An audit of the contribution to post-mortem examination diagnosis of individual analyte results obtained from biochemical analysis of the vitreous

Abstract

Biochemical analysis of the vitreous humor from the eye is an accepted accessory test for post-mortem investigation of cause of death. Modern biochemical analyzers allow testing of a range of analytes from a sample. However, it is not clear which analytes should be requested in order to prevent unnecessary testing (and expense). The means and standard deviation of the values obtained from analysis of the vitreous humor for sodium, potassium, chloride, osmolality, glucose, ketones (β-hydroxybutyrate), creatinine, urea, calcium, lactate, and ammonia were calculated from which the contribution of each analyte was reviewed in the context of post-mortem findings and final cause of death. For sodium 32 cases were regarded as high (more than one standard deviation above the mean), from which 9 contributed to post-mortem diagnosis [drowning (4), heat related death (2), diabetic hyperglycemia (2), and dehydration (1)], but 25 low values (greater than one standard deviation below the mean) made no contribution. For chloride 29 high values contributed to 4 cases--3 drowning and 1 heat-related, but these were all previously identified by a high sodium level. There were 29 high and 35 low potassium values, none of which contributed to determining the final cause of death. Of 22 high values of creatinine, 12 contributed to a diagnosis of renal failure. From 32 high values of urea, 18 contributed to 16 cases of renal failure (2 associated with diabetic hyperglycemia), 1 heat-related death, and one case with dehydration. Osmolarity contributed to 12 cases (5 heat-related, 4 diabetes, 2 renal failure, and 1 dehydration) from 36 high values. There was no contribution from 32 high values and 19 low values of calcium and there was no contribution from 4 high and 2 low values of ammonia. There were 11 high values of glucose, which contributed to the diagnosis of 6 cases of diabetic hyperglycemia and 21 high ketone levels contributed to 8 cases: 4 diabetic ketosis, 3 hypothermia, 3 ketosis of unknown cause, and 2 alcohol related deaths. A high lactate was identified in 25 cases, which contributed to 1 case with a diagnosis of metformin toxicity (1), but none of the 22 low lactate values contributed. The results of this audit have been used to reduce vitreous biochemistry test requests for sodium, osmolality, glucose, ketones, urea, and creatinine in most cases. Critical appraisal of each part of the post-mortem process should be undertaken to provide evidence to justify any investigative methods used in an autopsy.