Validation of the i-STAT®1 Analyzer for Postmortem Vitreous Humor Electrolytes and Glucose Analysis

Validation of the i-STAT®1 Analyzer for Postmortem Vitreous Humor Electrolytes and Glucose Analysis Abstract The analytical value of vitreous humor as a specimen in postmortem forensic toxicology has been known for some time. Numerous medical examiner laboratories outsource the analysis of this important specimen for electrolyte and glucose measurements. This can be both time-consuming and costly. The utility of the i-STAT®1 medical device to measure electrolytes and glucose in whole blood samples has been demonstrated for over two decades in a clinical setting through single-use disposable cartridges that introduce samples to the i-STAT®1. Different cartridge types allow for the analysis of various analytes including sodium, potassium, chloride, creatinine, urea nitrogen and glucose. With only 100 μL of sample, results are obtained in under 4 min. In this study, we utilized the i-STAT®1 using an alternative specimen matrix, postmortem vitreous humor and quantitatively determined the validity and reliability of the instrument for this purpose. Acceptable criterion was used for each test as suggested by the Scientific Working Group for Forensic Toxicology. All analytes of interest, except creatinine, demonstrated a percent error < ±10% for both accuracy and precision studies. Drug interference and stability studies were performed with many of the analytes demonstrating a percent error < ±20%. Throughout drug interference and stability studies, all analytes of interest were detectable except for potassium, which gave inconclusive results. Significant interference with commonly found drugs were shown for creatinine and chloride but must be evaluated carefully. Volume additions to ethanol spiked samples caused significant interference for all analytes and is considered a limitation for this method of analysis that requires additional studies. As vitreous humor continues to be used in forensic medicine to aid in diagnostic interpretation, the i-STAT®1 has the potential to give accurate results in a timely and cost-effective manner. vitreous humor, electrolytes, i-STAT®1, postmortem, chemical analyzer Introduction The chemical analysis of vitreous humor has long been utilized in conjunction with autopsies to assist the medical examiner in determining cause of death (1–5). When no other cause can be found, vitreous humor can give examiners a clue; deviations from clinical ranges for electrolytes and glucose can be indicative of significant clinical causes (4, 6). Vitreous humor is the semi-clear liquid-gel housed between the lens and the retina of the eye, giving the eye its shape. Vitreous humor is more protected from autolysis and putrefaction compared to other body fluids upon death (5, 7, 8). The vitreous is made up of 99% water; however the viscosity of vitreous is roughly 4.20 (mPa), more viscous than water (1.00 mPa) (9, 10). The thickness of this fluid is due to type II collagen fibers, as well as electrolytes, glucose and proteins (11). For decades, forensic science has utilized different techniques for the quantitation of electrolytes and glucose in vitreous humor. However, due to the high viscosity of vitreous humor, each technique requires pre-analytical treatment to enhance accuracy and precision of measured analytes. In a study by Blana et al., sodium, potassium and chloride were quantitated by flame photometry while calcium, glucose, lactate, urea and creatinine were quantitated by the Ion-Selective Electrode (ISE) method (7). Although flame photometry and ISE techniques are fast and reliable, pre-analytical treatment was still needed for improved accuracy and precision, causing an increase in analysis time. Furthermore, these instrumental techniques are infrequently available to forensic laboratories. Different pre-analytical treatments including heating, centrifugation and liquefying agents had been examined to determine the most effective technique. Nonetheless, pre-analytical treatments require time and can increase inaccuracy and imprecision (7, 12). The i-STAT®1 is a handheld point-of-care device created in 1983 by Abbott Laboratories for the analysis of whole blood at a patient’s bedside (13). Single-use disposable cartridges introduce 100 μL of sample to the i-STAT®1 handheld analyzer without the need for pre-analytical treatment. The i-STAT®1 not only uses a smaller amount of sample, but also has a shorter analysis time, as pre-analytical treatment is not required. As a result, the i-STAT®1 can decrease systematic errors due to pre-analytical treatment. This research was designed to establish the validity and reliability of the Abbot Point of Care i-STAT®1 handheld device for the analysis of glucose and electrolytes in postmortem vitreous humor. Validation criteria were followed from the Scientific Working Group for Forensic Toxicology (SWGTOX) (14). The authors assessed accuracy and precision, interference with drugs commonly found in postmortem vitreous humor, cartridge specimen stability at different time periods outside refrigeration, and the effects of shaking the specimen prior to analysis on the following analytes: sodium, potassium, chloride, glucose, vitreous urea nitrogen (VUN) and creatinine. Experimental Apparatus An electrochemical device, the i-STAT®1 utilizes three sensors: potentiometric, conductometric and amperometric. Signals from these sensors are used to determine analyte concentrations, cell concentrations and/or coagulation time, depending on the cartridge type. Through a mechanical process, the electrical internal conductor of the analyzer contacts the electrode of a cartridge and senses the potential(s) generated from analyzer/cartridge interaction. Conductometric sensors measure the alternating current generated between the analyzer/cartridge interaction. Potentiometric and amperometric sensors are forms of Ion-Selective Electrodes (ISE) that measure specific ion concentrations. Amperometric electrodes measure electrical changes caused by oxidation-reduction reactions. The analytes of interest were tested using two types of i-STAT®1 cartridges, CHEM8+and G. CHEM8+ cartridges test for sodium, potassium, chloride, total carbon dioxide, anion gap, ionized calcium, glucose, urea nitrogen, creatinine and hematocrit/calculated hemoglobin. The G cartridges test for glucose concentrations only. Cartridges are single-use, as the analyzer must detect movement of the sample across the sensors. Analyzed cartridge samples cannot be reanalyzed thus additional vitreous sample must be used. Due to continuous manufacturing process changes to the i-STAT®1 cartridges, Abbott Laboratories releases software (CLEW) updates twice a year to re-establish quality control reference ranges and to incorporate improvements to the internal quality monitoring system (15). An internal and an external Electronic Simulator performed quality control checks for the analyzer/cartridge signal-reading function. The Internal Electronic Simulator was automatically run before each analysis. The External Simulator was inserted into the analyzer, before each day’s analysis, mimicking the electrochemical signals produced by cartridge sensors and verifying the analyzer’s ability to interpret sensitive and precise measurements of voltage, resistance and electrical current from the cartridge. Quality controls Quality controls used throughout this study were aqueous fluids made by Abbott Laboratories. Quality controls were analyzed at the beginning of each day to assure the proper function of cartridges. Abbott Laboratories manufactures a five-level set of quality controls (Calibration Verification) that spans the reportable ranges of the six analytes (sodium, potassium, chloride, glucose, urea