Background:

Crilical hyperkalemia is a serious life‐threalening situation. However, we often encounter fictitious laboratory hyperkalemia due to specimen contamination or hemolysis during phlebotomy. This study determined the frequency of true critical hyperkalemia (potassium > 6.5 mEq/L) among hospitalized patients in an urban tertiary‐care academic center.

Methods:

This was a retrospective analysis of an administrative database of adult inpatient potassium (K) values from January 1, 2008, to November 30, 2009, Using a customized electronic query, questionable or unreliable specimens (i.e., classified by the laboratory as “contaminated,” “quantity not sufficient,” “clotted,” “hemolyzed,” etc.) were excluded. All 918 will be reviewed. The interim analysis of the 319 highest values are presented. The top critical values were obtained and further screened through electronic laboratory data and chart review. The criteria for true critical hyperkalemia were (1) the value was repeated and verified and (2) the clinician acted on the abnormal value.

Results:

A total of 670,094 inpatient potassium values were included, and 918 instances (0.14%) of critical hyperkalemia were identified. Only 15.9% (51) of the top 319 values fulfilled the criteria for true critical hyperkalemia (mean 7.8 mEq/L, range 7.4–9.8 mEq/L], indicating that the total incidence was 0.02% of all K measurements. Almosl half (44%) of the fictitious values were venous panel specimens obtained from the emergency department, Fifty‐one percent of all true critical hyperkalemic events were associated with chronic kidney disease, 49% with acute kidney injury, 13.7% with heart failure, and 9.8% with cancer. Thirty‐seven percent of these patients were taking angiolensin‐converting enzyme inhibitors, whereas eleclrocardiogram abnormalities were noted in 43% of these palients.

Conclusions:

Though critical hyperkalemia among hospitalized patient is rare, up to 15.9% of critical values represent a true medical emergency.

Author Disclosure:

A. Briones, none; M. Bundy, none; R. Jervis, none.