Letter to the Editor: The effects of hyperlipidaemia, hyperbilirubinaemia and haemolysis on tests performed by the Olympus AU 5000 multiple analyser

Increasing amounts ofIntralipid (Kabivitrum), bilirubin (NBS standard 916) or haemolysate were added to a serum pool of 40 patients, as previously described [1]. Added Intralipid yielded samples with triglyceride levels of 0"8-6-3 g/l, added bilirubin yielded samples having total bilirubin levels of 4-325 mg/1 and added haemolysate yielded samples with haemoglobin concentrations of 0-1"58 g/1. All analyses were performed in duplicate using procedures provided by the manufacturer (Olympus Corp., Success, MY, USA).

We systematically investigated the effects of hyperlipidaemia, hyperbilirubinaemia and haemolysis on 20 tests performed by the Olympus AU 5000. This brief report should be useful for users of this new analyser or those using similar analytical approaches.
Added Intralipid yielded samples with triglyceride levels of 0"8-6-3 g/l, added bilirubin yielded samples having total bilirubin levels of 4-325 mg/1 and added haemolysate yielded samples with haemoglobin concentrations of 0-1"58 g/1. All analyses were performed in duplicate using procedures provided by the manufacturer (Olympus Corp., Success, MY, USA).
The amount ofeach interfering substance causing at least a 10% negative or positive interference is indicated in table 1. In all instances where interference was observed, higher concentrations of the interferent resulted in larger interferences. For instance, the effects of bilirubin on serum creatinine at bilirubin concentrations of 130 and 190 mg/1 produced negative interferences of 28% and 38%, respectively. Intralipid is intended to simulate the light scattering caused by lipoproteins. However, the micelles of Intralipid are more homogeneous than those found in vivo and this approach may not always produce results comparable to those with patients' samples [1]. Nevertheless, the lipaemic interference could in every instance be eliminated by ultracentrifugation. As the Table 1 " Concentration ofinterferring substance which causes a 10% negative (-) or positive (+) interference.
.D ilution of Intralipid to a ratio of 1" 10 and yielding measured triglyceride concentrations of 6"3 g/l.
Interferences not reported in the manufacturer's literature. ND not determined. bilirubin added to the serum pool was unconjugated and as endogenous bilirubin may include conjugated bilirubin, biliverdin and other metabolites, our in vitro study may not totally reflect the analytical results obtained in vivo. Because visible haemolysis does not occur until the haemoglobin concentration is approximately 0"2 g/l, the effect of haemolysis on some analytes with interference from haemolysis at this concentration could be missed.
For this reason we have already changed our total bilirubin method for the AU 5000 [2]. Seven of the interferences that we detected are not described in the manufacturer's literature, emphasizing the importance of independently evaluating potential interferences with any new procedure and/or analyser. applications in the areas of clinical chemistry, pollution monitoring and also process control. All of these make interesting reading, but they are subject areas in themselves which require a book in their own rights. No attempt has been made to show how the approaches described in these chapters can relate to the analytical chemistry approach, which is a pity.
The one criticism that I have of the book is that the important areas of specification, economics and the effects on management, are not covered at all in this book. It is true that they are touched upon briefly, but they are vital to successful automation and they need to be reviewed.
These areas no doubt are outside the experience of the authors, but are significant. None the less, the book is a collossal achievement and it should find a good sales potential in academic and industrial establishments.
I congratulate the authors most warmly and will enjoy keeping their book as a ready reference source for many years to come. 9O