Repeated Thoracic Discharges From a Stun DeviceDaniel J. Valentino, MD, Robert J. Walter, PhD, Kimberly Nagy, MD, Andrew J. Dennis, DO,Jerry Winners, BS, Faran Bokhari, MD, Dorion Wiley, MD, Kimberly T. Joseph, MD,and Roxanne Roberts, MD

Background: Little objective labora-tory data are available describing thephysiologic effects of stun guns or electro-muscular incapacitation (EMI) devices,but increasing morbidity and even deathsare associated with their use. We hypoth-esized that exposure to EMI discharges ina model animal system would induceclinically significant acidosis and car-diac arrhythmia.

Methods: Ten Yucatan mini-pigs,six experimental and four sham con-trols, were anesthetized with ketamine,xylazine, and glycopyrrolate. Experi-mental pigs were exposed to two 40-second discharges from an EMI deviceover the left thorax. Electrocardio-grams, troponin I, blood gases, and lac-

tate levels were obtained pre-exposure,at 5, 15, 30, 60 minutes, and at 24, 48,and 72 hours postdischarge.

Results: No acute or delayed cardiacarrhythmias were seen. Heart rate wasnot affected significantly (p > 0.05). Asubclinical increase in troponin I was seenat 24 hours postdischarge (0.040 � 0.030ng/mL, p > 0.05). Central venous bloodpH (7.432 � 0.014) and pCO2 (36.1 � 0.9mm Hg) were not changed significantly(p > 0.05) during the 60-minute postdis-charge period. A moderate significant in-crease in lactate occurred in the 5-minutepostdischarge group (4.9 � 0.3 mmol/L,p � 0.0179). All blood chemistry and vitalsigns were normal at 24, 48, and 72 hourspostdischarge.

Conclusions: Although significantchanges in some parameters were seen,these changes were small and of little clin-ical significance. Lengthy EMI exposuresdid not cause extreme acidosis or cardiacarrhythmias. These findings may differfrom those seen with other EMI devicesbecause of the unique MK63 waveformcharacteristics or to specific characteris-tics of the model systems.

Key Words: Electromuscular inca-pacitation, stun gun, swine, cardiac effects,metabolic acidosis, respiratory acidosis, ar-rhythmia, ventricular fibrillation, ketamine,xylazine, electrocardiogram, troponin I, lac-tate, MK63, stun baton, heart, blood pres-sure, creatine kinase-MB isoform, pH,potassium, creatinine.

J Trauma. 2007;62:1134–1142.

Electrical discharges may produce a wide spectrum ofinjury.1 The manifestations of electrical injury dependon the wave form characteristics of the current applied

(AC, DC, mixed, strength, duration, frequency) and its ana-tomic location and path through the tissues of the body.2–4

Effects may include skin burns, neuromuscular incapacitation,skeletal muscle death, cardiac arrhythmia, osteocyte and os-teoblast death, and blood vessel endothelium dysfunction.3–8

Electromuscular incapacitation (EMI) devices utilize high-voltage (20–250 kV), low-frequency (10–100 Hz), time-varying amperage (up to 18 A) DC current to produce painand strong muscle contractions resulting in the incapacitationof volitional movement. The discharge times of stun devicesas used in the field vary greatly. Often short bursts (�5

seconds) are sufficient to subdue most subjects, but the de-vices are capable of delivering very prolonged, continuousdischarges. The only limit on the discharge length is theamount of battery power available, so continuous dischargescould be administered in the field for more than 10 minutesand instances of discharges longer than 90 seconds have beenreported. The utility of these devices in personal defense andlaw enforcement has led to their proliferation and use as analternative to lethal force. EMI devices have been shown tobe very effective when used to incapacitate combative indi-viduals while reducing risk to officers, suspects, andbystanders.9

All EMI devices generate time varying DC currents withwaveforms that are similar but distinctive to each specificdevice. The immediate effects and safety profile of thesedischarges on living organisms are poorly understood.10–12

Amnesty International has compiled a list of more than 100fatalities associated with EMI use in the United States for the2001 to 2004 period.9 This growing list of fatalities hasdrawn a great deal of public attention and raised questionsabout the safety of EMI devices and their potential com-plications, especially their association with fatal ventriculararrhythmia.10,12,13 More than 100 types of EMI devices aremarketed to law enforcement agencies around the world tosubdue combative subjects and increasingly by private citi-zens for personal protection. It is estimated that millions ofpeople in the United States are at risk of exposure to EMI

