The value of c-reactive protein to albumin ratio in predicting long term mortality among hfref patients with implantable cardiac defibrillators

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Patients with heart failure with reduced ejection fraction (HFrEF) who received implantable cardiac defibrillator (ICD) still remain at high risk due to pump failure and comorbid conditions. C-reactive protein to albumin ratio (CAR) may be of value for identifying those with high risk for mortality despite ICD implantation. Methods Those who were implanted ICD for HFrEF in our institution between 2009 and 2019 were included. CAR was calculated as ratio of C-reactive protein (CRP) to serum albumin concentration. Patients were grouped into tertiles in accordance to CAR at the time of implantation. After follow up of 48 ± 35 months, survival times of tertiles were compared by using Kaplan-Meier survival method. Results Thousand and eleven patients constituted study population. Ischemic cardiomyopathy was primary diagnosis in 92.3%. Of those 14.5% had died after discharge. Patients in tertile 3 (T3) had higher risk of appropriate shock (19.3% vs 23.7% vs 38.0%) and mortality (4.2% vs 11.0% vs 28.5%) compared to those in other tertiles. Multivariable analysis revealed that when patients in T1 were considered as reference, both those in T2 and T3 had independently higher risk of appropriate shocks and mortality. These effects were consistent in the unadjusted and adjusted multivariable models. Conclusion Among patients with HFrEF and ICD, elevated CAR increased the risk of appropriate device shock and mortality at long term. Table 1Admission C-reactive protein/Albumin ratio (n = 1011)T1 (n = 337)T2 (n = 337)T3 (n = 337)Mortality, %4.211.028.5Mortality, HR (95% CI)Model 1: unadjusted1[Reference]3.85 (2.12 - 11.20)8.14 (2.46 - 28.56)Model 2: adjusted for age, sex1[Reference]3.20 (1.90 - 9.48)6.32 (2.12 - 20.12)Model 3: adjusted for comorbiditesa1[Reference]4.85 (2.06 - 14.12)10.86 (4.12 - 44.82)Model 4: adjusted for covariatesb1[Reference]2.72 (1.66 - 7.12)5.72 (2.04 - 18.05)Frequency, %19.323.738.0Appropriate shock, HR (95% CI)Model 1: unadjusted1[Reference]1.38 (0.44 - 6.88)2.48 (1.34 - 5.82)Model 2: adjusted for age, sex1[Reference]1.42 (0.48 - 7.24)3.02 (1.52 - 6.24)Model 3: adjusted for comorbiditesa1[Reference]1.34 (0.38 - 6.66)2.74 (1.40 - 7.28)Model 4: adjusted for covariatesb1[Reference]1.30 (0.34 - 5.68)2.28 (1.16 - 8.28)Cox proportional analysis and logistic regression models for the appropriate shock and the long-term mortality by CAR.Abstract Figure 1