A simple heart failure detection score for rheumatology practice: findings from the KURAMA cohort
Kosaku Murakami, Hideaki Inazumi, Akira Onishi, Shuichiro Nakabo, Takayuki Fujii, Koichi Murata, Masao Tanaka, Koh Ono, Akio Morinobu, Eri KatoAbstract
Objectives
Cardiovascular disease (CVD) accounts for approximately 40% of deaths among patients with rheumatoid arthritis (RA), yet the burden of heart failure (HF) within this population remains poorly characterized. Using the KURAMA cohort, we aimed to quantify the prevalence of HF among RA outpatients and develop a practical HF screening tool that uses variables readily available to rheumatologists in routine clinical practice.
Methods
A cross-sectional study of 542 outpatients with RA was conducted. Their HF status was determined using a prespecified multistep algorithm that integrated clinical history, loop diuretic use, N-terminal pro-B-type natriuretic peptide (NT-proBNP), echocardiography, and careful differentiation from interstitial lung disease (ILD). Adaptive LASSO regression was applied to identify independent factors associated with HF and construct a detection score using non-cardiac variables.
Results
Heart failure was detected in 26.5% of patients with RA. Older age, lower haemoglobin (Hb) levels, higher serum creatinine (CRE) levels, and higher Simplified Disease Activity Index (SDAI) were identified as independent non-cardiac factors associated with HF. A 4-factor HF detection score was constructed based on adjusted odds ratios. The area under the receiver operating characteristic curve (AUC) for the discrete score based on variables routinely available in rheumatology clinics was 0.793, comparable to that of NT-proBNP ≥ 125 pg/mL alone (AUC = 0.814).
Conclusion
HF affects over one in four RA outpatients. A simple 4-factor score may serve as a practical first-line triage tool to identify patients who warrant further cardiac evaluation and cardiology referral.