Whole-Blood Glucose and Lactate.
Rapid glucose and lactate monitoring is essential for management of critically ill and diabetic patients. Numerous drugs can alter laboratory test results and therefore can confuse the physician. The clinical significance of drug interference with a laboratory method should be understood when interpreting test results. Generally, problems with interfering substances have slowed the clinical implementation of whole-blood biosensors. Spurious data from drug interference have been reported for glucose measurements. Little is known about drug interference with lactate methods. Changes in glucose and lactate levels due to metabolism of the sample in vitro also must be considered. Therefore, the objectives of this study were (a) to study trilayer electrochemical biosensors for whole-blood analysis of glucose and lactate, (b) to investigate potential interference with glucose and lactate measurements by 30 drugs frequently used in critical care, (c) to quantify the effects of metabolism on glucose and lactate levels in vitro, and (d) to recommend practice guidelines for measurements of these 2 analytes.
MATERIALS AND METHODS
Trilayer Electrochemical Biosensors
Figure 1 shows the basic structure of the electrochemical bio sensors. The measurement principle is based on the oxidation of either glucose or lactate following diffusion from the whole-blood sample to the enzyme layer (equations 1 and 2).
(1) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
(2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
(3) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Hydrogen peroxide produced in the enzymatic reactions diffuses to the platinum electrode and is oxidized at constant potential of 675 mV (equation 3), generating current proportional to the glucose or lactate levels (activity, molality) present in the sample.[3,4] Glucose and lactate levels are calibrated using a 2-point process.
The glucose and lactate biosensors are used in conjunction with an ABL625-GL whole-blood analyzer (Radiometer America, Indianapolis, Ind), which also measures electrolytes ([Na.sup.+], [K.sup.+], [Cl.sup.-], [Ca.sup.++]), blood gasses ([PO.sub.2], [PCO.sub.2]]), and pH, and can perform oximetry (total hemoglobin, [SO.sup.2], carboxyhemoglobin, reduced and oxygenated hemoglobin, methemoglobin) on approximately 210 [micro]L or less of whole blood, depending on the analytes selected. The minimum volume of blood required for glucose or lactate measurement is 30 [micro]L. The analysis time in these experiments was 57 seconds. The cycle time was 126 seconds.
Drug Selection and Concentrations
Thirty drugs were selected for study based on pharmacy statistics for critical care and hospital patients at the University of California, Davis, Medical Center. Table 1 summarizes the therapeutic, toxic, and test concentrations. Tested concentrations were (a) those recommended by the National Committee for Clinical Laboratory Standards (NCCLS); (b) 10 times the highest therapeutic concentrations used[6,7]; or, if these concentrations were not known, (c) the concentration equivalent to the highest therapeutic dose assumed to be distributed in 5 L of blood volume.
Table 1. Effect of Drugs on Glucose and Lactate Measurements
Therapeutic Toxic Level, Test Level, [micro]g/mL Concentration, Drug [micro]g/mL [micro]g/mL(*) Acetaminophen 10-20 150 200 Aminophyllin 10-20 30-40 250 Ampicillin 5 ... 50 Ascorbic acid 8-12 ... 30 Cefazolin 400 ... 4000 Cimetidine 1-10 ... 100 Dexamethasone ... ... 4 Digoxin [is greater 0.0017-0.0033 0.0033 than] 0.0008 Dobutamine 35-50 ... 0.05 Dopamine ... ... 130 Epinephrine ... ... 0.2 Erythromycin 2-20 ... 20 Furosemide 1-3 ... 20 Gentamycin 8-12 10-30 120 Glipizide ... ... 4 Heparin sodium ... ... 8 Hydralazine 0.1 ... 1 Lidocaine 1.5-6 9-14 60 Mannitol ... ... 20000 Nitroglycerin 0.0012-0.011 ... 0.11 Nitroprusside ... ... 7.5 Norepinephrine ... ... 10 Penicillin G ... ... 2400 Phentolamine ... ... 5 Phenytoin 10-20 20-40 + 100 Procainamide 4-10 10-12 + 100 Quinidine 0.3-6 10 50 Regular [is less than or ... 1000 insulin equal to] 100 Salicylate 20-100 150-300 500 Warfarin 1-10 ... 100 Drug Effect [Delta] Glucose, Lactate, Drug mmol/L [Delta] mmol/L Acetaminophen -0.039 0.010 Aminophyllin 0.085 -0.020 Ampicillin 0.032 -0.008 Ascorbic acid 0.127 -0.030 Cefazolin -0.032 0.003 Cimetidine 0.057 -0.020 Dexamethasone 0.037 0.003 Digoxin 0.118 0.040 Dobutamine 0.053 -0.025 Dopamine 0.055 -0.028 Epinephrine 0.020 -0.035 Erythromycin 0.049 -0.020 Furosemide 0.078 -0.002 Gentamycin 0.100 0.010 Glipizide 0.095 0.025 Heparin sodium -0.016 0.000 Hydralazine 0.081 0.025 Lidocaine 0.007 0.005 Mannitol -0.766 -0.225 Nitroglycerin 0.002 0.005 Nitroprusside -0.032 -0.023 Norepinephrine 0.111 -0.013 Penicillin G -0.021 -0.035 Phentolamine 0.032 -0.008 Phenytoin 0.000 -0.003 Procainamide 0.097 0.033 Quinidine 0.025 0.075 Regular insulin 0.016 0.045 Salicylate 0.065 0.003 Warfarin -0.009 0.007
(*) The unit of drug concentrations is given in [micro]g/mL, except heparin sodium (U/mL) and regular insulin ([micro]U/mL).