nitrogen and creatinine) (Table I). The quality controls do not contain any human serum. However, each quality control contains buffer and preservatives; the composition of the buffer and preservatives are proprietary. The quality controls were stored in refrigeration (−3.0–4.5°C). Table I. Instrument and calibration verification liquid quality control ranges. Analyte (units)  Instrument reportable rangea,b  Level 1 range and mean  Level 3 range and mean  Level 5 range and mean  Na mmol/L  100–180  95.2–102.2  126.3–134.3  168.0–180.0  98.7  130.3  174  K mmol/L  2.0–9.0  2.09–2.61  3.51–4.09  7.18–8.24  2.35  3.80  7.71  Cl mmol/L  65–140  60.7–67.7  87.7–97.9  117.2–132.8  64.2  92.8  125  Glu mg/dL  20–700  506.9–636.9  116.1–142.5  22.1–36.5  571.9  129.3  29.3  BUN mg/dL  3–140  102.8–130.4  8.3–13.5  1.7–5.9  116.6  10.9  3.8  Crea mg/dL  0.2–200  11.58–16.62  0.71–1.43  0.0–0.42  14.10  1.07  0.06  Analyte (units)  Instrument reportable rangea,b  Level 1 range and mean  Level 3 range and mean  Level 5 range and mean  Na mmol/L  100–180  95.2–102.2  126.3–134.3  168.0–180.0  98.7  130.3  174  K mmol/L  2.0–9.0  2.09–2.61  3.51–4.09  7.18–8.24  2.35  3.80  7.71  Cl mmol/L  65–140  60.7–67.7  87.7–97.9  117.2–132.8  64.2  92.8  125  Glu mg/dL  20–700  506.9–636.9  116.1–142.5  22.1–36.5  571.9  129.3  29.3  BUN mg/dL  3–140  102.8–130.4  8.3–13.5  1.7–5.9  116.6  10.9  3.8  Crea mg/dL  0.2–200  11.58–16.62  0.71–1.43  0.0–0.42  14.10  1.07  0.06  aReportable range is the range of values in which the instrument has been valid. bSymbol Technologies Corporation. i-STAT®1 System Manual. US Patent 5,532,469, 7 March 2013. Sample preparation Vitreous samples used for this project were samples that had exceeded the Los Angeles County Department of Medical Examiner-Coroner retention policy of one year. As a true blank matrix was not viable, ten vitreous samples were removed from long-term storage (5.0–7.2°C) and pooled. Pooled samples, previously analyzed, contained oxycodone, ethanol, trazodone, salicylate and metformin. Fortified and controlled samples were prepared from this pool for all studies. Samples were stored at −2.0–3.0°C throughout the study. No preservatives were used for the collection and storage of vitreous humor. All fortified and control samples were taken out of refrigeration for 30 min to allow for room temperature acclimation (depending on study), per recommendation by Abbott Laboratories. Fortified and control samples were subsequently shaken gently before 100 μL was added to a single-use cartridge using a 25 G7/8 disposable syringe. Analytical performance Validation consisted of determining the accuracy, precision, interference and stability of the method. All validation studies were performed on both G and CHEM8+ cartridges. The maximum acceptable criterion for all tests was ±20% (14), as ±20% was the highest total error believed to still produce reliable results (16). All interference and stability studies were performed by comparing fortified pooled vitreous samples against their non-fortified counterparts. Three quality control concentrations (Calibration Verification Levels 1, 3 and 5, Abbott Laboratories) spanning the reportable range of the six analytes were each analyzed in triplicate on five separate days to assess accuracy and precision. A one-way ANOVA statistical analysis was performed to calculate within-run and between-run precision. Nine pooled samples were fortified by the addition of methanolic drug and metabolite stock solutions and brought up to 1 mL not exceeding 10% volume displacement with stock solution. All fortified samples were analyzed in duplicated and compared to non-fortified controls (Table II). A 1:1 dilution of 0.394% standard stock solution of ethanol and pooled vitreous was used to create a fortified 0.197% ethanol sample. This dilution of vitreous humor equated to a 50% volume displacement and required the use of a control with the same vitreous humor dilution factor of 1:1 made by water and pooled vitreous. Table II. Pooled samples spiked with different drug groups. Sample  Drugs and metabolites  Concentration  1  Codeine, Morphine, Hydrocodone, Hydromorphone, Oxycodone, Oxymorphone, Buprenorphine, Fentanyl  0.75 μg/mL  6-Monoacetylmorphine and Desomorphine  0.20 μg/mL  2  Cocaine, Cocaethylene, Benzoylecgonine  0.75 μg/mL  Levamisole  0.25 μg/mL  3  d-Amphetamine and d-Methamphetamine  1.0 μg/mL  Sympathomimetic Amine Mix (Phentermine, Ephedrine, Pseudoephedrine, Methcathinone, Mephedrone, 3,4-Methylenedioxyamphetamine, 3,4-Methylenedioxymethamphetamine, 1-Benzylpiperazine and 3-Trifluoromethylphenylpiperazine)  0.25 μg/mL  4  Basics Drug Mix (Bupropion, Norfluoxetine, Fluoxetine, Lidocaine, Doxylamine, Metoprolol, Dextromethorphan, Amitriptyline, Nortriptyline, Cyclobenzaprine, Norchlorcyclizine, Norsertraline, Sertraline, Duloxetine, Paroxetine, Zolpidem, Haloperidol, Quetiapine, Buspirone, Trazodone, Benztropine, Chlorpheniramine, Citalopram, Norcitalopram, Diltiazem, Diphenhydramine, Flurazepam, Hydroxyzine, Ketamine, Meperidine, Methadone, Mirtazapine, N-desmethyltramadol, Normeperidine, Olanzapine, Promethazine, Tramadol and Verapamil)  0.50 μg/mL  5  Benzodiazepine Low Mix (Alprazolam, 7-Aminoclonazepam, Clonazepam, Estazolam, Lorazepam, Midazolam and Triazolam)  50 ng/mL  Benzodiazepine High Mix (Diazepam, Nordazepam, Flurazepam, Chlordiazepoxide, Oxazepam and Temazepam)  6  Barbiturates and Phenytoin  15 μg/mL  7  Acid/Neutrals Mix (Ibuprofen, Meprobamate, Caffeine, Antipyrine, Carisoprodol, Methocarbamol, Methaqualone, Metaxalone, Topiramate, Procainamide, Primidone, Carbamazepine, Oxcarbazepine, 10-OH-Carbazepine, Lamotrigine, Modafinil and Naproxen)  10 μg/mL  Amantadine, Memantine, Nicotine and Cotinine  1 μg/mL  Valproic acid and Levetiracetam  20 μg/mL  8  Phencyclidine  0.1 μg/mL  9  Ethanol  0.197%  Sample  Drugs and metabolites  Concentration  1  Codeine, Morphine, Hydrocodone, Hydromorphone, Oxycodone, Oxymorphone, Buprenorphine, Fentanyl  0.75 μg/mL  6-Monoacetylmorphine and Desomorphine  0.20 μg/mL  2  Cocaine, Cocaethylene, Benzoylecgonine  0.75 μg/mL  Levamisole  0.25 μg/mL  3  d-Amphetamine and d-Methamphetamine  1.0 μg/mL  Sympathomimetic Amine Mix (Phentermine, Ephedrine, Pseudoephedrine, Methcathinone, Mephedrone, 3,4-Methylenedioxyamphetamine, 3,4-Methylenedioxymethamphetamine, 1-Benzylpiperazine and 3-Trifluoromethylphenylpiperazine)  0.25 μg/mL  4  Basics Drug Mix (Bupropion, Norfluoxetine, Fluoxetine, Lidocaine, Doxylamine, Metoprolol, Dextromethorphan, Amitriptyline, Nortriptyline, Cyclobenzaprine, Norchlorcyclizine, Norsertraline, Sertraline, Duloxetine, Paroxetine, Zolpidem, Haloperidol, Quetiapine, Buspirone, Trazodone, Benztropine, Chlorpheniramine, Citalopram, Norcitalopram, Diltiazem, Diphenhydramine, Flurazepam, Hydroxyzine, Ketamine, Meperidine, Methadone, Mirtazapine, N-desmethyltramadol, Normeperidine, Olanzapine, Promethazine, Tramadol and Verapamil)  0.50 μg/mL  5  Benzodiazepine Low Mix (Alprazolam, 7-Aminoclonazepam, Clonazepam, Estazolam, Lorazepam, Midazolam and Triazolam)  50 ng/mL  Benzodiazepine High Mix (Diazepam, Nordazepam, Flurazepam, Chlordiazepoxide, Oxazepam and Temazepam)  6  Barbiturates and Phenytoin  15 μg/mL  7  Acid/Neutrals Mix (Ibuprofen, Meprobamate, Caffeine, Antipyrine, Carisoprodol, Methocarbamol, Methaqualone, Metaxalone, Topiramate, Procainamide, Primidone, Carbamazepine, Oxcarbazepine, 10-OH-Carbazepine, Lamotrigine, Modafinil and Naproxen)  10 μg/mL  Amantadine, Memantine, Nicotine and Cotinine  1 μg/mL  Valproic acid and Levetiracetam  20 μg/mL  8  Phencyclidine  0.1 μg/mL  9  Ethanol  0.197%  A control sample was analyzed in triplicate to establish a baseline for stability studies and was compared to experimental data. Duplicate control samples