Advantages and Potential Pitfalls of Point of Care Glucose and Lactate Monitoring

Point of care (POC) monitoring is becoming common in equine patients. Evaluation of POC glucose and lactate monitors suggests that these monitors may be clinically useful in practice, but individual monitors should be compared with a standard laboratory technique regularly because of the variation in agreement when using horse blood or plasma.

1. Introduction

Decision making about the emergency patient is highly time sensitive. In emergency patients and patients undergoing treatment, certain parameters, such as lactate and glucose concentrations, should be available rapidly. This way, treatment may be changed in response to altered patient condition; additionally, time-sensitive decisions, such as referral, may be made appropriately. A patient’s test results should ideally be available to the clinician within minutes of sample collection. Therefore, point of care (POC) testing has become more popular in the emergency and critical-care settings [1].

2. Lactate Monitoring

In human medicine, admission blood lactate concentration has been shown to be a useful indicator of disease severity, and numerous studies have linked hyperlactatemia with increased morbidity and mortality [2-5]. Similarly, lactate concentration has been suggested as a useful clinical indicator in a small number of studies examining equine colic patients and critically ill equine neonates [6-11]. Although laboratory-based methods of measuring lactate concentration are readily available to clinicians working at large referral hospitals, a convenient measure of lactate concentration has not been accessible to many practitioners in the past. Inexpensive pocket-sized lactate meters designed for use by human athletes in training have become available in recent years and are being used in some practices and referral centers. The accuracy and reliability of some of these meters has been evaluated in clinically normal, exercising humans and horses. However, their accuracy has not yet been established for clinically abnormal horses, where the performance of these meters might not be ideal. To be clinically useful, the accuracy of such an analyzer should be excellent in the near normal ranges and remain acceptable over the clinically relevant range in the pertinent patient population. To be acceptable to clinicians and veterinary technicians, these analyzers must be sensitive and simple to operate, and they must provide reproducible results. At values outside of the normal range, accuracy becomes slightly less critical but should remain sufficiently reliable so that trends may be followed. In a recent study of 217 adult horses presenting to a referral hospital [12], lactate concentration measured in heparinized plasma samples on a commercially available POC lactate monitor [a] generated values with excellent agreement to those produced by a blood gas monitor [b]. The blood gas monitor is also a POC device, but it is less portable. The observed average agreement between the blood gas monitor and the lactate meter was close to 0, and the 95% confidence intervals (or limits of agreement) were relatively narrow. Using concordance analysis, the relationship between values generated by the lactate meter, again using heparinized plasma, and the blood gas meter was more disparate for lactate concentrations between 6 and 10 mmol/l; however, the results were qualitatively similar, and the information seems to be clinically useful.

The blood lactate monitor measures plasma lactate concentration, but whole blood concentration can be calculated using a proprietary algorithm. The monitor’s reported range is 0.8 - 22.0 mmol/l for whole blood lactate concentration and 0.7 - 26.0 mmol/l for plasma lactate concentration. The study showed that a POC lactate meter measures lactate concentration in plasma samples from adult equine emergent patients as well as the laboratory-based standard method. The agreement was only moderate when whole blood samples were used, and discrepancies may be considerable over the range of values seen in horses presenting as emergency admissions. However, samples allowed to sit for 10 - 20 min may provide a "clean" plasma sample to provide accurate results, and this length of time should not significantly alter lactate concentration. The use of POC lactate analysis will decrease analytical time (reduces turn-around time) and make an important diagnostic parameter immediately available in the emergency and critical care setting. Furthermore, the monitor and test strips are inexpensive, making frequent lactate measurement possible and economical.

3. Glucose Monitoring

"Stall-side" or POC glucometry has become a standard approach to measuring blood glucose concentration in foals and adult horses all over the world. However, glucometry has not been fully evaluated in horses in a systematic manner. Aberrations in glucose homeostasis are not unusual in these patients, and they can range from hypoglycemia to hyperglycemia. Routine monitoring of blood glucose concentration has become more common, because hypoglycemia may be life threatening in neonates, and hyperglycemia has been associated with poor outcome in horses with colic; therefore, veterinarians need to have a better understanding of glucometry [13].

Glucose measurements in horses are usually performed using venous blood obtained by direct venipuncture of the jugular or peripheral vein. Unlike in human hospital critical care units, capillary blood is rarely used [14-19]. Blood glucose is measured by either a reagent strip and glucometry, a multi-electrode blood gas analyzer, or a serum/plasma chemistry analyzer. Although it has been generally accepted that these measurements are "close enough" and that they might be used interchangeably for an individual patient during any treatment period, understanding of the relationships between measurement techniques is important for evaluating results and developing rational treatments. There are differences in agreement between techniques, and potentially clinically important bias does exist.

One source of discrepancy in glucometer-derived glucose readings is in the variation in reagent test strips and individual monitors. Significant differences in results using different lots of reagent strips in the same glucometer have been reported, even after manufacturer calibration instructions have been followed [20]. It is vital that glucometers be calibrated for each new reagent lot. Reagent strip age and handling is another potential source of error [21-23]. Almost all glucometer reagent strips are sensitive to exposure to humidity and wide temperature fluctuations. Removing the reagent strip from its container before it is needed or leaving the cap to the reagent strip container loose can quickly degrade the accuracy of the test [21,22]. There is even a variation in results when two different glucometers of the same model using the same lot of reagent strips are compared [21,22]. Each glucometer should only be used within its specified temperature range.

