Capillary Samples and Hemolysis: Further Considerations

We are responding to the thoughtful letter by Cembrowski and Füzéry, “Deep Lancing or Over Manipulation,” in this issue of the Journal of Diabetes Science and Technology, ¹ which was written in response to our initial letter, “How Much Hemolysis Is Acceptable When Undertaking Deep Lancing for Finger Stick Derived Capillary Plasma Glucose Measurement?” ²

The purpose of our initial correspondence was to encourage open sharing of data and insights on the mechanisms and effects of hemolysis, in relation to capillary sample acquisition for plasma glucose measurement. ² We are therefore pleased that Cembrowski and Füzéry provided new data and also progressed this discussion. ¹ Hemolysis appears to be common when collecting capillary “finger stick” samples: We reported hemolysis indices from 42 participants obtained during a previous study. ³ Standardized procedures for capillary sample collection were used throughout this study. ⁴ Hemolysis (in the categories “slight,” “moderate,” or “severe”) occurred in over half of the 0.6 mL samples obtained from deep lancing. ² Aggregating the results quoted in Cembrowski and Füzéry’s correspondence describing visible hemolysis in shallow lanced samples from a previously unpublished study, 20 out of 141 “shallow” samples showed visible hemolysis after spinning. ¹

Why was the focus of our original correspondence on deep lancing? This was because its context related to the recent paper by Klonoff et al, describing a blood glucose monitor system surveillance protocol. ⁵ Their protocol suggests use of deep lancing techniques to obtain capillary samples for plasma glucose measurement on comparative glucose instruments. ⁵ Our initial correspondence commented on the paucity of published information on hemolysis, specifically within this context. We are nevertheless pleased that Cembrowski and Füzéry expanded the discussion to include shallow lancing and also the potential for non-equivalence of samples obtained using “shallow” and “deep” lancing techniques.

Cembrowski and Füzéry provide multiple additional observations on potential sources of preanalytical and analytical error in the context of capillary glucose measurements. ¹ These include (but are not limited to) equivalence of different reporting methods for hemolysis and the different mechanisms whereby hemolysis might interfere with glucose measurement, whether in the laboratory or when using glucose meters. They also discuss the impact of “over manipulation,” both of sampling sites and also of specimens, during specimen preparation. We would be keen for the discussion around optimal acquisition techniques for finger stick samples, to be expanded further. For example, routine warming (“arterializing”) of the hand appears to improve blood flow during sample collection, ⁶ perhaps minimizing hemolysis. Could “arterialization” of samples cause unintended consequences? For example, arterialization might change the proportion of glucose-rich arterial blood compared to glucose-depleted blood from venules, within the collected sample.

We concur with Cembrowski and Füzéry’s suggestion that identification of potential sources of pre-analytical and analytical error within the context of measuring capillary plasma glucose, allows better exploration of ways to minimize and mitigate these errors.