This study followed the guidelines of the Human Subjects Committee. Volunteer subjects were healthy and had not taken medications prior to blood donation. Blood was collected in lithium heparin Vacutainers (Becton Dickinson, Rutherford, NJ) and pooled after inverting the tubes 20 times. Plasma was separated by centrifuging whole blood at 2500g for 10 minutes.
Protocol for Study of Metabolism In Vitro
To assess the effects of metabolism on glucose and lactate levels in vitro at room temperature, whole-blood glucose and lactate levels were measured every 3 minutes for 45 minutes at room temperature. The metabolism experiment was performed 3 times. The mean effects of metabolism on analyte levels were plotted versus time.
Drug Interference Protocol
The volume of drug added to a blood sample was less than 10% of the total volume per NCCLS guidelines. Drug stocks were diluted with saline. Tablet drugs were ground into fine powder, then dissolved completely in saline at 37 [degrees] C. Each drug stock solution was prepared and used the same day that the drug was studied. Drugs were added to whole blood to obtain the test concentrations listed in Table 1. The total volume of blood and drug for each sample was 2 mL. Control samples were made by substituting an equal volume of saline for the volume of drug added to the whole-blood sample, so that the dilutions of the spiked and control samples were identical. Each experiment consisted of a control sample and 2 duplicate spiked samples. Analyte levels in each sample were measured in duplicate, starting with the control sample and then the 2 spiked samples. Exact times of glucose and lactate measurements were documented.
For each drug studied, control means were subtracted from drug spiked means to calculate the paired differences.
[Delta] [Glucose] = [[Glucose].sub.spiked] - [[Glucose].sub.control]
[Delta] [Lactate] = [[Lactate].sub.spiked] - [[Lactate].sub.control]
The measurement cycle time of the analyzer caused a time delay between serial measurements. The time elapsed between the first and last measurements was 5 cycles x 126 seconds per cycle = 630 seconds or 10.5 minutes. Metabolic changes in glucose and lactate levels were included in the calculation of the net effect of the drug, [Delta][[Analyte].sub.Drug Effect], at the time of analyte measurement.
Total Change = [Delta][[Analyte].sub.Drug Effect]
[Delta][[Analyte].sub.Drug Effect] = Total Change - [Delta][[Analyte].sub.Metabolism]
The change in analyte level versus time for the first 15 minutes of the study of metabolism in vitro was used to fit a least squares linear regression line. The slope of this line was used to calculate the change in analyte level due to metabolism, [Delta][[Analyte].sub.Metabolism], for each drug at the exact time of analyte measurement in the preceding equation.
Quantitative Criteria for Drug Interference and Dose-Response Relationships
The error tolerance for drug effects on analyte measurements was [+ or -] 2 SD, where the SD was obtained from the within-day precision evaluated in this study. If drug effects exceeded this error tolerance, the drug was tested at 5 levels to obtain a dose-response relationship for whole-blood samples. Dose-response relationships with plasma sample measurements also were evaluated.
Calibration and Precision
The analyzer automatically performed a 2-point calibration every 4 hours after start-up and a 1-point calibration after every 10 measurements. The analyzer also was calibrated if there was drift or instability in measurements. Day-to-day precision of the glucose and lactate biosensors was determined using 3 levels of aqueous quality control (QC) at the beginning and end of each experiment. Within-day precision was determined for aqueous QC solutions and for whole blood.
Statistics and Units
Interference of the largest mannitol concentration (24 mg/dL) was assessed using a Student t test for paired differences; a P value less than .05 was considered statistically significant. Drug units are those conventionally used. Analyte values are reported in both conventional and System International units. For glucose, mmol/L = mg/dL x 0.05551, and for lactate, mmol/L = mg/ dL x 0.1111.
