The LP2 and ABL agreed well in arterial blood taken from both healthy volunteers and intensive care patients. Capillary blood lactate values in the fingertip were significantly higher than the corresponding arterial values in both ICU patients and healthy volunteers. Capillary blood lactate values in the earlobe were significantly higher than arterial blood lactate values in the ICU group but not in the healthy volunteer group.
The earlobe seems to be a better sample site than the fingertip, possibly because it is more central than the fingertip. However, more importantly, it is less sensitive to the variation between two measurements taken consecutively. Contenti et al. used the earlobe as a sample site for capillary blood and found that capillary blood lactate was higher than both venous and arterial lactate in patients in an ED, while in our study, ICU patients had higher capillary blood lactate than arterial blood lactate but almost the same as venous blood lactate (20).
Studies performed with other handheld devices in ICU populations also report good agreement between capillary and arterial lactate, but the devices were validated in lower lactate ranges (21-23).
Collange et al. is the only other study that compared both capillary and arterial blood samples on a handheld device instead of comparing capillary blood measured on a handheld device with arterial blood measured on a reference method. Like us, they found that capillary lactate was higher than arterial blood lactate. However, in contrast to our findings, they reported that the handheld instrument measured slightly lower lactate values than the reference instrument using arterial blood (22).
Interpretations of capillary blood lactate levels in patients suffering from shock may be challenging because of discrepancies between capillary and arterial blood due to changes in peripheral perfusion following vasopressor use. Vasopressors may also increase lactate production via stimulation of skeletal muscle beta-2 adrenergic receptors, and reduce lactate clearance in peripheral tissues, and thereby increase the difference between arterial, venous and capillary lactate values (24). Hence, vasopressors may also influence the difference in capillary blood lactate between the fingertip and earlobe. Pattharanitima et al. found no evidence of norepinephrine use effecting the relationship between capillary or arterial lactate (23). We did not adjust for vasopressor as a confounder in the ICU group, and therefore our results may not be directly applicable to the prehospital patients who do not receive vasopressor.
Lactate monitoring in EMS may help the provider in detecting shock at an earlier stage and may be used as an early trigger for blood transfusion in haemorrhaging patients or for fluid therapy in septic patients (14). In general, arterial blood is preferred. Capillary lactate should only be measured when rapid measurement is necessary or when arterial blood is not available. Since patients in the prehospital setting in most cases do not receive an arterial line, a handheld capillary blood lactate measuring device may be well suited for EMS services. Measuring lactate values in the field is quick and should not delay other treatments or interventions on-scene. Our results show that LP2 also has the potential to overestimate the lactate values, which may lead to overtreatment or overtriage of patients. This may, however, be of less importance than the potential consequences of under-triage in situations where EMS providers fail to detect the early stages of shock.
Despite overestimation of the actual values in a single reading, this value may be helpful as an adjunct to the other vital signs during assessment of these patients. Further, multiple lactate readings may provide even more information to the EMS provider because a lactate trend may reflect the clinical course of shock and the potential effect of resuscitation. This may be of value for clinical decision making and for guiding treatment. When interpreting the lactate values, one must consider both the instrumental bias of the LP2 and the physiological discrepancy between arterial and capillary blood lactate values in different sample sites and different haemodynamical states.
Strengths and limitations
The strength of this study lies in the standardized procedure for collecting blood and analysing lactate. In contrast to other studies comparing capillary and arterial blood lactate we measured both capillary and arterial/venous blood on the same instrument. Most other studies compare the capillary lactate value measured on the handheld device with the arterial value measured by the reference method. We believe our method is more correct because it separates the instrument agreement from the difference in the different blood samples. Another strength is that we validated LP2 in a wide range of lactate values (from 1-25 mmol/l) in both healthy volunteers and intensive care patients, which shows the differences in lactate distribution in capillary and arterial/venous blood in these two haemodynamically different groups. This study has some limitations of note. We have previously observed that cold LP2 strips may influence the measured value randomly; therefore, we made sure that the LP2 and the strips were stored at room temperature. This limitation in the instrument may have caused a bias in patients with low body temperature. We are considering further research on this issue. The manufacturer states that the device must be used between 5 and 40 degrees Celsius and should be adjusted to the surroundings for at least 20 minutes. In the pre-hospital environment, the temperature may exceed these limits. This requires the device to be stored in a temperature-controlled environment, e.g. in an isolated casing.
Regarding capillary measurements, the disinfecting agent can cause haemolysis, which may increase the lactate concentration. This study includes few study subjects, which limits the generalizability of the study findings. The number of healthy volunteers included is low due to the ethical aspect of arterial cannulation and the associated risk of complication. The number of missing values in our data set, partly due to the detection limits of the LP2 also constitutes a limitation.