RA patients compared with controls. Restricted cubic splines with 4 knots were used to adjust for age with median age as reference.

Results: In total, 29 163 patients with incident RA and 145 792 matched controls were identified. The 5-year CIP of death for a 60-year-old RA patient decreased from 8.2 % in 1996 to 2000 to 3.1 % in 2011 to 2015, while the decrease seen in matched controls was from 4.9 % to 2.1 % (Table 1). The age-adjusted all-cause CIP of death was higher for patients with RA compared to matched controls in all calendar periods, with RDs ranging from 3.4 % in 1996 to 2000 to -1.7 % in 2011 to 2015 when compared to the reference group. In the same period the RD for matched controls decreased with -2.8 %. Similar, the RR decreased from 1.3 in 1996 to 2000 to 0.9 in 2011 to 2015 for patients with RA, and to 0.8 for matched controls in 2011 to 2015 when compared to the reference group. Sex-specific analyses revealed similar temporal and RD patterns in men and women.

Conclusion: Five-year all-cause mortality risk decreased over time for both patients with RA and matched controls indicating temporal improvements in mortality related factors in both groups. In line with the established excess mortality in RA patients, a higher percentage of patients with RA died compared to matched controls across all calendar periods. The differences between the groups attenuated with time. This probably reflects the improved treatment regimens for RA, increased focus on treating comorbidities and lifestyle factors in RA patients.

Disclosures: B. Soussi, Novo Nordisk; K. Duch, None; R. Cordtz, None; C. Bork, None; S. Kristensen, None; E. Schmidt, None; J. Lindhardsen, None; L. Dreyer, Bristol-Myers Squibb(BMS).