were left out of refrigeration in capped red top vials for 30 min, per manufacturer’s recommendations, to assess for stability. These were compared to capped samples analyzed at time zero, 30 min (baseline), and 2 h. The manufacturer’s procedure states that samples must be shaken before uptake into the syringe and dispensed into the cartridge well. Unshaken samples were compared to samples gently shaken for 10–15 s to assess stability of the analytes. Calibration, carry over and limits of detection and quantitation were not applicable due to i-STAT®1 design and the use of single-use cartridges. The i-STAT®1 is calibrated by the manufacturer; therefore, in-laboratory calibrations were not applicable. Manufactured calibrations were verified through liquid quality controls. Analytical case results that were above or below the manufacturer’s reportable range were displayed on the analyzer screen as “>[max value]” or “<[min value]”. Single-use disposable cartridges introduce samples to the i-STAT®1 handheld analyzer. As a result, carryover was not a concern. However, attention should be made to not overfill the cartridge. While this does not lead to contamination issues, it can lead to poor analyzer results. Abbott Laboratories established the limit of detection and quantitation. Results Accuracy and precision At Calibration Verification Levels 1, 3 and 5 all analytes demonstrated a percent error < ±10% except creatinine (Table III). Calibration Verification Level 5 exhibited a mean creatinine value of 0.06, however, the instrument reportable range for creatinine spans from 0.2 to 20.0 mg/dL. Therefore, creatinine at Calibration Verification Level 5 was considered by the authors to yield inconclusive results. The low percent error for all analytes agreed with the closeness of results to the true value reported by the manufacturer. Table III. Percent error for bias demonstrated by corresponding analytes. Analyte  Level 1  Level 3  Level 5  Na  0.236%  0.281%  0.07%  K  1.28%  0.175%  0.476%  Cl  −1.66%  −0.287%  0.0%  Glu  4.82%  −5.44%  −7.62%  VUN  1.09%  0.917%  −7.02%  Crea  7.33%  −4.05%  < >a  Glu (G)b  3.43%  0.490%  −0.796%  Analyte  Level 1  Level 3  Level 5  Na  0.236%  0.281%  0.07%  K  1.28%  0.175%  0.476%  Cl  −1.66%  −0.287%  0.0%  Glu  4.82%  −5.44%  −7.62%  VUN  1.09%  0.917%  −7.02%  Crea  7.33%  −4.05%  < >a  Glu (G)b  3.43%  0.490%  −0.796%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. All analytes had a within-run coefficient of variation (CV) <15% (Table IV) and a between-run CV <15% (Table V), demonstrating adequate precision at all levels over five different runs. Table IV. Within-run precision. Analytes  Level 1  Level 3  Level 5  Na  0.639%  0.625%  0.593%  K  1.88%  0.678%  0.471%  Cl  1.16%  1.18%  1.32%  Glu  1.33%  0.597%  3.82%  VUN  1.33%  0.00%  14.6%  Crea  3.14%  2.51%  < >a  Glu (G)b  0.782%  0.199%  1.54%  Analytes  Level 1  Level 3  Level 5  Na  0.639%  0.625%  0.593%  K  1.88%  0.678%  0.471%  Cl  1.16%  1.18%  1.32%  Glu  1.33%  0.597%  3.82%  VUN  1.33%  0.00%  14.6%  Crea  3.14%  2.51%  < >a  Glu (G)b  0.782%  0.199%  1.54%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. Table V. Between-run precision. Analyte  Level 1  Level 3  Level 5  Na  0.593%  0.625%  0.611%  K  1.72%  0.678%  0.871%  Cl  1.35%  1.05%  1.19%  Glu  1.90%  0.940%  6.98%  VUN  1.44%  0.00%  14.6%  Crea  5.09%  4.70%  < >a  Glu (G)b  0.982%  0.579%  2.54%  Analyte  Level 1  Level 3  Level 5  Na  0.593%  0.625%  0.611%  K  1.72%  0.678%  0.871%  Cl  1.35%  1.05%  1.19%  Glu  1.90%  0.940%  6.98%  VUN  1.44%  0.00%  14.6%  Crea  5.09%  4.70%  < >a  Glu (G)b  0.982%  0.579%  2.54%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. Drug interferences Most samples fortified with commonly encountered drugs exhibited < ±20% deviations from their respective controls. However, there was substantial interference with creatinine in many of the fortified samples. Several fortified samples also exhibited substantial interference with chloride (Table VI). Potassium was consistently out of the instrument’s reportable range and could not be quantitated. Analytes that had a percent error < ±20% only had regular variations with no interference in the fortified samples. Table VI. Percent error due to common drug interferences. Vial name  Na  K  Cl  Glu  VUN  Crea  Glu (G)  Opioid  5.11%  < >  Inconclusive  7.1%  9.62%  −33.3%a  −4.42%  Coke  4.74%  < >  Inconclusive  12.2%  9.62%  −38.9%  −4.42%  Amine  3.99%  < >  −12.4%  14.7%  7.69%  −27.8%  −1.77%  Basic  5.49%  < >  −5.31%  7.14%  7.69%  −11.1%  −1.77%  Benzo  4.36%  < >  Inconclusive  16.0%  5.77%  −66.7%  −3.10%  Barb  6.61%  < >  Inconclusive  14.7%  9.62%  −66.7%  −4.42%  Acid  0.623%  < >  −3.10%  5.88%  −1.92%  −5.56%  −1.77%  PCP  5.11%  < >  Inconclusive  18.5%  5.77%  −50.0%  −5.75%  EtOH  < >b  Inconclusive  Inconclusive  < >  Inconclusive  50.0%  Inconclusive  Vial name  Na  K  Cl  Glu  VUN  Crea  Glu (G)  Opioid  5.11%  < >  Inconclusive  7.1%  9.62%  −33.3%a  −4.42%  Coke  4.74%  < >  Inconclusive  12.2%  9.62%  −38.9%  −4.42%  Amine  3.99%  < >  −12.4%  14.7%  7.69%  −27.8%  −1.77%  Basic  5.49%  < >  −5.31%  7.14%  7.69%  −11.1%  −1.77%  Benzo  4.36%  < >  Inconclusive  16.0%  5.77%  −66.7%  −3.10%  Barb  6.61%  < >  Inconclusive  14.7%  9.62%  −66.7%  −4.42%  Acid  0.623%  < >  −3.10%  5.88%  −1.92%  −5.56%  −1.77%  PCP  5.11%  < >  Inconclusive  18.5%  5.77%  −50.0%  −5.75%  EtOH  < >b  Inconclusive  Inconclusive  < >  Inconclusive  50.0%  Inconclusive  aBold results represent unacceptable results. b“< >” represents data that are out of instrument range. For the ethanol and accompanying control samples, all analytes gave inconclusive or failed results with deviations either above the acceptable criterion, outside the manufactured reportable range or dependent on other tests that were out of instrument range (Table VI). Stability The percent error exhibited at time zero in capped sample vials for Na, Cl, Glu (Chem8+), VUN, Cre and Glu (G) were <−1, 1, 3, 2, 4 and 3%, respectively. Therefore, samples were stable when analyzed straight from the refrigerator without reaching room temperature; however, analyzes took around 5 min compared to the typical 3–4 min as samples were brought to 37°C by the automatic thermal control subsystem in the device. The percent error exhibited at 2 h out of refrigeration for Na Cl, Glu (Chem8+), VUN, Crea and Glu (G) were <1, −1, 6, −2, 1, 4 and −1%, respectively. Potassium gave inconclusive results at both time periods. Therefore, capped samples were stable out of the refrigerator for up to 2 h. The percent error exhibited for unshaken samples was <1, −1, 6, 2 and 9% for Na, Cl, Glu (Chem8+), VUN and Glu (G), respectively. Creatinine gave a percent error greater than 40% and potassium gave inconclusive results. Discussion Limitations While the i-STAT®1 device has been successfully utilized clinically using whole blood for some time, there is relatively little published information about its potential use on postmortem samples like vitreous humor. At the time of writing, the authors have identified only one published manuscript using the i-STAT®1 device for vitreous humor analysis, and there were no validation parameters presented (17). Additionally, Oles et al. (18) presented a comparison of the i-STAT®1 to the Synchron LX 20 high throughput clinical laboratory analyzer from freshly procured vitreous humor at autopsy and found comparable results for glucose and creatinine, but differences for sodium, chloride and urea nitrogen. However, it is unclear if this work attempted validation criteria or considered potential contaminating species present in vitreous humor. Although there was substantial interference for creatinine and chloride in the drug fortified samples, research by