Point-of-care β-hydroxybutyrate and glucose as candidate screening methods for ketoacidosis-associated death in forensic autopsy investigations.

OBJECTIVES: Ketone bodies, like β-hydroxybutyrate (BHB), derived from fatty acid breakdown, can cause fatal ketoacidosis if levels are excessively high. Postmortem diagnosis of ketoacidosis is challenging due to non-specificity of rapid chromogenic tests and the time required for LC-MS/MS analysis. This study investigates the feasibility of using point-of-care (POC) BHB and glucose testing to diagnose ketoacidosis-related deaths and distinguish between diabetic and other types of ketoacidosis, post-mortem.

DESIGN AND METHODS: This study evaluated the Nova StatStrip meter's analytical and post-mortem performance for measuring BHB and glucose in decedent whole blood and vitreous humor. Precision, linearity, and recovery were assessed. BHB and glucose were measured in whole blood and vitreous humor from 100 autopsy cases (both ketoacidosis and non-ketoacidosis deaths). Results were compared quantitatively and qualitatively with standard laboratory methods to determine the meter's accuracy and reliability in predicting ketoacidosis-related deaths. Receiver operating characteristic (ROC) curves were constructed for each matrix/analyte combination to evaluate screening capabilities for ketoacidosis- and diabetic ketoacidosis-related deaths.

RESULTS: Imprecision was highest in decedent vitreous humor samples for both BHB and glucose and both assays exhibited acceptable linearity. ROC curve analysis indicated comparable post-mortem performance between methods and matrices. Whole blood BHB showed the best performance for predicting all-cause ketoacidosis on the meter, despite exhibiting a higher device error messages than vitreous humor. Glucose in vitreous humor exhibited the most optimal performance (100 % sensitivity and specificity) but showed the highest rate of error messages (64 %). Thus, whole blood glucose (80 % sensitivity, 98 % specificity) would be the preferred matrix to identify potential DKA-related deaths. Agreement between meter and laboratory methods was excellent despite differing thresholds (96-100 %).

CONCLUSIONS: This study suggests that post-mortem BHB levels in whole blood and vitreous humor can be conveniently obtained using the StatStrip meter and have comparable performance to current standards, making it suitable as a screening tool for ketoacidosis-related death. Screening thresholds recommended are vitreous humor (≥ 0.9 mmol/L) or whole blood (≥ 1.6 mmol/L) BHB for ketoacidosis-related death, followed by whole blood glucose (≥ 27.3 mmol/L) for DKA-related death.