Dyskalemias in patients with acute kidney injury presenting to the emergency department are common and independent predictors of adverse outcome

Link to article at PubMed

Int J Clin Pract. 2020 Aug 8:e13653. doi: 10.1111/ijcp.13653. Online ahead of print.

ABSTRACT

BACKGROUND: No data concerning the prevalence and risk factors of dyskalemia in acute kidney injury (AKI) exists. We investigated A.) prevalence rates, B.) risk factors and C.) outcome of hypo- and hyperkalemia in emergency patients.

METHODS: In this cross-sectional analysis, all patients admitted to the emergency department of a large public hospital in Switzerland between January 1st 2017 and December 31st 2018 with measurements of creatinine and potassium were included. Baseline characteristics, medication and laboratory data were extracted. Chart reviews were performed to identify patients with a diagnosis of chronic kidney disease (CKD) and to extract their baseline creatinine. For all other patients, the ADQI backformula was used in order to calculate baseline creatinine. AKI was graduated using creatinine criteria of the acute kidney injury network. Binary logistic regression analysis was used to identify risk factors for appearance of hyperkalemia and outcome.

RESULTS: AKI was found in 8% of patients. Hyperkalemia was present in 13% and hypokalemia in 11% of patients with AKI. AKI stage, potassium-sparing diuretics, ACE-inhibitors and underlying CKD were the strongest risk factors for hyperkalemia. Hyperkalemia as well as profound hypokalemia were independently associated with prolonged length of stay and in-hospital mortality. The study is limited by its dependency on chart review data in order to identify patients with chronic kidney disease and by limitations of the ADQI backformula to calculate baseline creatinine.

CONCLUSIONS: Dyskalemias are common in emergency patients with AKI and are independent risk factors for adverse outcomes. Potassium-sparing diuretics, ACE-inhibitors, AKIN stage and CKD are predictors of hyperkalemia in AKI.

PMID:32770846 | DOI:10.1111/ijcp.13653

Leave a Reply

Your email address will not be published.