Within-day precision for whole blood was 1.7% (mean 81.4, SD 1.4 mg/dL; 4.52, 0.08 mmol/L) for glucose (n = 20) and 6.6% (9.5, SD 0.6 mg/dL; 1.06, 0.07 mmol/L) for lactate (n = 10). Table 2 presents the precision results for aqueous QC solutions. Coefficients of variation ranged from 2.03% to 3.94% for between-day precision and from 0.92% to 3.90% for within-day precision. The [+ or -] 2 SD interference thresholds, taken from the within-day precision results, were from QC level 2 (4.39-6.0 mmol/L, 79-108 mg/dL) for glucose and from QC level 3 (0.9-1.7 mmol/ L, 8.1-15.3 mg/dL) for lactate. These QC levels corresponded to the control levels measured in normal subjects and also to published normal levels. The error tolerances (ie, [+ or -] 2 SD) for determination of drug interference were [+ or -] 0.09 X 2 = [+ or -] 0.18 mmol/L (3.24 mg/dL) for glucose and [+ or -] 0.05 X 2 = [+ or -] 0.10 mmol/L (0.9 mg/dL) for lactate.
Table 2. Glucose and Lactate Precision
Quality Control Range Between-day Precision (n = 30)(*) Level mmol/L mg/dL Mean SD CV, % Glucose 1 12.5-15.5 225-279 13.81 0.28 2.03 2 4.39-6.0 79-108 5.37 0.16 2.98 3 1.5-2.5 27-45 2.19 0.05 2.28 Lactate 1 7.7-10.1 69.3-90.9 8.58 0.27 3.15 2 3.8-5.2 34.2-46.8 4.35 0.13 2.99 3 0.9-1.7 8.1-15.3 1.27 0.05 3.94 Within-day Precision (n = 30)(*) Level Mean SD CV, % Glucose 1 14.08 0.13 0.92 2 5.37 0.09 1.68 3 2.12 0.06 2.83 Lactate 1 8.94 0.22 2.46 2 4.48 0.10 2.23 3 1.28 0.05 3.90
(*) Units are in millimoles per liter.
The metabolic change in glucose was -2.3% (0.3% SD) at 15 minutes, -4.6% (0.2% SD) at 30 minutes, and -6.4% (0.5% SD) at 45 minutes at room temperature. The change in lactate was 11.4% (2.0% SD) at 15 minutes, 20.6% (3.8% SD) at 30 minutes, and 26.7% (4.8% SD) at 45 minutes. Figure 2 shows the trends in glucose and lactate. The metabolic rates during the first 15 minutes were -0.00848 mmol/L (-0.153 mg/dL) per minute for glucose ([r.sup.2] = 0.985) and 0.0100 mmol/L (0.090 mg/dL) per minute for lactate ([r.sup.2] = 0.992). The average time elapsed between measurements of the spiked and control samples in the screening tests was 6.25 to 10.25 minutes. Table 1 summarizes (after subtracting changes from metabolism) the net drug effects on glucose and lactate measurement for all 30 drugs.
[Figure 2 ILLUSTRATION OMITTED]
For control conditions (n = 30) for drug testing, the mean and range of the analytes were 4.5 (0.5 SD) mmol/ L (81 [9.6 SD] mg/dL) and 3.5 to 5.3 mmol/L (63-96 mg/ dL) for glucose and 1.5 (0.4 SD) mmol/L (13.5 [3.6 SD] mg/dL) and 0.7 to 2.5 mmol/L (6.3-22.5 mg/dL) for lactate. The 2 right-hand columns in Table 1 show that of the 30 drugs tested, based on the error tolerance, only mannitol ([C.sub.6][H.sub.14][O.sub.6]) interfered significantly with the analyte measurements. At the test concentration of 20 mg/mL, mannitol produced an average error, with respect to control measurements, of -0.766 mmol/L (-13.8 mg/dL) in glucose concentration and -0.225 mmol/L (-2.0 mg/dL) in lactate concentration. These errors were both outside the precision-based error tolerance. Therefore, dose-response experiments were performed covering this range of mannitol concentrations. Other drugs that demonstrated intermediate interference, but did not exceed the [+ or -] 0.18 mmol/L threshold for glucose, were ascorbic acid (error = 0.127 mmol/L), digoxin (0.118 retool/L), gentamicin (0.100 mmol/L), and norepinephrine (0.111 mmol/L). With the exception of mannitol, interference of drugs with lactate measurements was uniformly insignificant.
Figure 3 shows the metabolism-corrected dose-response relationships for mannitol. For whole-blood samples, mannitol-induced error is outside the error tolerance at a mannitol concentration of approximately 5.6 mg/dL for glucose and 18.2 mg/dL for lactate. At a concentration of 24 mg/ mL, mannitol caused errors of -0.71 (SD 0.13) mmol/L (-12.8 [SD 2.3] mg/dL) in glucose and -0.16 (SD 0.09) mmol/L (-1.4 [SD 0.79] mg/dL) in lactate whole-blood measurements (P [is less than] .05). Errors in measurements for plasma samples were consistently smaller. For example, at the same mannitol concentration of 24 mg/mL, the glucose error was -0.41 (SD 0.07) mmol/L (-7.4 [SD 1.2] mg/dL), and the lactate error was -0.083 (SD 0.03) mmol/L (-1.5 [SD 0.52] mg/dL) for plasma samples (P [is less than] .05).
