How sensitive and specific is urinalysis 'dipstick' testing for detection of hyperglycaemia and ketosis? An audit of findings from coronial autopsies

Abstract

<h4>Aims</h4>Biochemical analysis of glucose and ketones in the vitreous humour obtained at post-mortem examination is representative of the levels in the blood prior to death. Elevated levels can be indicative of conditions including diabetic ketoacidosis, which can be a cause for unexpected death. A rapid screening test for such conditions can be performed during the autopsy through urinalysis using test strips (urine 'dipstick' testing). The aim of this study was to assess the utility of urinalysis testing for post-mortem detection of derangements of glucose and ketone levels.<h4>Methods</h4>The results of vitreous biochemical analysis and urinalysis were collated from 188 forensic autopsy cases.<h4>Results</h4>A vitreous glucose result of above 10 mmol/L was regarded as high. When this was compared to urinalysis results it was found that any urinalysis result above negative had a sensitivity of 0.83 and a specificity of 0.93. A vitreous ketone level of above 5 mmol/L was regarded as significantly elevated; a urinalysis result above negative had a sensitivity of 1, but a specificity of 0.12.<h4>Conclusions</h4>Urinalysis ('dipstick' testing) for glucose has a good sensitivity and specificity for high vitreous glucose levels, which are regarded as indicative of pathological hyperglycaemia during life. It was found that urine testing for ketones either has an excellent sensitivity with low specificity or a poor sensitivity with a good specificity; however, this finding has to be viewed in the context of uncertainty of the biochemical level of significant ketosis.