Abstract
Objective To assess trends in incidence of heart failure (HF) in patients with incident rheumatoid arthritis (RA) in 1980-2009 and to compare different HF definitions in RA.
Methods The study population comprised Olmsted County, Minnesota residents with incident RA (age ≥ 18 years, 1987 ACR criteria met in 1980-2009). All subjects were followed until death, migration, or 04/30/2019. Incident HF events were defined as: 1) Framingham criteria for HF; 2) Diagnosis of HF (outpatient or inpatient) by a physician; 3) ICD-9/10 codes for HF. Patients with HF prior to RA incidence/index date were excluded. Cox proportional hazards models were used to compare incident HF events by decade, adjusting for age, sex and cardiovascular risk factors. HF definitions 2 and 3 were compared to the Framingham criteria.
Results The study included 905 patients with RA (mean age 55.9 years; 68.6% female; median follow-up 13.4 years). The 10-year cumulative incidence of HF event by any chartreviewed method in RA cohort in the 1980s was 11.66% (95%CI 7.86-17.29%), 1990s was 12.64% (95%CI 9.31-17.17%), and 2000s was 7.67% (95%CI 5.36-10.97%). Incidence of HF did not change across the decades of RA incidence using any of the HF definitions. Physician diagnosis of HF and ICD-9/10 code-based definitions of HF performed well compared to Framingham criteria, showing moderate-to-high sensitivity and specificity.
Conclusion Incidence of HF in patients with incident RA in 2000s versus 1980s was not statistically significantly different. Physician diagnosis of HF and ICD-9/10 codes for HF performed well against Framingham criteria.