[Figure 3 ILLUSTRATION OMITTED]
Glucose and Lactate in Rapid Decision Making
Careful control of the blood glucose level is important in preventing nephropathy, retinopathy, and neuropathy. In hospitalized patients, Hirsch et al and Queale et al recommended algorithmic insulin treatment based on frequent glucose measurements to maintain blood glucose within the therapeutic brackets for good glycemic control. Recommended therapeutic brackets are 120 to 200 mg/dL for hospitalized patients, and for nonhospitalized patients, 80 to 120 mg/dL preprandially and 100 to 140 mg/dL at bedtime. Lactate, the product of anaerobic metabolism of glucose, is important in a number of clinical situations (for a detailed review, please see references 11 and 12). Recent interesting data published by Bakker et al suggest that elevated lactate levels are prognostic of poor outcome in patients with multiple organ failure following septic shock. Interestingly, these data show that the magnitude, duration, and integrated trend of lactate levels are predictive of poor outcomes, including increased morbidity and mortality. For example, patients who died during the first 24 hours of septic shock had higher blood lactate levels (9.6 [+ or -] 5.3 SD mmol/L) than those who died later (5.6 [+ or -] 3.7 mmol/L). Thus, the ability to follow acute changes in glucose and lactate levels is important for medical decision making in critical care settings.
Practice Guidelines for Whole-Blood Analysis and Specimen Processing in Critical Care Settings
During life-saving resuscitation speed is of the essence. Common critical care problems, such as acute myocardial infarction, seizures, and shock, demand rapid response. Immediate diagnosis and therapy are equally important. Diagnostic and therapeutic processes[11,12] evolve quickly. Diagnostic test results may be needed in as few as 5 minutes.[14,15] The most pragmatic approach for analyzing glucose and lactate levels is to provide direct, biosensor-based, whole-blood measurements because of speed and blood volume conservation.[16,17] Whole-blood analysis can be performed at the point of care or elsewhere if rapid transport systems are available. The whole-blood analyzer used in this study is transportable, can be used on a cart at the bedside, and is well suited for a near-patient, satellite, or rapid-response laboratory. A compact model using the same biosensor electrochemistry has been released recently. In general, transportable whole-blood analyzers can provide simultaneous measurements of electrolytes ([K.sup.+], [Ca.sup.++], [Na.sup.+], [Cl.sup.-], [Mg.sup.++]), blood gases ([PO.sub.2], [[PCO.sub.2]]), pH, total carbon dioxide content, hematocrit, hemoglobin, and metabolites (glucose, lactate, urea nitrogen, creatinine), as well as co-oximetry in 1 to 2 minutes, depending on the tests selected. Handheld devices for point-of-care glucose and lactate[21,22] testing also have been studied in critical care settings.
Table 3 summarizes metabolic changes in glucose and lactate levels documented in this and other studies. Glycolytic inhibitors, such as fluoride, may interfere with enzymatic reactions in electrochemical biosensors and also may cause hemolysis. Therefore, glycolytic inhibitors were not used in this study. Heparin did not interfere with measurements (see Table 1). Therefore, heparin was used as the anticoagulant. The results of this study, which was limited to healthy subjects, show that room temperature is acceptable for specimen processing if glucose and lactate measurements are performed within a few minutes (see Figure 2). Table 4 summarizes practice guidelines for whole-blood analysis specimen processing. The recommended therapeutic turnaround time for critical emergencies is 5 minutes.[14,15] If analyzed within this short period of time, glucose and lactate levels change less than 1 mg/ dL at room temperature. Metabolic changes were about -5% for glucose at 30 minutes and greater than 10% for lactate at 15 minutes. These changes may be large enough to affect diagnosis or treatment in critical care.[23,24] Longer delays may impair medical decisions based on glucose and lactate values, as well as on other analyte levels measured in whole blood, especially if there are rapid disease-related changes in actual patients. For example, a decrease in glucose of about 5 mg/dL (at 30 minutes) may move a patient's level into or out of the therapeutic brackets and may warrant treatment, and an increase in the lactate level above 2 mmol/L accumulates "lactime" (abnormal prognosticator) when trending shock patients. Immediate whole-blood analysis should be performed if (a) the patient's history is not known, (b) metabolism in the specimen is not normal eg, leukocytosis [is greater than] 50 000/[micro]L or platelets [is greater than] 600 000/[micro]L), (c) oxygen tension in the specimen is not stable (eg, hemoglobin [is less than] 7.5 g/dL or P[O.sub.2] [is greater than] 200 mm Hg), or (d) there is a critical emergency or resuscitation where rapid response is lifesaving.[14,15,25]
Table 3. Rates of Glucose and Lactate Metabolism In Vitro