Coe (2) emphasized the stability of postmortem vitreous creatinine. Deviations in clinical ranges for creatinine could indicate dehydration, renal failure or uremia (4). However, we cannot rule out the possibility that the dilution of the fortified samples with drug stock solution contributed to these results because volume matched controlled blanks were not prepared. However, if the results were solely due to volumetric differences we would have expected more unacceptable or inconclusive results for the other analytes measured, especially sodium. It is, therefore, likely that interference due to drug groups being present in vitreous humor is minimal and that altering sample volume by drug spiking additions or other solutions is a noteworthy limitation in the analysis of postmortem vitreous humor (4, 6). An elevated ethanol concentration of 0.197% as well as volume displacement matched controls significantly interfered with all analytes. However, some private laboratories use similar ion-selective electrode techniques with no ethanol interference (19). Therefore, the inconclusive and out-of-range results shown are likely best explained by the dilution of vitreous humor. Based on the overall interference studies, inconclusive or out-of-range results may become apparent in analysis at or near a 10% volume dilution displacement of vitreous humor. Future studies will include performing parallel analyzes of vitreous specimens between the i-STAT®1 and an independent laboratory. Such a study could offer information on which method provides the greatest turnaround time and cost effectiveness. Conclusion Overall, the i-STAT®1 handheld device is accurate and precise for the analysis of postmortem vitreous humor. There were, however, potential drug interferences when analyzing for chloride and creatinine, as both analytes had either uncharacteristic signals or percent error greater than ±20%. These results, including those for the ethanol fortified samples, also indicate sensitivity to volumetric changes in the sample preparation and highlight an inherent difficulty in controlling for endogenous analytes using this matrix. In order to maximize analyte stability in matrix, samples should be properly stored in refrigeration (−2.0–3.0°C) and shaken gently before analysis. Overall samples were stable up to 2 h out of refrigeration. However, research by Gagajewski et al. (20) supports temperature dependent changes in vitreous humor’s sodium, potassium and chloride concentrations depending on storage conditions. When unshaken, most analytes present had a percent error <10%, except for creatinine that had a percent error greater than 40%. This result could indicate creatinine’s sensitivity to sample shaking. The i-STAT®1 has been used in the medical field for over two decades for whole blood chemical analysis. This study has shown that the i-STAT®1 can also be used to analyze vitreous humor to aide with the cause of death determination, though additional testing must be performed to ensure the i-STAT®1’s reliability and cost effectiveness for in-house testing. Acknowledgments The authors would like to acknowledge the Los Angeles County Department of Medical Examiner-Coroner Office for providing the resources needed to complete this project. The authors would also like to thank the staff at the Los Angeles Department of Medical Examiner-Coroner Office for their support, especially Jessica Gadway for her guidance in quality assurance procedures. References 1 Coe, J.I. ( 1974) Postmortem chemistry: practical considerations and a review of the literature. Journal of Forensic Science , 19, 13– 32. 2 Coe, J.I. ( 1993) Postmortem chemistry update emphasis on forensic application. American Journal of Forensic Medicine and Pathology , 14, 91– 117. Google Scholar CrossRef Search ADS PubMed  3 Palmiere, C., Mangin, P. ( 2012) Postmortem chemistry update part Ι. International Journal of Legal Medicine , 126, 187– 198. Google Scholar CrossRef Search ADS PubMed  4 Rose, K.L., Collins, K.A. ( 2008) Vitreous postmortem chemical analysis. College of American Pathologists. http://www.cap.org/apps/docs/newspath/0812/vitreous_postmortem_chemical_analysis.pdf (accessed July 7, 2017). 5 Coe, J.I. ( 1969) Postmortem chemistries on human vitreous humor. American Journal of Clinical Pathology , 51, 741– 750. Google Scholar CrossRef Search ADS PubMed  6 Coe, J.I., Apple, F.S. ( 1985) Variations in vitreous humor chemical values as a result of instrumentation. Journal of Forensic Science , 30, 828– 835. Google Scholar CrossRef Search ADS   7 Blana, S.A., Muβhoff, F., Hoeller, T., Fimmers, R., Madea, B. ( 2011) Variations in vitreous humor chemical values as a result of pre-analytical treatment. Forensic Science International , 210, 263– 270. Google Scholar CrossRef Search ADS PubMed  8 Madea, B., Musshoff, F. ( 2007) Postmortem biochemistry. Forensic Science International , 165, 165– 171. Google Scholar CrossRef Search ADS PubMed  9 Aretz, S., Krohne, T.U., Kammerer, K., Warnken, U., Hotz-Wagenblatt, A., Bergmann, M., et al.  . ( 2013) In-depth mass spectrometric mapping of the human vitreous proteome. Proteome Science , 11, 1– 10. Google Scholar CrossRef Search ADS PubMed  10 Swindle, K.E., Ravi, N. ( 2007) Recent advances in polymeric vitreous substitutes. Expert Review Ophthalmology , 2, 255– 265. Google Scholar CrossRef Search ADS   11 Murthy, K.R., Goel, R., Subbannayya, Y., Jacob, H., Murthy, P.R., Srinivas Manda, S., et al.  . ( 2014) Proteomic analysis of human vitreous humor. Clinical Proteomics , 11, 1– 11. Google Scholar CrossRef Search ADS PubMed  12 Garg, U., Althahabi, R., Amirahmadi, V., Brod, M., Blanchard, C., Young, T. ( 2004) Hyauronidase as aliquefying agent for chemical analysis of vitreous fluid. Journal of Forensic Science , 49, 1– 4. Google Scholar CrossRef Search ADS   13 Chin, C.D., Linder, V., Sia, S.K. ( 2012) Commercialization of microfluidic point-of-Care diagnostic devices. Lab on a chip , 12, 2118– 2134. Google Scholar CrossRef Search ADS PubMed  14 SWGTOX. ( 2013) Scientific working group for forensic toxicology (SWGTOX) standard practices for method validation in forensic toxicology. Journal of Analytical Toxicology , 37, 452– 474. CrossRef Search ADS PubMed  15 (2017) i-STAT®1 Product Update Software and Cartridge Test Information: Update May 2017. https://www.pointofcare.abbott/download?docUri=/technical-library/static-assets/technical-documentation/731668-00H.pdf (accessed 7 July 2017). 16 Shah, V.P., Midha, K.K., Findlay, J.A., Hill, H.M., Hules, J.D., McGilveray, I.J., et al.  . ( 2000) Bioanalytical method validation- a revisit with a decade progress. Pharmaceutical Research , 17, 1551– 1557. Google Scholar CrossRef Search ADS PubMed  17 Antonides, H., Marinetti, L. ( 2011) Ethanol production in a postmortem urine sample. Journal of Analytical Toxicology , 35, 516– 518. Google Scholar CrossRef Search ADS PubMed  18 Oles, M.A., Juhascik, M.P., Jenkins, A.J. Evaluation of the i-STAT 1 handheld analyzer for postmortem vitreous humor chemistry analysis. Abstract Presented at the Society for Forensic Toxicology Meeting, San Francisco, CA, 2011. 19 Caplan, Y.H., Levine, B. ( 1990) Vitreous humor in the evaluation of postmortem blood ethanol concentrations. Journal of Analytical Toxicology , 14, 305– 307. Google Scholar CrossRef Search ADS PubMed  20 Gagajewski, A., Murakami, M.A., Kloss, J., Edstrom, M., Hillyer, M., Peterson, G.F., et al.  . ( 2004) Measurement of chemical analytes in vitreous humor: stability and precision studies. Journal of Forensic Science , 49, 1– 4. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Analytical Toxicology Oxford University Press

Validation of the i-STAT®1 Analyzer for Postmortem Vitreous Humor Electrolytes and Glucose Analysis

Loading next page...
