# Comparison of Blood Gas Analysis and Auto-Analyzer Results for Sodium and Potassium Levels in Elderly and Non-elderly Adult Emergency Department Patients

## TL;DR

This study compares blood gas and auto-analyzer results for sodium and potassium in elderly and non-elderly emergency department patients, finding greater variability in the elderly group.

## Contribution

The study reveals that diagnostic precision for electrolyte levels may need adjustment for elderly patients due to greater measurement discrepancies.

## Key findings

- Elderly patients showed a larger bias and wider limits of agreement for sodium measurements compared to non-elderly patients.
- Potassium measurements in elderly patients also had greater discrepancies and lower concordance correlation coefficients.
- The study highlights the need for tailored diagnostic approaches for elderly emergency department patients.

## Abstract

Objectives

This study aims to evaluate the concordance between blood gas and biochemical measurement methods for sodium and potassium levels in elderly and non-elderly patients within an emergency department (ED) setting.

Methods

A retrospective method comparison study was conducted at an ED from February 1, 2023, to March 1, 2023. The study included 414 patients, categorized into "elderly" (aged 65 and above; n = 138, 33.3%) and "non-elderly" (aged 18 to 64; n = 276, 66.7%) groups. Concordance was assessed using Bland-Altman, Passing-Bablok, and Lin's concordance correlation methods.

Results

In sodium measurements, the elderly group exhibited an average bias of −1.52 mEq/L (95% confidence interval [CI] −2.12 to −0.92), with lower and upper limits of agreement (LoA) at −8.46 and 5.42 mEq/L, respectively, indicating a broader variance than non-elderly patients, who showed an average bias of −0.82 mEq/L with limits of −4.97 to 3.32 mEq/L. For potassium, the elderly group's average bias was −0.46 mEq/L (95% CI −0.36 to −0.57), with limits of agreement from −1.68 to 0.75 mEq/L, compared to non-elderly patients with a bias of −0.29 mEq/L and limits of −0.71 to 0.13 mEq/L. Furthermore, concordance correlation coefficients revealed a reduced agreement in the elderly for both sodium (rccc = 0.799) and potassium (rccc = 0.529) compared to the non-elderly cohort (sodium rccc = 0.821, potassium rccc = 0.715).

Conclusion

The study identifies significant discrepancies in sodium and potassium levels between elderly and non-elderly patients, suggesting a need for diagnostic precision. It emphasizes the importance of customizing diagnostic approaches to better serve the elderly population in EDs.