The development and validation of an ABL835 FLEX Blood Gas Analyser method for the quantitation of glucose, lactate, and electrolytes in post-mortem vitreous humour

Van Vuuren, Andrew

The development and validation of an ABL835 FLEX Blood Gas Analyser method for the quantitation of glucose, lactate, and electrolytes in post-mortem vitreous humour

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Department of Pathology

Faculty

Faculty of Health Sciences

Abstract

Post-mortem biochemistry provides valuable insights into the biochemical status of an individual at the time of death. Vitreous humour (VH) has an anatomically isolated location within the body, making it less prone to the effects of putrefaction and autolysis after death. As such, VH has frequently been described as the specimen of choice in biochemical forensic investigations to identify potential ante-mortem conditions and to estimate the post-mortem interval. The lack of a validated method for post-mortem biochemistry analysis using VH as part of routine toxicology testing within the Western Cape province of South Africa motivated the primary aim of this study, which was to develop and validate a method for the analysis of glucose, lactate, and electrolytes (potassium (K+), sodium (Na+), chloride (Cl-) and calcium (Ca2+)) using an ABL835 FLEX blood gas analyser. Method development included instrument verification and sample preparation optimisation, which assessed the effect of centrifugation and dilution. Validation included tests for accuracy, precision, sample stability, drug interferences and between-eye differences. Sample preparation experiments determined that untreated VH may be analysed directly or diluted in a 1:1 ratio with ultrapure water, when required. All analytes demonstrated acceptable bias, within-run and between-run precision at concentrations above 1.6 mmol/L. Results indicated suitable stability for K+, Na+, Cl-, glucose and lactate across all storage conditions over 31 days, while Ca2+ values decreased over time when frozen at -20°C. Drug interference was minimal for most analytes, with effects remaining within acceptable ranges (%error < ± 20%), with the exception of Ca2+. No statistically significant between-eye differences (p > 0.05) were observed for any analytes in both undiluted and diluted (1:1) samples. Overall, the method performed satisfactorily and was deemed fit for purpose for routine analysis. While important limitations regarding sample stability and potential interferences must be considered, this method provides a valuable resource for post-mortem biochemical analysis in cases of unnatural deaths in the Western Cape to support both individual case resolutions and broader public health insights.