 
/lp/ou_press/validation-of-the-i-stat-1-analyzer-for-postmortem-vitreous-humor-L3oJl3HFq5
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0146-4760
eISSN
1945-2403
D.O.I.
10.1093/jat/bkx084
Publisher site
See Article on Publisher Site

Abstract

Abstract The analytical value of vitreous humor as a specimen in postmortem forensic toxicology has been known for some time. Numerous medical examiner laboratories outsource the analysis of this important specimen for electrolyte and glucose measurements. This can be both time-consuming and costly. The utility of the i-STAT®1 medical device to measure electrolytes and glucose in whole blood samples has been demonstrated for over two decades in a clinical setting through single-use disposable cartridges that introduce samples to the i-STAT®1. Different cartridge types allow for the analysis of various analytes including sodium, potassium, chloride, creatinine, urea nitrogen and glucose. With only 100 μL of sample, results are obtained in under 4 min. In this study, we utilized the i-STAT®1 using an alternative specimen matrix, postmortem vitreous humor and quantitatively determined the validity and reliability of the instrument for this purpose. Acceptable criterion was used for each test as suggested by the Scientific Working Group for Forensic Toxicology. All analytes of interest, except creatinine, demonstrated a percent error < ±10% for both accuracy and precision studies. Drug interference and stability studies were performed with many of the analytes demonstrating a percent error < ±20%. Throughout drug interference and stability studies, all analytes of interest were detectable except for potassium, which gave inconclusive results. Significant interference with commonly found drugs were shown for creatinine and chloride but must be evaluated carefully. Volume additions to ethanol spiked samples caused significant interference for all analytes and is considered a limitation for this method of analysis that requires additional studies. As vitreous humor continues to be used in forensic medicine to aid in diagnostic interpretation, the i-STAT®1 has the potential to give accurate results in a timely and cost-effective manner. vitreous humor, electrolytes, i-STAT®1, postmortem, chemical analyzer Introduction The chemical analysis of vitreous humor has long been utilized in conjunction with autopsies to assist the medical examiner in determining cause of death (1–5). When no other cause can be found, vitreous humor can give examiners a clue; deviations from clinical ranges for electrolytes and glucose can be indicative of significant clinical causes (4, 6). Vitreous humor is the semi-clear liquid-gel housed between the lens and the retina of the eye, giving the eye its shape. Vitreous humor is more protected from autolysis and putrefaction compared to other body fluids upon death (5, 7, 8). The vitreous is made up of 99% water; however the viscosity of vitreous is roughly 4.20 (mPa), more viscous than water (1.00 mPa) (9, 10). The thickness of this fluid is due to type II collagen fibers, as well as electrolytes, glucose and proteins (11). For decades, forensic science has utilized different techniques for the quantitation of electrolytes and glucose in vitreous humor. However, due to the high viscosity of vitreous humor, each technique requires pre-analytical treatment to enhance accuracy and precision of measured analytes. In a study by Blana et al., sodium, potassium and chloride were quantitated by flame photometry while calcium, glucose, lactate, urea and creatinine were quantitated by the Ion-Selective Electrode (ISE) method (7). Although flame photometry and ISE techniques are fast and reliable, pre-analytical treatment was still needed for improved accuracy and precision, causing an increase in analysis time. Furthermore, these instrumental techniques are infrequently available to forensic laboratories. Different pre-analytical treatments including heating, centrifugation and liquefying agents had been examined to determine the most effective technique. Nonetheless, pre-analytical treatments require time and can increase inaccuracy and imprecision (7, 12). The i-STAT®1 is a handheld point-of-care device created in 1983 by Abbott Laboratories for the analysis of whole blood at a patient’s bedside (13). Single-use disposable cartridges introduce 100 μL of sample to the i-STAT®1 handheld analyzer without the need for pre-analytical treatment. The i-STAT®1 not only uses a smaller amount of sample, but also has a shorter analysis time, as pre-analytical treatment is not required. As a result, the i-STAT®1 can decrease systematic errors due to pre-analytical treatment. This research was designed to establish the validity and reliability of the Abbot Point of Care i-STAT®1 handheld device for the analysis of glucose and electrolytes in postmortem vitreous humor. Validation criteria were followed from the Scientific Working Group for Forensic Toxicology (SWGTOX) (14). The authors assessed accuracy and precision, interference with drugs commonly found in postmortem vitreous humor, cartridge specimen stability at different time periods outside refrigeration, and the effects of shaking the specimen prior to analysis on the following analytes: sodium, potassium, chloride, glucose, vitreous urea nitrogen (VUN) and creatinine. Experimental Apparatus An electrochemical device, the i-STAT®1 utilizes three sensors: potentiometric, conductometric and amperometric. Signals from these sensors are used to determine analyte concentrations, cell concentrations and/or coagulation time, depending on the cartridge type. Through a mechanical process, the electrical internal conductor of the analyzer contacts the electrode of a cartridge and senses the potential(s) generated from analyzer/cartridge interaction. Conductometric sensors measure the alternating current generated between the analyzer/cartridge interaction. Potentiometric and amperometric sensors are forms of Ion-Selective Electrodes (ISE) that measure specific ion concentrations. Amperometric electrodes measure electrical changes caused by oxidation-reduction reactions. The analytes of interest were tested using two types of i-STAT®1 cartridges, CHEM8+and G. CHEM8+ cartridges test for sodium, potassium, chloride, total carbon dioxide, anion gap, ionized calcium, glucose, urea nitrogen, creatinine and hematocrit/calculated hemoglobin. The G cartridges test for glucose concentrations only. Cartridges are single-use, as the analyzer must detect movement of the sample across the sensors. Analyzed cartridge samples cannot be reanalyzed thus additional vitreous sample must be used. Due to continuous manufacturing process changes to the i-STAT®1 cartridges, Abbott Laboratories releases software (CLEW) updates twice a year to re-establish quality control reference ranges and to incorporate improvements to the internal quality monitoring system (15). An internal and an external Electronic Simulator performed quality control checks for the analyzer/cartridge signal-reading function. The Internal Electronic Simulator was automatically run before each analysis. The External Simulator was inserted into the analyzer, before each day’s analysis, mimicking the electrochemical signals produced by cartridge sensors and verifying the analyzer’s ability to interpret sensitive and precise measurements of voltage, resistance and electrical current from the cartridge. Quality controls Quality controls used throughout this study were aqueous fluids made by Abbott Laboratories. Quality controls were analyzed at the beginning of each day to assure the proper function of cartridges. Abbott Laboratories manufactures a five-level set of quality controls (Calibration Verification) that spans the reportable ranges of the six