Some issues to consider with the use of serum indices

Haemolysis is a preanalytical interference commonly encountered in specimens drawn for laboratory testing.1 2 As haemoglobin moieties absorb light between 300 and 600 nm, haemolysis can cause direct spectrophotometric interference in assays which absorb at similar wavelengths (eg, iron, lipase). Haemolysis itself can increase the concentration of intracellular analytes in serum or plasma (eg, potassium, lactate dehydrogenase and asparate aminotransferase) and conversely dilute extracellular molecules (eg, glucose and sodium).1 Therefore, reliable detection of haemolysis is an important prerequisite to the production of reliable laboratory test results. In some instances, knowing the degree of specimen haemolysis can help rationalise aberrant or unexpected test results. The use of automated, preanalytical processes for the detection of haemolysis (H-index), icteric (I-index) and lipaemia (L-index) in patient specimens has superseded their visual inspection.3 4 This is largely due to advances in instrumentation driven by increased demands on our medical laboratory services. In a high throughput laboratory, it is impossible to physically examine every specimen received. Aside from this, visual inspection of specimens is considered arbitrary at best.3 4 In this case, this New York-based laboratory uses two operators to examine specimens when the automated H-index is elevated.5 Here, the ‘high’ H index is set at >90, corresponding, on Roche instrumentation, to a haemoglobin concentration of 0.9 g/L (56 µmol/L). With a Roche H-index measuring range of 5–1200 mg/dL (~3–745 µmol/L), this cut-off of 90 mg/dL could be perceived as relatively low. The lower this index is set the more frequently ‘haemolysed’ samples will be encountered and the more laboratorian time is required for manual specimen inspection. Ideally, such interferents would not be present in patient serum/plasma, but alas they are. …