Submitted for publication November 17, 2006.Accepted for publication February 8, 2007.Copyright © 2007 by Lippincott Williams & Wilkins, Inc.From the Department of Trauma (D.J.V., R.J.W., K.N., A.J.D., J.W.,

F.B., D.W., K.T.J., R.R.), Stroger Hospital of Cook County, Chicago, IL; andthe Department of General Surgery (D.J.V., R.J.W., K.N., A.J.D., F.B.,D.W., K.T.J., R.R.), Rush University Medical Center, Chicago, IL.

Presented at the Chicago Committee on Trauma of the AmericanCollege of Surgeons meeting on June 7, 2006 and at the 14th Annual TraumaSymposium, Indianapolis, IN, November 3, 2006.

Address for reprints: Robert J. Walter, PhD, Department of Trauma,Rm 1300, Stroger Hospital of Cook County, 1900 West Polk St., Chicago, IL60612; email: rwalter@cookcountytrauma.org.

DOI: 10.1097/TA.0b013e3180479858

The Journal of TRAUMA� Injury, Infection, and Critical Care

1134 May 2007

daily and that thousands are exposed annually, creating agrowing public health concern.14 Despite the increasing us-age of EMI devices, there is no consensus in the medicalliterature regarding the safety and type of injuries producedby EMI. Many of the initial studies on stun devices wereperformed using the much less powerful first- or second-generation devices.15–18 The current peer-reviewed literatureon fourth-generation EMI devices is slowly emerging, butmany of the results are conflicting. Some studies show noevidence of acute arrhythmia10,19 in swine and no acidosisor hyperkalemia in healthy human volunteers.18,20 Othersindicate the potential for the development of significantacidosis21 or arrhythmia.22,23 Such conflicting results havemade it difficult to arrive at a consensus regarding the needfor treatment or monitoring of exposed individuals.

As a result of the increasing usage and deployment ofEMI devices, a growing number of individuals are presentingwith injuries related to their exposure to these devices and agrowing number of morbidities and mortalities are being ob-served. We hypothesize that the time-varying DC current uti-lized by certain EMI devices may produce significant myocar-dial injury, acute arrhythmia, acidosis, electrolyte or biochemicalabnormalities. We have developed a model system to study theeffects of EMI devices in anesthetized miniature swine.

MATERIALS AND METHODSAnimals and Groups

Three- to 4-month-old Yucatan mini-pigs (Sinclair Re-search, Columbia, MO) weighing between 16 and 33 kg wereused. The experimental group (80-second thoracic discharge)and the negative (or sham control) group, were comprised ofsix and four animals, respectively. Previous studies in thisfield have not included sham controls and used only intra-animal baseline data as control data. As a result, the data fromthose studies have been somewhat difficult to interpret. Pre-liminary experiments showed that, in view of the small mag-nitude and high reproducibility of the physiologic changesthat occur, it would not be necessary to use more animalssimply to improve the chances of finding statistical signifi-cance when such differences would not possess clinical sig-nificance. The Institutional Animal Care and Use Committeefor the Hektoen Institute for Medical Research reviewed andapproved this project.

Animals were sedated with intramuscular ketamine(Ketaset; Fort Dodge Animal Health, Fort Dodge, IA) andxylazine (Anased; Lloyd, Shenandoah, IA) and respiratorysecretions were inhibited using glycopyrrolate (Robinul; FortDodge Animal Health, Fort Dodge, IA) in the ratio 30/3/0.01mg/kg. During EMI discharge and for all subsequent moni-toring, animals were anesthetized with ketamine and xylazine(5.6/0.8 mg/mL) in sterile saline instilled intravenously usingan infusion pump (Flogard 6200; Travenol, Deerfield, IL)through a 21-G or 23-G cannula placed into an ear vein at arate of 3 mL/hr/kg (16.8/2.4 mg/kg). Animals were intubatedusing cuffed endotracheal tubes (5.0–6.5 mm, Rusch; Kernen,

Germany) after anesthetizing the larynx with 0.25 to 1.0 mLof sprayed 20% benzocaine (Hurricaine; Beutlich Pharm.,Waukegan, IL). Breathing was controlled (15 breaths perminute; tidal volume � 10 mL/kg; min volume � 150 mL/kg).Animals were maintained in dorsal recumbence for all elec-trical discharges and monitoring procedures. At the conclu-sion of each monitoring session from which animals were torecover, intravenous yohimbine (0.05–0.15 mg/kg; Yobine;BenVenue Labs, Bedford, OH) was used to reverse the ef-fects of xylazine and to speed recovery from anesthesia.