analytes (sodium, potassium, chloride, glucose, urea nitrogen and creatinine) (Table I). The quality controls do not contain any human serum. However, each quality control contains buffer and preservatives; the composition of the buffer and preservatives are proprietary. The quality controls were stored in refrigeration (−3.0–4.5°C). Table I. Instrument and calibration verification liquid quality control ranges. Analyte (units)  Instrument reportable rangea,b  Level 1 range and mean  Level 3 range and mean  Level 5 range and mean  Na mmol/L  100–180  95.2–102.2  126.3–134.3  168.0–180.0  98.7  130.3  174  K mmol/L  2.0–9.0  2.09–2.61  3.51–4.09  7.18–8.24  2.35  3.80  7.71  Cl mmol/L  65–140  60.7–67.7  87.7–97.9  117.2–132.8  64.2  92.8  125  Glu mg/dL  20–700  506.9–636.9  116.1–142.5  22.1–36.5  571.9  129.3  29.3  BUN mg/dL  3–140  102.8–130.4  8.3–13.5  1.7–5.9  116.6  10.9  3.8  Crea mg/dL  0.2–200  11.58–16.62  0.71–1.43  0.0–0.42  14.10  1.07  0.06  Analyte (units)  Instrument reportable rangea,b  Level 1 range and mean  Level 3 range and mean  Level 5 range and mean  Na mmol/L  100–180  95.2–102.2  126.3–134.3  168.0–180.0  98.7  130.3  174  K mmol/L  2.0–9.0  2.09–2.61  3.51–4.09  7.18–8.24  2.35  3.80  7.71  Cl mmol/L  65–140  60.7–67.7  87.7–97.9  117.2–132.8  64.2  92.8  125  Glu mg/dL  20–700  506.9–636.9  116.1–142.5  22.1–36.5  571.9  129.3  29.3  BUN mg/dL  3–140  102.8–130.4  8.3–13.5  1.7–5.9  116.6  10.9  3.8  Crea mg/dL  0.2–200  11.58–16.62  0.71–1.43  0.0–0.42  14.10  1.07  0.06  aReportable range is the range of values in which the instrument has been valid. bSymbol Technologies Corporation. i-STAT®1 System Manual. US Patent 5,532,469, 7 March 2013. Sample preparation Vitreous samples used for this project were samples that had exceeded the Los Angeles County Department of Medical Examiner-Coroner retention policy of one year. As a true blank matrix was not viable, ten vitreous samples were removed from long-term storage (5.0–7.2°C) and pooled. Pooled samples, previously analyzed, contained oxycodone, ethanol, trazodone, salicylate and metformin. Fortified and controlled samples were prepared from this pool for all studies. Samples were stored at −2.0–3.0°C throughout the study. No preservatives were used for the collection and storage of vitreous humor. All fortified and control samples were taken out of refrigeration for 30 min to allow for room temperature acclimation (depending on study), per recommendation by Abbott Laboratories. Fortified and control samples were subsequently shaken gently before 100 μL was added to a single-use cartridge using a 25 G7/8 disposable syringe. Analytical performance Validation consisted of determining the accuracy, precision, interference and stability of the method. All validation studies were performed on both G and CHEM8+ cartridges. The maximum acceptable criterion for all tests was ±20% (14), as ±20% was the highest total error believed to still produce reliable results (16). All interference and stability studies were performed by comparing fortified pooled vitreous samples against their non-fortified counterparts. Three quality control concentrations (Calibration Verification Levels 1, 3 and 5, Abbott Laboratories) spanning the reportable range of the six analytes were each analyzed in triplicate on five separate days to assess accuracy and precision. A one-way ANOVA statistical analysis was performed to calculate within-run and between-run precision. Nine pooled samples were fortified by the addition of methanolic drug and metabolite stock solutions and brought up to 1 mL not exceeding 10% volume displacement with stock solution. All fortified samples were analyzed in duplicated and compared to non-fortified controls (Table II). A 1:1 dilution of 0.394% standard stock solution of ethanol and pooled vitreous was used to create a fortified 0.197% ethanol sample. This dilution of vitreous humor equated to a 50% volume displacement and required the use of a control with the same vitreous humor dilution factor of 1:1 made by water and pooled vitreous. Table II. Pooled samples spiked with different drug groups. Sample  Drugs and metabolites  Concentration  1  Codeine, Morphine, Hydrocodone, Hydromorphone, Oxycodone, Oxymorphone, Buprenorphine, Fentanyl  0.75 μg/mL  6-Monoacetylmorphine and Desomorphine  0.20 μg/mL  2  Cocaine, Cocaethylene, Benzoylecgonine  0.75 μg/mL  Levamisole  0.25 μg/mL  3  d-Amphetamine and d-Methamphetamine  1.0 μg/mL  Sympathomimetic Amine Mix (Phentermine, Ephedrine, Pseudoephedrine, Methcathinone, Mephedrone, 3,4-Methylenedioxyamphetamine, 3,4-Methylenedioxymethamphetamine, 1-Benzylpiperazine and 3-Trifluoromethylphenylpiperazine)  0.25 μg/mL  4  Basics Drug Mix (Bupropion, Norfluoxetine, Fluoxetine, Lidocaine, Doxylamine, Metoprolol, Dextromethorphan, Amitriptyline, Nortriptyline, Cyclobenzaprine, Norchlorcyclizine, Norsertraline, Sertraline, Duloxetine, Paroxetine, Zolpidem, Haloperidol, Quetiapine, Buspirone, Trazodone, Benztropine, Chlorpheniramine, Citalopram, Norcitalopram, Diltiazem, Diphenhydramine, Flurazepam, Hydroxyzine, Ketamine, Meperidine, Methadone, Mirtazapine, N-desmethyltramadol, Normeperidine, Olanzapine, Promethazine, Tramadol and Verapamil)  0.50 μg/mL  5  Benzodiazepine Low Mix (Alprazolam, 7-Aminoclonazepam, Clonazepam, Estazolam, Lorazepam, Midazolam and Triazolam)  50 ng/mL  Benzodiazepine High Mix (Diazepam, Nordazepam, Flurazepam, Chlordiazepoxide, Oxazepam and Temazepam)  6  Barbiturates and Phenytoin  15 μg/mL  7  Acid/Neutrals Mix (Ibuprofen, Meprobamate, Caffeine, Antipyrine, Carisoprodol, Methocarbamol, Methaqualone, Metaxalone, Topiramate, Procainamide, Primidone, Carbamazepine, Oxcarbazepine, 10-OH-Carbazepine, Lamotrigine, Modafinil and Naproxen)  10 μg/mL  Amantadine, Memantine, Nicotine and Cotinine  1 μg/mL  Valproic acid and Levetiracetam  20 μg/mL  8  Phencyclidine  0.1 μg/mL  9  Ethanol  0.197%  Sample  Drugs and metabolites  Concentration  1  Codeine, Morphine, Hydrocodone, Hydromorphone, Oxycodone, Oxymorphone, Buprenorphine, Fentanyl  0.75 μg/mL  6-Monoacetylmorphine and Desomorphine  0.20 μg/mL  2  Cocaine, Cocaethylene, Benzoylecgonine  0.75 μg/mL  Levamisole  0.25 μg/mL  3  d-Amphetamine and d-Methamphetamine  1.0 μg/mL  Sympathomimetic Amine Mix (Phentermine, Ephedrine, Pseudoephedrine, Methcathinone, Mephedrone, 3,4-Methylenedioxyamphetamine, 3,4-Methylenedioxymethamphetamine, 1-Benzylpiperazine and 3-Trifluoromethylphenylpiperazine)  0.25 μg/mL  4  Basics Drug Mix (Bupropion, Norfluoxetine, Fluoxetine, Lidocaine, Doxylamine, Metoprolol, Dextromethorphan, Amitriptyline, Nortriptyline, Cyclobenzaprine, Norchlorcyclizine, Norsertraline, Sertraline, Duloxetine, Paroxetine, Zolpidem, Haloperidol, Quetiapine, Buspirone, Trazodone, Benztropine, Chlorpheniramine, Citalopram, Norcitalopram, Diltiazem, Diphenhydramine, Flurazepam, Hydroxyzine, Ketamine, Meperidine, Methadone, Mirtazapine, N-desmethyltramadol, Normeperidine, Olanzapine, Promethazine, Tramadol and Verapamil)  0.50 μg/mL  5  Benzodiazepine Low Mix (Alprazolam, 7-Aminoclonazepam, Clonazepam, Estazolam, Lorazepam, Midazolam and Triazolam)  50 ng/mL  Benzodiazepine High Mix (Diazepam, Nordazepam, Flurazepam, Chlordiazepoxide, Oxazepam and Temazepam)  6  Barbiturates and Phenytoin  15 μg/mL  7  Acid/Neutrals Mix (Ibuprofen, Meprobamate, Caffeine, Antipyrine, Carisoprodol, Methocarbamol, Methaqualone, Metaxalone, Topiramate, Procainamide, Primidone, Carbamazepine, Oxcarbazepine, 10-OH-Carbazepine, Lamotrigine, Modafinil and Naproxen)  10 μg/mL  Amantadine, Memantine, Nicotine and Cotinine  1 μg/mL  Valproic acid and Levetiracetam  20 μg/mL  8  Phencyclidine  0.1 μg/mL  9  Ethanol  0.197%  A control sample was analyzed in triplicate to establish a baseline for stability studies and was compared to experimental