First Reference Analyte Sample Temperature Sazama et al Glucose Serum in sterile Room (1979) Vacutainer (before temperature cell separation) Chan et al Glucose Whole blood Room (1989) heparinized temperature Sacks (1999) Glucose Nonhemolyzed sterile 25 [degrees] C serum (cells 4 [degrees] C separated) Present study Glucose Whole blood Room heparinized temperature Geyssant et Lactate Whole blood 20 [degrees] C al (1985) (heparinized?) 4 [dergrees] C Wandrup Lactate Whole blood Room (1989) heparinized temperature Toffaletti et Lactate Whole blood Room al (1992) heparinized temperature Ice bath Present study Lactate Whole blood Room heparinized temperature Rate or Change and Time Interval First Reference mmol/L mg/dL Sazama et al -0.00178 per min -0.032 per min (1979) Chan et al -0.08 at 15 min -1.4 at 15 min (1989) -0.16 at 30 min -2.9 at 30 min -0.25 at 45 min -4.5 at 45 min -0.32 at 60 min -5.8 at 60 min Sacks (1999) Stable 8 h Stable 72 h Present study -0.00848 per min to -0.153 per min to 15 min 15 min -0.26 at 30 min -4.68 at 30 min -0.36 at 45 min -6.49 at 45 min Geyssant et 0.5 per 60 min 4.5 per 60 min al (1985) Stable 2 h Wandrup 0.42 per 60 min 3.78 per 60 min (1989) Toffaletti et 0.5 at 30 min 4.5 at 30 min al (1992) 0.1 at 60 min 0.9 at 60 min 0.2 at 120 min 1.8 at 120 min Present study 0.010 per min to 0.90 per min to 15 min 15 min 0.27 at 30 min 2.43 at 30 min 0.35 at 45 min 3.15 at 45 min
Table 4. Practice Guidelines for Whole-Blood Analysis Specimen Processing(*)
Time Limit Container Temperature Immediate Glass syringe Room temperature 15 min Plastic syringe Room temperature 30 min Plastic syringe Room temperature 30 min Plastic syringe Ice slush 1h Glass syringe or capillary Ice slush 2h Glass syringe or capillary Ice slush Time Limit Measurements Immediate [PO.sub.2] when [is greater than] 200 mm Hg 15 min [PO.sub.2], [O.sub.2] saturation, lactate 30 min Glucose 30 min Acid-base (pH, [PCO.sub.2]], [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]), electrolytes ([Na.sup.+], [K.sup.+], [Cl.sup.-], [Ca.sup.++]) 1 h [K.sup.+] (plus the analytes below) 2 h pH, [PCO.sub.2], [PO.sub.2], [O.sub.2] saturation, hemoglobin, hematocrit, [Ca.sup.++], and other electrolytes (except [K.sup.+])
(*) These guidelines are from Burnett et al, with the exception of glucose and lactate, for which the guidelines for sample processing are based on the results reported in this article. Immediate means as soon as possible, ideally within 5 minutes. Caution is advised when potassium measurements from samples stored in ice slush are delayed more than 30 minutes.
Importance of Plasma Measurements
The patient's drug history may not be known initially during emergency resuscitation. Nevertheless, glucose and lactate levels are needed immediately for diagnosis and treatment. Therefore, it is desirable to use a whole-blood method that is free of significant drug interference when test results are needed quickly. In this study, only mannitol interfered significantly with whole-blood glucose measurements. The maximum mannitol glucose interference, -0.71 mmol/L (-12.8 mg/dL), could mislead clinical decisions pivoting near the thresholds for hypoglycemia or hyperglycemia for hospitalized patients. Interference with lactate measurements, -0.16 mmol/L (-1.4 mg/dL) at the maximum mannitol level tested of 24 mg/ dL, is not clinically important in most patients.
Mannitol is an osmotic agent used in critical care for the treatment of increased cerebrospinal or intraocular pressure and in the prevention of acute renal failure due to hemoglobinuria or myoglobinuria.[26-28] Clinical doses may be as high as 120 g/d.[6,7] Interference has been observed on macroanalyzers. A concentration range of 0 to 137 mmo1/L was used in one study of mannitol interference with an automated serum phosphate assay. The range used here, 0 to 24 mg/mL (0-131.8 mmol/L), is comparable. However, higher blood concentrations may occur in mannitol-induced renal failure.[31-36] Hence, in this situation, measurements of glucose and lactate in plasma samples may be better (compared with whole blood) for decision making because of less interference. However, recent studies[37-40] show that lactate levels in plasma may be higher than those in whole blood.
Mechanisms of Mannitol Interference
Mannitol may interfere with glucose and lactate measurements by suppressing the formation of [H.sub.2][O.sub.2] in the enzymatic reaction of the analytes in the middle layer of the biosensor, by repartitioning water from erythrocytes to plasma, or by other mechanisms, such as changes in viscosity, direct inhibition of glucose oxidase or lactate oxidase, and hemolysate formation with the release of electroactive glutathione from erythrocytes. Mannitol forms a stable equimolar compound with [H.sub.2][O.sub.2] and lowers [H.sub.2][O.sub.2] concentrations.[41,43] Gillbe et al suggested that mannitol directly reacted with [H.sub.2][O.sub.2] or decreased its formation. Therefore, in the presence of mannitol the diffusion of [H.sub.2][O.sub.2] may be limited and the amount of [H.sub.2][O.sub.2] available for oxidation at the electrode surface in the biosensor may be reduced, which would lower the glucose and lactate readings.