Instead of using inhaled halothane or isoflurane anesthe-sia, ketamine/xylazine was used throughout this study. Theprimary local electrical injury anticipated with these wave-forms was membrane electroporation, particularly of nerveand muscle.7 This effect is sensitive to the presence of lipidsor highly lipid-soluble agents such as isoflurane, halothane,or barbiturates. These anesthetics may act to artifactuallyreverse electroporation effects generated in this experimentalsystem. Ketamine and xylazine have some lipid character, butless than isoflurane or halothane, so they are preferred anes-thetics for this study. The ketamine/xylazine combinationused here has been shown to be an effective general anes-thetic in swine24,25 and our data confirm this (see below).

Test DeviceAnecdotal reports from the MK63 manufacturer (Aegis

Industries, Bellevue, ID) indicated that discharges from thedevice caused severe pain and complete voluntary muscleincapacitation. The MK63 stun baton and specifically thecomponent containing the electronic circuitry for generatingEMI discharges (e-pod) was studied. The e-pod was incor-porated into a custom-made laboratory apparatus fashionedfrom 1.25- and 2-inch diameter polyvinyl chloride (PVC)pipe and weights (Fig. 1). This apparatus was held verticallyby clamping the 2-inch PVC pipe to a heavy stand and wasassembled such that the 1.25-inch diameter PVC pipe con-tained the e-pod and this pipe could slide freely up and downwithin the larger pipe while maintaining uniform downwardforce resulting from a final total mass of 1.5 kg for thee-pod/pipe assembly. A 12.0 VDC, 800 mA power supplywas used as the power source.

Experimental Setup and EMI DischargeWhile in dorsal recumbence, all four limbs of the animal

were restrained to the table. The e-pod was placed over theleft midclavicular line at the level of the fourth or fifth rib anddischarges were administered with the electrodes orientedparallel to the cardiac axis from the base of the heart to itsapex. Based on our preliminary experiments using a broadrange of discharge times, we concluded that 80-second dis-charges would be feasible and that any ill effects that mightoccur would most likely be seen with such a prolongeddischarge. Discharges performed for longer times resultedin effects very similar to those seen with 80 seconds, solonger discharges offered no advantage here. The e-pod

Repeated Thoracic Discharges From a Stun Device

Volume 62 • Number 5 1135

was discharged in two separate 40-second intervals, for atotal of 80 seconds, during which time the ventilator wasshut off but spontaneous breaths were permitted. Twoventilated breaths (during 10 seconds) were administeredbetween the 40-second discharges.

Cardiac rhythm was evaluated and monitored continu-ously during anesthesia using a five-lead electrocardiogram(EKG) and monitor (Datex Instruments, Helsinki, Finland) ateach experimental time point; 10- to 15-second tracings wereprinted and retained. EKGs were also recorded during thedischarge. There were eight time points at which centralvenous blood was drawn from the precaval venous complex,and vital signs (tissue oxygen saturation, heart rate, and bloodpressure) and additional EKGs were recorded. The samplingtime points were pre-discharge (time 0); 5, 10, 15, 30, and 60minutes; and 24, 48, and 72 hours postdischarge. Animalswere euthanized according to American Veterinary MedicalAssociation standards after the 72-hour time point by switch-ing the anesthesia to 5% inhaled isoflurane and subsequentlyinjecting 3 mol/L KCl into the heart.

Immediately after drawing, each blood sample wasplaced into heparinized and plain Vacutainer tubes. The hep-arinized blood was tested using an iSTAT analyzer (AbbottPoint-of-Care, Abbott Park, IL) using CG8�, CG4�, creat-inine, and troponin I (TnI) cartridges. These cartridgesreturn data on a variety of parameters including pH, pCO2,bicarbonate, lactate, potassium, TnI, and creatinine. Bloodsamples were stored on ice for a maximum of 2 hours,centrifuged (3000� g for 15 minutes at 4°C), plasma andserum aliquoted into 400-�L microcentrifuge tubes, and sam-ples stored at �85°C until use. Serum from each time pointwas thawed and assayed for creatine kinase-MB isoform

(CK-MB) and myoglobin using microplate enzyme-linkedimmunosorbent assays (ELISAs).

