A comparison of venous versus capillary blood samples when measuring blood glucose using a point-of-care, capillary-based glucometer

Background: Blood glucose level (BGL) is routinely assessed by paramedics in the out-of-hospital setting. Most commonly, BGL is measured using a blood sample of capillary origin analyzed by a hand-held, point-of-care glucometer. In some clinical circumstances, the capillary sample may be replaced by blood of venous origin. Given most point-of-care glucometers are engineered to analyze capillary blood samples, the use of venous blood instead of capillary may lead to inaccurate or misleading measurements. Hypothesis/Problem: The aim of this prospective study was to compare mean difference in BGL between venous and capillary blood from healthy volunteers when measured using a capillary-based, hand-held, point-of-care glucometer. Methods: Using a prospective observational comparison design, 36 healthy participants provided paired samples of blood, one venous and the other capillary, taken near simultaneously. The BGL values were similar between the two groups. The capillary group had a range of 4.3mmol/l, with the lowest value being 4.4mmol/l and 8.7mmol/l the highest. The venous group had a range of 2.7mmol/l, with the lowest value being 4.1mmol/l and 7.0mmol/l the highest. For the primary research question, the mean BGL for the venous sample group was 5.3mmol/l (SD = 0.6), compared to 5.6mmol/l (SD = 0.8) for the capillary group. This represented a statistically significant difference of 0.3mmol/l (P = .04), but it did not reach the a priori established point of clinical significance (1.0mmol/l). Pearson’s correlation coefficient for capillary versus venous indicated moderate correlation (r = 0.42). Conclusion: In healthy, non-fasted people in a non-clinical setting, a statistically significant, but not clinically significant, difference was found between venous- and capillary-derived BGL when measured using a point-of-care, capillary-based glucometer. Correlation between the two was moderate. In this context, using venous samples in a capillary-based glucometer is reasonable providing the venous sample can be gathered without exposure of the clinician to risk of needle-stick injury. In clinical settings where physiological derangement or acute illness is present, capillary sampling would remain the optimal approach.