Evaluation of published equations for calculating serum osmolality in a South African tertiary hospital patient population.
Ebonwu, Okechukwu Emmanuel
Ebonwu, Okechukwu Emmanuel
Citations
Altmetric:
Abstract
Background:
Many studies have assessed the predictive accuracy of serum osmolality equations in various settings. Thirty published equations were evaluated using patient
data from a regional hospital laboratory.
Materials and methods:
Laboratory records were extracted, with same-sample results for measured serum osmolality, sodium, potassium, urea and glucose analysed National
Health Laboratory Service at Steve Biko Academic Hospital, Pretoria, South Africa, between 1/1/2017–31/12/2018. Calculated osmolality was compared to measured serum osmolality using Passing-Bablok regression and Bland-Altman analysis. Equations with osmolal gaps closest to zero, bias of ≤0.7% and osmolal gap reference limits within ±10 mOsm/kg were identified.
Results:
Majority of equations produced proportionally negative-biased results. Two equations (EQ6, EQ19) were identified with bias of ≤0.7% but unworkable osmolal gap
reference limits. Of the two equations, EQ19 performed near-optimally over a range of 200 - 400 mOsm/kg and required no further adaptation. Several equations produced acceptable bias and osmolal gap reference limits within ±10 mOsm/kg after bias adjustment (EQ13, EQ21, EQ27, and EQ29).
Conclusion:
Few published equations are immediately usable. Validating an osmolality equation before routine use is highly recommended. Bias adjustment generated several usable equations. Future research should investigate the feasibility of harmonising predicted osmolality and osmolal gap reporting between South African laboratories
Description
Submitted in partial fulfilment of the requirements for the degree, Master Technologiae: Biomedical Technology in the Department of Biomedical Sciences in the Faculty of Science at the Tshwane University of Technology.
Date
2020-04-01
Journal Title
Journal ISSN
Volume Title
Publisher
Tshwane University of Technology
Research Projects
Organizational Units
Journal Issue
Keywords
Osmolality, Serum, Equations, National Health Laboratory Service, Steve Biko Academic Hospital, Pretoria, South Africa