The regression analysis was corrected for sex, age, severity of illness, AKI, mean potassium, mean potassium squared and potassium variability.
Data are adjusted for sex, age, acute kidney injury (AKI), severity of illness (Acute Physiology and Chronic Health Evaluation II (APACHE II) score), mean potassium, mean potassium squared and potassium variability observed between 24 hours and 7 days after admission.
Moreover, to study the relationship between the combined mean serum glucose and mean serum potassium concentrations on ICU mortality, patients were divided into eight equal subgroups for their mean serum glucose and eight equal subgroups for their mean serum potassium values.
The average serum potassium concentration, as well as average levels of other relevant covariates was calculated as the mean of the first month and each consecutive third month potassium value until the occurrence of the renal event.
In response to the potassium-deficient meal plus 35 mmol potassium, the serum potassium did not increase but the hourly mean potassium excretion increased sharply.
In addition, our data on the relationship between the mean potassium level during follow-up, which reflects the exposure to a high serum potassium load during a definitive period of time more accurately than a single value, and renal outcomes displayed a similar association between increased serum potassium and adverse renal outcomes.
Minimum, maximum and mean potassium levels, as well as potassium variability, were determined for every patient.
The potassium metrics that we determined included mean potassium, potassium variability (defined as the standard deviation of all potassium levels) and percentage of ICU time below and above the reference range (3.5 through 5.0 mmol/L).
In order to test for a U-shaped relationship of mean potassium with hospital mortality, the mean potassium concentration was both included directly and squared.
For the different algorithms we also contrasted the mean, median and IQRs of serum potassium values for those who were positive for the code compared with patients with hospital encounters who had no evidence of the code (ie, code negative).