Mannitol also increases plasma osmolality, without necessarily penetrating healthy erythrocyte membranes. Increased plasma osmolality will cause water to diffuse out of erythrocytes (and, in vivo, out of other tissues), thereby diluting plasma glucose and lactate concentrations; however, the level of dilution may be unequal because the intracellular versus extracellular distribution of glucose and lactate differ. The experiments performed on whole-blood and plasma samples suggest that at least 2 of these mechanisms come into play in decreasing glucose and lactate readings in the presence of increasing concentrations of mannitol (see Figure 3). However, mannitol effects in vivo may be more complicated because of the volume of distribution of this drug and its equilibration outside the vascular compartment.
Other Drugs and Metabolites
In general, potential for interference depends on a drug's molecular weight, polarity, concentration, and time of exposure (factors which determine electrode access) and on biochemical and electrochemical interactions in the electrode. Ascorbic acid and acetaminophen interfere with glucose measurements performed on other analyzers.[46,47] In vivo nitroprusside degrades quickly to form cyanide, which then is metabolized to thiocyanate. Randell and St. Louis reported that thiocyanate interferes with electrochemical measurements of glucose in whole blood. While we found no significant interference (performed with no sample illumination) from nitroprusside, we cannot rule out the possibility that metabolites in vivo might affect glucose or lactate measurements. Other drug metabolites, as well as other drugs not tested, also may interfere with the glucose or lactate trilayer biosensors. Future studies should document whether substances that we did not study will interfere. However, the list of drugs in Table 1 covers the bulk of important drugs used in our health system in the treatment of critically ill patients.
The trilayer electrochemical biosensor eliminated significant interference by strong oxidizing substances, which have been shown to interfere with glucose measurements by other methods. Mannitol interfered with both glucose and lactate measurements. The interference was statistically significant, but may not reach clinical significance in routine practice, with the possible exception of mannitol-induced renal failure and other conditions that produce high levels of mannitol. In these cases, glucose interference could affect clinical decisions.
Mechanisms by which mannitol may produce this interference include interaction with [H.sub.2][0.sub.2] in the coupled enzymatic oxidation in the biosensor and repartitioning of water between erythrocytes and plasma. Mannitol interference should be taken into consideration with critically ill patients. Plasma measurements minimize interference. Future studies should address other drugs and drug metabolites, and other instruments, since rapid whole-blood analysis has several advantages in critical care settings.
Whole-blood measurements should be performed within 15 minutes for lactate and within 30 minutes for glucose to avoid preanalytic error from metabolism in vitro. These recommendations extend current practice guidelines for whole-blood analysis. In cases where the patient's history is not known, the patient has leukocytosis or thrombocytosis, or there is critical need for rapid response, immediate whole-blood analysis is clinically appropriate.
The authors thank the vendor for supplying equipment and reagents and thank other donors for contributing to the Point-of-Care Testing Center for Teaching and Research, University of California, Davis, where the research was performed.
[1.] Christiansen TF, Jakobsen KM. The slow penetration of enzyme-based biosensors into clinical chemistry analysis. Acta Anaesthesiol Scand. 1995;(suppl 104):31-35.
[2.] Palleschi G, Rahni MAN, Lubrano GJ, et al. A study of interferences in glucose measurement in blood by hydrogen peroxide based glucose probes. Anal Biochem. 1986;159:114-121.
[3.] Clark LC, Lyons C. Electrode system for continuous monitoring in cardiovascular surgery. Ann N Y Acad Sci. 1962;10:29-45.
[4.] Clark LC, Noyes LK, Grooms TA, et al. Rapid micromeasurement of lactate in whole blood. Crit Care Med. 1984;12:461-464.
[5.] Powers DM, Boyd JC, Glick MR, et al. Interference Testing in Clinical Chemistry. EP7-P, Vol 6, No 13. Villanova, Pa: National Committee for Clinical Laboratory Standards; 1986.
[6.] Goodman G. Goodman & Gilman's the Pharmacological Basis of Therapeutics. 9th ed. New York, NY: McGraw-Hill; 1996.
[7.] Zaloga GP Drew M. The Critical Care Handbook. 2nd ed. New York, NY: Mosby Year Book; 1997.
[8.] Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329: 977-986.
[9.] Hirsch IB, Brunzell J, Paauw DS. Inpatient management of adults with diabetes. Diabetes Care. 1995;18:870-878.
