Simple methods for estimating the maximum 24-hour urinary potassium excretion in kidney failure without replacement...

Area of Science:

• Nephrology
• Clinical Biochemistry
• Medical Informatics

Background:

• Dietary potassium adjustment based on 24-hour urinary potassium excretion is crucial for preventing hyperkalemia.
• Accurate and convenient methods for calculating maximum 24-hour urinary potassium excretion are lacking in patients with kidney failure not on replacement therapy.
• This study addresses the need for reliable assessment tools in this high-risk population.

Purpose of the Study:

• To develop and validate novel predictive models for maximum 24-hour urinary potassium excretion.
• To provide a tool for estimating the upper limit of dietary potassium consumption in patients with kidney failure.
• To improve the management of hyperkalemia in this specific patient group.

Main Methods:

• Data from 145 patients with kidney failure and hyperkalemia were analyzed.
• Two predictive models were developed using multilayer perceptron and stepwise multiple linear regression on a training set of 102 patients.
• Model performance was validated on a separate set of 43 patients.

Main Results:

• The developed models demonstrated significantly lower bias and mean absolute error compared to existing methods.
• High accuracy (83.45% and 84.14%) and strong correlation (Spearman 0.72, ICC 0.67-0.70) with measured 24-hour urine potassium were observed.
• Excellent agreement was found, with narrow 95% limits of agreement (13.20-13.70 mmol/24h).

Conclusions:

• The newly developed models offer high clinical utility for calculating maximum 24-hour urinary potassium excretion.
• These models can aid in safely adjusting dietary potassium intake for patients with kidney failure and hyperkalemia.
• The findings support the application of these models for better hyperkalemia management in this cohort.