data. Duplicate control samples were left out of refrigeration in capped red top vials for 30 min, per manufacturer’s recommendations, to assess for stability. These were compared to capped samples analyzed at time zero, 30 min (baseline), and 2 h. The manufacturer’s procedure states that samples must be shaken before uptake into the syringe and dispensed into the cartridge well. Unshaken samples were compared to samples gently shaken for 10–15 s to assess stability of the analytes. Calibration, carry over and limits of detection and quantitation were not applicable due to i-STAT®1 design and the use of single-use cartridges. The i-STAT®1 is calibrated by the manufacturer; therefore, in-laboratory calibrations were not applicable. Manufactured calibrations were verified through liquid quality controls. Analytical case results that were above or below the manufacturer’s reportable range were displayed on the analyzer screen as “>[max value]” or “<[min value]”. Single-use disposable cartridges introduce samples to the i-STAT®1 handheld analyzer. As a result, carryover was not a concern. However, attention should be made to not overfill the cartridge. While this does not lead to contamination issues, it can lead to poor analyzer results. Abbott Laboratories established the limit of detection and quantitation. Results Accuracy and precision At Calibration Verification Levels 1, 3 and 5 all analytes demonstrated a percent error < ±10% except creatinine (Table III). Calibration Verification Level 5 exhibited a mean creatinine value of 0.06, however, the instrument reportable range for creatinine spans from 0.2 to 20.0 mg/dL. Therefore, creatinine at Calibration Verification Level 5 was considered by the authors to yield inconclusive results. The low percent error for all analytes agreed with the closeness of results to the true value reported by the manufacturer. Table III. Percent error for bias demonstrated by corresponding analytes. Analyte  Level 1  Level 3  Level 5  Na  0.236%  0.281%  0.07%  K  1.28%  0.175%  0.476%  Cl  −1.66%  −0.287%  0.0%  Glu  4.82%  −5.44%  −7.62%  VUN  1.09%  0.917%  −7.02%  Crea  7.33%  −4.05%  < >a  Glu (G)b  3.43%  0.490%  −0.796%  Analyte  Level 1  Level 3  Level 5  Na  0.236%  0.281%  0.07%  K  1.28%  0.175%  0.476%  Cl  −1.66%  −0.287%  0.0%  Glu  4.82%  −5.44%  −7.62%  VUN  1.09%  0.917%  −7.02%  Crea  7.33%  −4.05%  < >a  Glu (G)b  3.43%  0.490%  −0.796%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. All analytes had a within-run coefficient of variation (CV) <15% (Table IV) and a between-run CV <15% (Table V), demonstrating adequate precision at all levels over five different runs. Table IV. Within-run precision. Analytes  Level 1  Level 3  Level 5  Na  0.639%  0.625%  0.593%  K  1.88%  0.678%  0.471%  Cl  1.16%  1.18%  1.32%  Glu  1.33%  0.597%  3.82%  VUN  1.33%  0.00%  14.6%  Crea  3.14%  2.51%  < >a  Glu (G)b  0.782%  0.199%  1.54%  Analytes  Level 1  Level 3  Level 5  Na  0.639%  0.625%  0.593%  K  1.88%  0.678%  0.471%  Cl  1.16%  1.18%  1.32%  Glu  1.33%  0.597%  3.82%  VUN  1.33%  0.00%  14.6%  Crea  3.14%  2.51%  < >a  Glu (G)b  0.782%  0.199%  1.54%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. Table V. Between-run precision. Analyte  Level 1  Level 3  Level 5  Na  0.593%  0.625%  0.611%  K  1.72%  0.678%  0.871%  Cl  1.35%  1.05%  1.19%  Glu  1.90%  0.940%  6.98%  VUN  1.44%  0.00%  14.6%  Crea  5.09%  4.70%  < >a  Glu (G)b  0.982%  0.579%  2.54%  Analyte  Level 1  Level 3  Level 5  Na  0.593%  0.625%  0.611%  K  1.72%  0.678%  0.871%  Cl  1.35%  1.05%  1.19%  Glu  1.90%  0.940%  6.98%  VUN  1.44%  0.00%  14.6%  Crea  5.09%  4.70%  < >a  Glu (G)b  0.982%  0.579%  2.54%  a“< >” represents data that are out of instrument range. b(G) represents the analysis of glucose using the G cartridge. Drug interferences Most samples fortified with commonly encountered drugs exhibited < ±20% deviations from their respective controls. However, there was substantial interference with creatinine in many of the fortified samples. Several fortified samples also exhibited substantial interference with chloride (Table VI). Potassium was consistently out of the instrument’s reportable range and could not be quantitated. Analytes that had a percent error < ±20% only had regular variations with no interference in the fortified samples. Table VI. Percent error due to common drug interferences. Vial name  Na  K  Cl  Glu  VUN  Crea  Glu (G)  Opioid  5.11%  < >  Inconclusive  7.1%  9.62%  −33.3%a  −4.42%  Coke  4.74%  < >  Inconclusive  12.2%  9.62%  −38.9%  −4.42%  Amine  3.99%  < >  −12.4%  14.7%  7.69%  −27.8%  −1.77%  Basic  5.49%  < >  −5.31%  7.14%  7.69%  −11.1%  −1.77%  Benzo  4.36%  < >  Inconclusive  16.0%  5.77%  −66.7%  −3.10%  Barb  6.61%  < >  Inconclusive  14.7%  9.62%  −66.7%  −4.42%  Acid  0.623%  < >  −3.10%  5.88%  −1.92%  −5.56%  −1.77%  PCP  5.11%  < >  Inconclusive  18.5%  5.77%  −50.0%  −5.75%  EtOH  < >b  Inconclusive  Inconclusive  < >  Inconclusive  50.0%  Inconclusive  Vial name  Na  K  Cl  Glu  VUN  Crea  Glu (G)  Opioid  5.11%  < >  Inconclusive  7.1%  9.62%  −33.3%a  −4.42%  Coke  4.74%  < >  Inconclusive  12.2%  9.62%  −38.9%  −4.42%  Amine  3.99%  < >  −12.4%  14.7%  7.69%  −27.8%  −1.77%  Basic  5.49%  < >  −5.31%  7.14%  7.69%  −11.1%  −1.77%  Benzo  4.36%  < >  Inconclusive  16.0%  5.77%  −66.7%  −3.10%  Barb  6.61%  < >  Inconclusive  14.7%  9.62%  −66.7%  −4.42%  Acid  0.623%  < >  −3.10%  5.88%  −1.92%  −5.56%  −1.77%  PCP  5.11%  < >  Inconclusive  18.5%  5.77%  −50.0%  −5.75%  EtOH  < >b  Inconclusive  Inconclusive  < >  Inconclusive  50.0%  Inconclusive  aBold results represent unacceptable results. b“< >” represents data that are out of instrument range. For the ethanol and accompanying control samples, all analytes gave inconclusive or failed results with deviations either above the acceptable criterion, outside the manufactured reportable range or dependent on other tests that were out of instrument range (Table VI). Stability The percent error exhibited at time zero in capped sample vials for Na, Cl, Glu (Chem8+), VUN, Cre and Glu (G) were <−1, 1, 3, 2, 4 and 3%, respectively. Therefore, samples were stable when analyzed straight from the refrigerator without reaching room temperature; however, analyzes took around 5 min compared to the typical 3–4 min as samples were brought to 37°C by the automatic thermal control subsystem in the device. The percent error exhibited at 2 h out of refrigeration for Na Cl, Glu (Chem8+), VUN, Crea and Glu (G) were <1, −1, 6, −2, 1, 4 and −1%, respectively. Potassium gave inconclusive results at both time periods. Therefore, capped samples were stable out of the refrigerator for up to 2 h. The percent error exhibited for unshaken samples was <1, −1, 6, 2 and 9% for Na, Cl, Glu (Chem8+), VUN and Glu (G), respectively. Creatinine gave a percent error greater than 40% and potassium gave inconclusive results. Discussion Limitations While the i-STAT®1 device has been successfully utilized clinically using whole blood for some time, there is relatively little published information about its potential use on postmortem samples like vitreous humor. At the time of writing, the authors have identified only one published manuscript using the i-STAT®1 device for vitreous humor analysis, and there were no validation parameters presented (17). Additionally, Oles et al. (18) presented a comparison of the i-STAT®1 to the Synchron LX 20 high throughput clinical laboratory analyzer from freshly procured vitreous humor at autopsy and found comparable results for glucose and creatinine, but differences for sodium, chloride and urea nitrogen. However, it is unclear if this work attempted validation criteria or considered potential contaminating species present in vitreous humor. Although there was