[10.] Queale WS, Seidler AJ, Brancati FL. Glycemic control and sliding scale insulin use in medical inpatients with diabetes mellitus. Arch Intern Med. 1997; 157:545-552.
[11.] Kost GJ. Point-of-care testing ?? the hybrid laboratory ?? knowledge optimization. In: Kost GJ, ed. Handbook of Clinical Automation, Robotics, and Optimization. New York, NY: John Wiley and Sons; 1996:757-838.
[12.] Kost GJ. Point-of-care testing in intensive care. In: Tobin M, ed. Principles and Practice of Intensive Care Monitoring. New York, NY: McGraw-Hill; 1998: 1267-1296.
[13.] Bakker J, Gris P, Coffemils M, Kahn RJ, Vincent LJ. Serial blood lactate levels can predict the development of multiple organ failure following septic shock. Am J Surg. 1996;171:221-226.
[14.] Kost GJ. Guidelines for point-of-care testing: improving patient outcomes. Am J Clin Pathol. 1995;104(4 suppl 1):S111-S127.
[15.] Kost GJ, Ehrmeyer SS, Chernow B, et al. The laboratory-clinical interface: point-of-care testing. Chest. 1999;115:1140-1154.
[16.] Kost GJ, Wiese DA, Bowen TP. New whole blood methods and instruments: glucose measurement and test menus for critical care. J Int Fed Clin Chem. 1991;3:160-172.
[17.] Kost GJ. New whole blood analyzers and their impact on cardiac and critical care. Crit Rev Clin Lab Sci. 1993;30:153-202.
[18.] Kost GJ. Planning and implementing point-of-care testing systems. In: Tobin M, ed. Principles and Practice of Intensive Care Monitoring. New York, NY: McGraw-Hill; 1998:1297-1328.
[19.] Kost GJ, Vu HT, Inn M, et al. Multicenter study of whole-blood creatinine, total carbon dioxide content, and chemistry profiling for laboratory and point-of-care testing in critical care in the United States. Crit Care Med. 2000. In press.
[20.] Kost GJ, Vu HT, Lee JH, et al. Multicenter study of oxygen-insensitive handheld glucose point-of-care testing in critical care/hospital/ambulatory patients in the United States and Canada. Crit Care Med. 1998;26:581-590.
[21.] Mizock BA. Lactate and point-of-care testing. Crit Care Med. 1998;26: 1474-1476.
[22.] Slomovitz BM, Lavery RF, Tortella BJ, et al. Validation of a hand-held lactate device in determination of blood lactate in critically injured patients. Crit Care Med. 1998;26:1523-1528.
[23.] Toffaletti JG. Blood lactate: biochemistry, laboratory methods, and clinical interpretation. Crit Rev Clin Lab Sci. 1991;28:253-268.
[24.] Toffaletti J. Elevations in blood lactate: overview of use in critical care. Stand J Clin Lab Invest. 1996;56(suppl 224):107-110.
[25.] Burnett RW, Covington AK, Fogh-Andersen N, et al. Recommendations on whole blood sampling, transport, and storage for simultaneous determination of pH, blood gases, and electrolytes: International Federation of Clinical Chemistry Scientific Division. J Int Fed Clin Chem. 1994;6:115-120.
[26.] Gennari FJ, Kassirer JP. Osmotic diuresis. N Engl J Med. 1974;291:714-720.
[27.] Ambuhl PM, Ballmer PE, Drahenbuhl S, et al. Qualification and predictors of plasma volume expansion from mannitol treatment. Intensive Care Med. 1977; 23:1159-1164.
[28.] Rudehill A, Gordon E, Ohman G, et al. Pharmacokinetics and effects of mannitol on hemodynamics, blood and cerebrospinal fluid electrolytes, and osmolality during intracranial surgery. J Neurosurg Anaesth. 1993;5:4-12.
[29.] Landesman PW, Lott JA, Zager RA. Mannitol interferes with the Du Pont aca methods for inorganic phosphorus. Clin Chem. 1982;28:1994-1995.
[30.] Eisenbrey AB, Mathew R, Kiechle FL. Mannitol interference in an automated serum phosphate assay. Clin Chem. 1987;33:2308-2309.
[31.] Dorman HR, Sondheimer JH, Cadnapaphornchai P. Mannitol-induced acute renal failure. Medicine. 1990;69:153-159.
[32.] Rabetoy GM, Fredericks MR, Hostettler CF. Where the kidney is concerned, how much mannitol is too much? Ann Pharmacother. 1993;27:25-28.
[33.] Weaver A, Sica DA. Mannitol-induced acute renal failure. Nephron. 1987; 45:233-235.
[34.] Whelan TV, Bacon ME, Madden M, et al. Acute renal failure associated with mannitol intoxication. Arch Intern Med. 1984;144:2053-2055.
[35.] Goldwasser P, Fotino S. Acute renal failure following massive mannitol infusion. Arch Intern Med. 1984; 144:2214-2216.