Serum Myoglobin and CK-MB DeterminationPlasma or serum myoglobin, TnI, and CK-MB have been

shown to be useful in evaluating possible cardiac muscledamage usually as a result of myocardial infarction.26–32 Thetime course for the appearance of each of these markers isknown. Levels of cardiac TnI, the most specific marker formyocardial damage, peak at 12 to 24 hours, and may remainelevated for several days. Serum myoglobin becomes ele-vated within 2 to 4 hours of myocardial injury. CK-MB isfound in cardiac and skeletal muscle but is present in muchhigher quantities in cardiac muscle. CK-MB levels becomeelevated within 3 to 4 hours of cardiac injury and remainelevated for 60 to 70 hours. Myoglobin and CK-MB canbecome elevated from noncardiac-related injuries such aschronic muscle disease, skeletal muscle trauma, and renalfailure.28,31,33 As a result, all three of these markers werestudied to determine the extent of cardiac and skeletal muscleinjury.

Serum samples stored at �85°C were thawed once andtested for myoglobin (20 �L/well) and CK-MB (25 �L/well)using solid phase microplate sandwich ELISAs (DiagnosticAutomation, Calabasas, CA). All samples and standards forthese assays were performed in duplicate and averaged. Stan-dard curves using four to seven reference standards of dif-ferent concentrations were generated for each run. Myoglobinand CK-MB concentrations for the experimental serumsamples were interpolated from these standard curves us-ing best-fit regression formulas generated by Excel (Mi-crosoft, Redmond, WA).

Fig. 1. e-Pod and MK63 lab simulator. The distance between the outer electrodes on the MK63 was 2.1 inches and the distance between theinner electrodes was 0.35 inches. The total mass of experimental device was 1.5 kg.

The Journal of TRAUMA� Injury, Infection, and Critical Care

1136 May 2007

Data Reduction and Statistical AnalysisEach of the animals described above were studied for all

EKGs and blood chemistry. To simplify comparisons, wher-ever possible the values for each of these parameters weregraphed during the entire time course studied including im-mediately pretreatment (0 time); 5, 10, 15, 30, and 60 min-utes; and 24, 48, and 72 hours postdischarge. All data pointsrepresent means � SEM. Normal values were drawn frompublished data for mini-pigs, full-sized swine, or humans inthat order of preference based on data availability andreliability.24,25,34–39

Parametric statistics including analysis of variance(ANOVA), paired or unpaired t tests, followed by Tukey’sor Student-Neuman-Keuls posttests were employed tocompare quantitative data and groups. Trends were eval-uated using linear or nonlinear regression. The experimen-tal groups were compared against their baseline for eachparameter to assess whether changes from baseline weresignificant. In addition, the experimental and controlgroups were compared with each other (Prism and InStat,GraphPad Software, San Diego, CA).

RESULTSVital Signs Were Moderately Altered byEMI Discharge

Heart rate (Fig. 2) in experimental animals showed asignificant (p � 0.05) decrease from baseline (117 � 3 bpmat t � 0) at 5 minutes postdischarge (104 � 4 bpm). Itreturned to baseline levels at subsequent time points. Noacute changes in pulse oximetry were observed at any time.Tachycardia was not seen in response to EMI discharge.

Blood pressure (Fig. 3) showed minor fluctuations dur-ing the 60-minute postdischarge time period (systolic �126 � 3 mm Hg, diastolic � 53 � 2 mm Hg). The systolicblood pressure decreased during the initial 60-minute moni-toring period reaching a nadir at 30 minutes (115 � 8 mmHg). This decrease in systolic blood pressure was not signif-icant (p � 0.05) compared with the baseline value (132 � 10mm Hg) in the experimental group. Blood pressure did notbecome elevated or show any other changes suggesting painperception.

