Deep Lancing or Overmanipulation?

In this issue of the Journal of Diabetes Science and Technology, Heenan et al ¹ compare 41 adults’ average capillary blood glucose meter (BGM) glucoses to their single central laboratory glucose (CLG) measurements of 0.6 mL lithium heparin plasma separator microtainers filled after “deep lancing” of the finger capillaries. The authors’ figure illustrates the relationship between the central laboratory instrument measurement of hemolysis (hemolysis index) and the ratios of the two sets of measurements. As hemolysis increases, it seems that the CLG increases slightly, but so does the variation in results. It is disappointing that the authors did not attempt to document hemolysis in the “shallow-lanced” specimens used to obtain the BGM measurements. Recently, three different groups of adult subjects had their capillary blood collected into plain microhematocrit capillary blood collection tubes after lancing with a commercially available lancet device. ² While the same trained phlebotomists employed the same technique to obtain the capillary samples, the visible hemolysis rates of the spun specimens varied from 0% (59 subjects) to 7% (42 subjects) to 43% (40 subjects). These experiments seem to complicate Heenan et al’s tacit assumption that “shallow-lancing” does not result in artifactual hemolysis.

Many questions arise from Heenan et al’s study. Why did many of the 0.6 ml lithium heparin plasma separator microtainer tubes exhibit hemolysis? How accurate (and transferrable) are the central laboratory hemolysis indices? What BGM was used? Did it demonstrate any sensitivity to hemoglobin interference? What laboratory instrument provided the CLG? Larger blood volume collections for BGM evaluations are often stabilized and sent to a central laboratory for highly accurate glucose testing. What artifacts have been introduced with these collections? Was there significant milking of the fingers to obtain that 0.6 ml of blood? Does the entire large sample collection process increase the secretion of catecholamines and glucose? What inaccuracies might be incurred in a manufacturer’s glucose meter evaluation due to the apparently nonequivalent shallow and deep collections? How can these inaccuracies be prevented and/or mitigated? Should only shallow collections be performed and blood samples be pooled to obtain adequate blood volumes for central laboratory testing? We thank Dr Heenan et al for this very interesting study.