[36.] Rello J, Triginer C, Sanchez AN, et al. Acute renal failure following massive mannitol infusion. Nephron. 1989;53:377-378.
[37.] Jackson EV, Wiese J, Sigal B, et al. Effects of crystalloid solutions on circulating lactate concentrations, I: implications for the proper handling of blood specimens obtained from critically ill patients. Crit Care Med. 1997;25:1840-1846.
[38.] Wiese J, Didwania A, Kerzner R, et al. Use of different anticoagulants in test tubes for analysis of blood lactate concentrations, II: implications for the proper handling of blood specimens obtained from critically ill patients. Crit Care Med. 1997;25:1847-1850.
[39.] Didwania A, Miller JA, Kassel D, et al. Effect of intravenous lactated Ringer's solution infusion on the circulating lactate concentration, III: results of a prospective, randomized, double-blind, placebo-controlled trial. Crit Care Med. 1997;25:1851-1854.
[40.] Hills L, Azurin G, Wang X, et al. Glutathione as an interferent in near patient whole blood glucose devices. In: Proceedings of the 17th International Symposium of the International Federation of Clinical Chemistry. Madison, Wis: Omnipress; 1998:207-219.
[41.] Windholz M, ed. The Merck Index. Rahway, NJ: Merck; 1999.
[42.] Young MW, Hung HL, Chan KH. Mannitol reduces plasma hydrogen peroxide free radical in patients undergoing coronary artery bypass graft surgery. Ma Tsui Hsueh Tsa Chi. 1992;30:65-70.
[43.] England MD, Cavorochi MC, O'Brien JF, et al. Influence of antioxidants (mannitol and allopurinol) on oxygen free radical generation during and after cardiopulmonary bypass. Circulation. 1986;74(suppl 3):111-113.
[44.] Gillbe CE, Sage FJ, Gutteridge JMC. Mannitol: molecule magnifique or a case of radical misinterpretation. Free Radic Res. 1996;24:1-7.
[45.] Wandrup JH, Walcker K. New Improvements of Stat-Lactate Measurements on Whole Blood; Clinical and Analytical Aspects. Westlake, Ohio: Radiometer America; 1997. Technical publication AS123.
[46.] Vinet B, Letellier G. The in vitro effect of drugs on biochemical parameters determined by a SMAC system. Clin Biochem. 1977;10:47-51.
[47.] Siest G, Appel W, Blijenberg GB, et al. Drug interference in clinical chemistry: studies on ascorbic acid. J Clin Chem Clin Biochem. 1978;16:103-110.
[48.] Rindone JP, Sloane EP. Cyanide toxicity from sodium nitroprusside: risks and management. Ann Pharmacother. 1992;26:515-519.
[49.] Randell EW, St. Louis P. Interference in glucose and other clinical chemistry assays by thiocyanale and cyanide in a patient treated with nitroprusside. Clin Chem. 1996;42:449-453.
[50.] Sazama K, Robertson EA, Chester RA. Is antiglycolysis required for routine glucose analysis? Clin Chem. 1979;25:2038-2039.
[51.] Chan AYW, Swaminathan R, Cockram CS. Effectiveness of sodium fluoride as a preservative of glucose in blood. Clin Chem. 1989;35:315-317.
[52.] Sacks DB. Carbohydrates. In: Burtis CA, Ashwood ER, eds. Tietz Textbook of Clinical Chemistry. 2nd ed. Philadelphia, Pa: WB Saunders; 1999:750-808.
[53.] Geyssant A, Dormois D, Barthelemy JC, et al. Lactate determination with the lactate analyser LA 640: a critical study. Scand J Clin Lab Invest. 1985;45: 145-149.
[54.] Wandrup J, Tvede K, Grinsted J, et al. "Stat" measurements of L-lactate in whole blood and cerebrospinal fluid assessed. Clin Chem. 1989;35:1740-1743.
[55.] Toffaletti JG, Hammes ME, Gray R, et al. Lactate measured in diluted and undiluted whole blood and plasma: comparison of methods and effect of hematocrit. Clin Chem. 1992;38:2430-2434.
Accepted for publication January 12, 2000. From Medical Pathology and clinical Chemistry, University of California, Davis, School of Medicine.
Presented at the Oak Ridge Conference of the America Association of Clinical Chemistry In San Jose, Calif, on April 23, 1999.
Reprints: Gerald J. Kost, MD, PhD, Medical Pathology and Biomedical Engineering, 3453 Tupper Hall, School of Medicine, University of California, Davis, CA 95616.
|Printer friendly Cite/link Email Feedback|
|Author:||Kost, Gerald J.; Nguyen, Tam H.; Tang, Zuping|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Date:||Aug 1, 2000|
|Previous Article:||Quality Control of Test Systems Waived by the Clinical Laboratory Improvement Amendments of 1988.|
|Next Article:||Effects of Different Hematocrit Levels on Glucose Measurements With Handheld Meters for Point-of-Care Testing.|