[Association between serum potassium levels and all-cause mortality in patients with acute heart failure]

Objective: To investigate the relationship between different serum potassium levels at admission and discharge and all-cause mortality in patients with acute heart failure (HF). Methods: A total of 2 621 patients with acute HF who were hospitalized in the Heart Failure Center of Fuwai Hospital from October 2008 to October 2017 were analyzed. Patients were divided into three groups according to the different serum potassium levels at admission: hypokalemia with serum potassium<3.5 mmol/L (n=329), normokalemia with 3.5-5.5 mmol/L (n=2 270), and hyperkalemia with serum potassium>5.5 mmol/L (n=22). Clinical data such as patient history, comorbidities, clinical examination and drug use were collected, and systematic outpatient review or telephone follow-up was performed after patients were discharged from the hospital until January 2020. The primary outcome was all-cause death at 90 days, 2 years, and 5 years of follow-up. We compared the clinical characteristics of patients with different serum potassium levels at admission and discharge, and used a multivariate Cox proportional hazards regression model to analyze the association between serum potassium levels at admission and discharge and all-cause mortality. Results: The age of all patients was (58.0±15.3) years old, and 1 877 patients (71.6%) were male. There were 329 (12.6%) and 22 (0.8%) patients with hypokalemia and hyperkalemia at admission, and 38 (1.4%) and 18 (0.7%) at discharge, respectively. The serum potassium levels of all patients were (4.01±0.50) and (4.25±0.44) mmol/L at admission and discharge, respectively. The follow-up time[M(Q₁,Q₃)] of this study was 2.63(1.00,4.42)years, and a total of 1 076 all-cause deaths were recorded at the last follow-up. Compared with patients with normokalemia at discharge, discharged patients with hypokalemia and hyperkalemia were followed up for 90 days (90.3% vs 76.3% vs 38.9%), 2 years (73.8% vs 60.5% vs 33.3%) and 5 years (63.4% vs 44.7% vs 22.2%), respectively, and the difference of which in cumulative survival rates were statistically significant (all P values<0.001). The multivariate-adjusted Cox regression analysis showed that hypokalemia (HR=0.979, 95%CI: 0.812-1.179, P=0.820) and hyperkalemia (HR=1.368, 95%CI: 0.805-2.325, P=0.247) at admission were not associated with all-cause mortality risk, however, hypokalemia (HR=1.668, 95%CI: 1.081-2.574, P=0.021) and hyperkalemia (HR=3.787, 95%CI: 2.264-6.336, P<0.001) at discharge were associated with increased all-cause mortality risk. Conclusions: Both hypokalemia and hyperkalemia at discharge in hospitalized patients with acute HF were associated with increased short-and long-term all-cause mortality, and serum potassium levels should be closely monitored.