Short-term prognosis of normalising serum potassium following an episode of hypokalaemia in patients with chronic heart failure

Abstract

It is well known that patients with chronic heart failure and hypokalaemia have increased mortality risk. We investigated the impact of normalising serum potassium following an episode of hypokalaemia on short-term mortality among patients with chronic heart failure. We identified 1673 patients diagnosed with chronic heart failure who had a serum potassium measurement under 3.5 mmol/l within 14 days and one year after initiated medical treatment with both loop diuretics and angiotensin-converting enzyme inhibitors or angiotensin-II receptor blockers. A second serum potassium measurement was required 8-30 days after the episode of hypokalaemia. All-cause mortality and cardiovascular mortality was examined within 90 days from the second serum potassium measurement. Mortality was examined according to six predefined potassium groups derived from the second measurement:<3.5 mmol/l (n = 302), 3.5-3.7 mmol/l (n = 271), 3.8-4.1 mmol/l (n = 464), 4.2-4.4 mmol/l (n = 270), 4.5-5.0 mmol/l (n = 272), and 5.1-8.0 mmol/l (n = 94). We used Cox regression to estimate both all-cause mortality risk and cardiovascular mortality, with serum potassium at 3.8-4.1 mmol/l as reference. After 90 days, the all-cause mortality in the six groups was 29.5%, 22.1%, 20.3%, 24.8%, 23.5% and 43.6%, respectively. In multivariable adjusted analysis, patients with serum potassium <3.5 mmol/l (hazard ratio: 1.51; 95% confidence interval: 1.13-2.02) and serum potassium 5.1-8.0 mmol/l (hazard ratio: 2.18; 95% confidence interval: 1.50-3.17) had an increased risk of all-cause mortality compared to the reference. After 90 days, the cardiovascular mortality in the six groups was 19.2%, 17.7%, 14.4%, 18.9%, 18.8% and 34.0%, respectively. In multivariable adjusted analysis, patients with serum potassium 5.1-8.0 mmol/l (hazard ratio: 2.32; 95% confidence interval: 1.51-3.56) had an increased risk of cardiovascular mortality compared to the reference, while serum potassium <3.5 mmol/l (hazard ratio: 1.37; 95% confidence interval: 0.97-1.95) had a trend toward increased risk of cardiovascular mortality compared to the reference. Patients with chronic heart failure and hypokalaemia, who after 8-30 days remained hypokalaemic, had a significantly higher 90-day all-cause mortality risk compared to patients in the reference group (3.8-4.1 mmol/l). Patients with chronic heart failure and hypokalaemia, who after 8-30 days had the serum potassium level increased to a level within 5.1-8.0 mmol/l, had both a significantly higher 90-day all-cause mortality risk and cardiovascular mortality risk compared to patients in the reference group (3.8-4.1 mmol/l).