Manufacturers are aware of the problems that result in discrepancies between glucometer test results and laboratory standard tests, and they have been modifying test reagent design to report plasma-like glucose values, despite testing whole blood. Some manufacturers even claim that their meters test plasma directly after the strip excludes cells from the reagent surface [23].

The question of clinically important differences becomes vital in certain critically ill patients. The normal range of glucose concentration, determined by standard laboratory chemistry technique, changes with time in the first week of life in neonatal foals, but a range of 60 - 180 mg/dl has historically been considered "acceptable" in our neonatal intensive care unit where one study was performed [24]. When evaluated by using whole blood samples from neonates, the glucometer [c] tended to err in the direction of hypoglycemia and err away from hyperglycemia compared with the laboratory standard [d]. If hypoglycemia is potentially the most deleterious condition for the patient, as it is in the case of most sick foals, then glucometry, depending on the glucometer used, may be considered a "safe" method to use for glucose monitoring of hypoglycemia. In this patient population, the glucometer would err on the side of caution (i.e., frequent warning of potential hypoglycemia). Aberrant, clinically inconsistent, or unexpected findings should be reevaluated using laboratory standards.

If hyperglycemia is the major clinical concern, which it can be in the case of sick foals and horses with gastrointestinal disease, then glucometry seems to be fairly reliable, because it underestimates the laboratory standard [d] in the study [24]. Use of glucometry in the management of hyperglycemic patients may require adjustment of parameters of concern to different "trigger points" for detection of hypoglycemia and hyperglycemia, depending on the particular glucometer used. Again, aberrant or unexpected findings should be reevaluated using laboratory standards.

A recent study testing a different glucometer [e] in 50 adult horses presenting for emergency treatment had different results [f]. This study evaluated a glucometer that uses both plasma and whole blood, and it compared its performance to both a blood gas monitor [b] and standard laboratory chemistry [d] plasma glucose determinations. Agreement determined by concordance analysis was excellent for the performance of the glucometer (when plasma was used) compared with the laboratory standard. However, agreement deteriorated when the glucometer (plasma sample) was compared with the blood gas monitor. There was little agreement between whole blood samples in the glucometer and both the blood gas monitor and the laboratory standard.

As practiced in many veterinary hospitals, POC glucometry is practical, rapid, and affordable. However, glucometry has less than ideal agreement with the laboratory plasma chemistry standard and other POC tests like blood gas analysis. Two recent studies performed in foals and adult horses using two different makes and models of glucometersf showed this disparity [24]. The glucometers and sample types (i.e., whole blood versus plasma) that will have the best agreement with laboratory standards are quite unpredictable. It is also hard to predict if a glucometer will overestimate or underestimate the laboratory standard. There are >30 glucometers currently marketed, and each may react differently to confounding effects, resulting in its own characteristic "signature." Each hospital should critically evaluate their glucometry by periodically comparing results with their laboratory standard and recognizing that the very competitive glucometry market drives constant modifications by the manufacturers. Differences may be clinically important, and decisions regarding the management of glucose concentrations in sick foals and adults should be made with these somewhat predictable differences in mind.

4. Conclusion

In both referral and practice situations, POC glucose and lactate monitoring is inexpensive, easily performed, and potentially very clinically useful. However, individual monitors should regularly be compared with a standard laboratory technique because of variation in agreement when using horse blood or plasma. Understanding the potential limitations of any individual monitor used will aid in interpretation of the results produced by the meter.

The author thanks the following collaborators: Raymond C. Boston, PhD; Brett Tennent-Brown, BVSc, Diplomate ACVIM; Catherine Russell, BVSc; Jonathan E. Palmer, VMD, Diplomate ACVIM; Anna R. Hollis, BVSc, MRCVS; Barbara L. Dallap Schaer, VMD, Diplomate ACVS, Diplomate ACVECC; Sue Lindborg, BS, CVT; Gail Russell, BS, CVT; and the staff of the Clinical Laboratory at New Bolton Center. Funding for some of the cited studies was provided by Jerry Cosgrove and Roche Diagnostics (Accutrend lactate monitor and initial supplies), the Raker-Tulleners’ Research Fund, and the Raymond Firestone Research Fund.

Footnotes

[a] Accutrend, Roche Diagnostics, Indianapolis, IN 46250.
[b] Critical Care Express, Nova Biomedical, Waltham, MA 02453.
[c] Accu-Chek Advantage, Roche Diagnostics Corporation, Indianapolis, IN 46250.
[d] Vitros Chemistry Products, Ortho-Clinical Diagnostics, Rochester, NY 14626.
[e] Accu-Chek Aviva, Roche Diagnostics Corporation, Indianapolis, IN 46250.
[f] Hollis AR, Dallap Schaer BL, Boston RC, Wilkins PA. Unpublished data, 2007.