substantial interference for creatinine and chloride in the drug fortified samples, research by Coe (2) emphasized the stability of postmortem vitreous creatinine. Deviations in clinical ranges for creatinine could indicate dehydration, renal failure or uremia (4). However, we cannot rule out the possibility that the dilution of the fortified samples with drug stock solution contributed to these results because volume matched controlled blanks were not prepared. However, if the results were solely due to volumetric differences we would have expected more unacceptable or inconclusive results for the other analytes measured, especially sodium. It is, therefore, likely that interference due to drug groups being present in vitreous humor is minimal and that altering sample volume by drug spiking additions or other solutions is a noteworthy limitation in the analysis of postmortem vitreous humor (4, 6). An elevated ethanol concentration of 0.197% as well as volume displacement matched controls significantly interfered with all analytes. However, some private laboratories use similar ion-selective electrode techniques with no ethanol interference (19). Therefore, the inconclusive and out-of-range results shown are likely best explained by the dilution of vitreous humor. Based on the overall interference studies, inconclusive or out-of-range results may become apparent in analysis at or near a 10% volume dilution displacement of vitreous humor. Future studies will include performing parallel analyzes of vitreous specimens between the i-STAT®1 and an independent laboratory. Such a study could offer information on which method provides the greatest turnaround time and cost effectiveness. Conclusion Overall, the i-STAT®1 handheld device is accurate and precise for the analysis of postmortem vitreous humor. There were, however, potential drug interferences when analyzing for chloride and creatinine, as both analytes had either uncharacteristic signals or percent error greater than ±20%. These results, including those for the ethanol fortified samples, also indicate sensitivity to volumetric changes in the sample preparation and highlight an inherent difficulty in controlling for endogenous analytes using this matrix. In order to maximize analyte stability in matrix, samples should be properly stored in refrigeration (−2.0–3.0°C) and shaken gently before analysis. Overall samples were stable up to 2 h out of refrigeration. However, research by Gagajewski et al. (20) supports temperature dependent changes in vitreous humor’s sodium, potassium and chloride concentrations depending on storage conditions. When unshaken, most analytes present had a percent error <10%, except for creatinine that had a percent error greater than 40%. This result could indicate creatinine’s sensitivity to sample shaking. The i-STAT®1 has been used in the medical field for over two decades for whole blood chemical analysis. This study has shown that the i-STAT®1 can also be used to analyze vitreous humor to aide with the cause of death determination, though additional testing must be performed to ensure the i-STAT®1’s reliability and cost effectiveness for in-house testing. Acknowledgments The authors would like to acknowledge the Los Angeles County Department of Medical Examiner-Coroner Office for providing the resources needed to complete this project. The authors would also like to thank the staff at the Los Angeles Department of Medical Examiner-Coroner Office for their support, especially Jessica Gadway for her guidance in quality assurance procedures. References 1 Coe, J.I. ( 1974) Postmortem chemistry: practical considerations and a review of the literature. Journal of Forensic Science , 19, 13– 32. 2 Coe, J.I. ( 1993) Postmortem chemistry update emphasis on forensic application. American Journal of Forensic Medicine and Pathology , 14, 91– 117. Google Scholar CrossRef Search ADS PubMed  3 Palmiere, C., Mangin, P. ( 2012) Postmortem chemistry update part Ι. International Journal of Legal Medicine , 126, 187– 198. Google Scholar CrossRef Search ADS PubMed  4 Rose, K.L., Collins, K.A. ( 2008) Vitreous postmortem chemical analysis. College of American Pathologists. http://www.cap.org/apps/docs/newspath/0812/vitreous_postmortem_chemical_analysis.pdf (accessed July 7, 2017). 5 Coe, J.I. ( 1969) Postmortem chemistries on human vitreous humor. American Journal of Clinical Pathology , 51, 741– 750. Google Scholar CrossRef Search ADS PubMed  6 Coe, J.I., Apple, F.S. ( 1985) Variations in vitreous humor chemical values as a result of instrumentation. Journal of Forensic Science , 30, 828– 835. Google Scholar CrossRef Search ADS   7 Blana, S.A., Muβhoff, F., Hoeller, T., Fimmers, R., Madea, B. ( 2011) Variations in vitreous humor chemical values as a result of pre-analytical treatment. Forensic Science International , 210, 263– 270. Google Scholar CrossRef Search ADS PubMed  8 Madea, B., Musshoff, F. ( 2007) Postmortem biochemistry. Forensic Science International , 165, 165– 171. Google Scholar CrossRef Search ADS PubMed  9 Aretz, S., Krohne, T.U., Kammerer, K., Warnken, U., Hotz-Wagenblatt, A., Bergmann, M., et al.  . ( 2013) In-depth mass spectrometric mapping of the human vitreous proteome. Proteome Science , 11, 1– 10. Google Scholar CrossRef Search ADS PubMed  10 Swindle, K.E., Ravi, N. ( 2007) Recent advances in polymeric vitreous substitutes. Expert Review Ophthalmology , 2, 255– 265. Google Scholar CrossRef Search ADS   11 Murthy, K.R., Goel, R., Subbannayya, Y., Jacob, H., Murthy, P.R., Srinivas Manda, S., et al.  . ( 2014) Proteomic analysis of human vitreous humor. Clinical Proteomics , 11, 1– 11. Google Scholar CrossRef Search ADS PubMed  12 Garg, U., Althahabi, R., Amirahmadi, V., Brod, M., Blanchard, C., Young, T. ( 2004) Hyauronidase as aliquefying agent for chemical analysis of vitreous fluid. Journal of Forensic Science , 49, 1– 4. Google Scholar CrossRef Search ADS   13 Chin, C.D., Linder, V., Sia, S.K. ( 2012) Commercialization of microfluidic point-of-Care diagnostic devices. Lab on a chip , 12, 2118– 2134. Google Scholar CrossRef Search ADS PubMed  14 SWGTOX. ( 2013) Scientific working group for forensic toxicology (SWGTOX) standard practices for method validation in forensic toxicology. Journal of Analytical Toxicology , 37, 452– 474. CrossRef Search ADS PubMed  15 (2017) i-STAT®1 Product Update Software and Cartridge Test Information: Update May 2017. https://www.pointofcare.abbott/download?docUri=/technical-library/static-assets/technical-documentation/731668-00H.pdf (accessed 7 July 2017). 16 Shah, V.P., Midha, K.K., Findlay, J.A., Hill, H.M., Hules, J.D., McGilveray, I.J., et al.  . ( 2000) Bioanalytical method validation- a revisit with a decade progress. Pharmaceutical Research , 17, 1551– 1557. Google Scholar CrossRef Search ADS PubMed  17 Antonides, H., Marinetti, L. ( 2011) Ethanol production in a postmortem urine sample. Journal of Analytical Toxicology , 35, 516– 518. Google Scholar CrossRef Search ADS PubMed  18 Oles, M.A., Juhascik, M.P., Jenkins, A.J. Evaluation of the i-STAT 1 handheld analyzer for postmortem vitreous humor chemistry analysis. Abstract Presented at the Society for Forensic Toxicology Meeting, San Francisco, CA, 2011. 19 Caplan, Y.H., Levine, B. ( 1990) Vitreous humor in the evaluation of postmortem blood ethanol concentrations. Journal of Analytical Toxicology , 14, 305– 307. Google Scholar CrossRef Search ADS PubMed  20 Gagajewski, A., Murakami, M.A., Kloss, J., Edstrom, M., Hillyer, M., Peterson, G.F., et al.  . ( 2004) Measurement of chemical analytes in vitreous humor: stability and precision studies. Journal of Forensic Science , 49, 1– 4. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Journal of Analytical ToxicologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off