No Evidence of Acute Arrhythmia or MyocardialInjury was Found

Rhythm strips taken before, during, and after dischargeof the MK63 over the thorax showed no acute changes incardiac rhythm at any time (Fig. 4). EKGs show continuedregular ventricular contractions throughout the EMI dis-charge. No arrhythmias were observed in any animal at anytime point. Mean CK-MB levels (Fig. 5) were not signifi-cantly affected in the experimental group as compared withcontrols or with experimental baseline values. CK-MB levelswere not elevated in any of the individual experimental ani-mals. Mean TnI values (Fig. 6) increased at the 24-hour timepoint in negative controls (0.023 � 0.019 ng/mL) and exper-imental animals (0.040 � 0.031 ng/mL). The observed in-crease in experimental animals, however, was not statisticallysignificant when compared with that of controls (p � 0.695)or with baseline levels for the experimental group (0.000 �0.000 ng/mL). A TnI value of 0.040 ng/mL represents theupper limit of normal.40 The largest value of TnI observed inan individual experimental animal was 0.190 ng/mL at 24hours postdischarge. TnI in this animal returned to baseline at48 hours.

Fig. 2. Heart rate before EMI discharge and during the 72-hourtime course after EMI discharge. Heart rate (mean � SEM) showedan initial decrease 5 minutes after discharge but was maintainedbetween 100 and 120 bpm at all time points in the experimentalgroup. None of these variations were statistically significant norwere there any significant differences between the control andexperimental groups (one-way ANOVA).

Fig. 3. Blood pressure versus time during the 72-hour time courseafter EMI discharge. Both systolic and diastolic pressures (mean �

SEM) are plotted for the experimental and control groups. Therewere no significant differences in the systolic and diastolic bloodpressures in the experimental group at any time point when eachwas compared with its baseline value. No significant differenceswere seen between the control and experimental groups for systolicand diastolic blood pressures (one-way ANOVA).

Repeated Thoracic Discharges From a Stun Device

Volume 62 • Number 5 1137

No Clinically Significant Acidosis was Seen WithEMI Discharge

Central venous blood pH (Fig. 7) showed an initial de-crease after EMI discharge at the 5-minute time point. Thischange (7.45 � 0.03 to 7.39 � 0.02) was statistically signif-icant (p � 0.05), although the observed values were stillwithin normal ranges. Central venous blood pH then returnedto baseline during the 60-minute monitoring session. Controlanimals had a significantly higher pH during the initial 60-minute time period. A slight acidosis was noted only in oneanimal (pH � 7.32) at the 5-minute time interval. All otheranimals maintained blood pH at or above normal levels forthe duration of the experiment.

Blood pCO2 was not significantly changed by EMI dis-charge. A small increase in pCO2 was seen at 5 minutes(39.2 � 2.8 mm Hg) postdischarge. This change was notsignificantly different (p � 0.05) from the baseline value of

36.6 � 2.7 mm Hg. The pCO2 returned to baseline during the60-minute monitoring period. One animal had pCO2 levelsexceeding normal (pCO2 � 49.6 mm Hg) at the 5-minutepostdischarge time. The levels for this animal then returned tonormal subsequently during the 60-minute time course.

Bicarbonate levels decreased at 5 minutes postdis-charge (24.3 � 0.6 mmol/L) from baseline values (25.2 �1.1 mmol/L), but this was not a significant change frombaseline ( p � 0.05). This decrease contrasted to a smallincrease in the control group. There was no significantdifference when comparing controls with experimental an-imals during the initial 60-minute period (one-wayANOVA, p � 0.05). All bicarbonate levels were within thenormal reference range. One animal in the experimentalgroup had a bicarbonate level of 20.6 mmol/L before EMIexposure. The bicarbonate level in this animal rose tonormal levels after exposure.

Fig. 4. Representative EKGs from predischarge through EMI discharge (A) and through 72 hours postdischarge (B). EKGs show continuedregular ventricular contractions throughout the EMI discharge. No evidence of acute arrhythmia was detected during, immediatelypostdischarge, or at later time points postdischarge. No arrhythmias were observed in any animal at any time point. Note that there is littlenoise in the bottom tracing (lead aVR) in (A) and that sinus rhythm is maintained throughout the discharge.

The Journal of TRAUMA� Injury, Infection, and Critical Care

1138 May 2007

Lactate values showed a small increase at the 5-minutetime point (4.89 � 0.7 mmol/L). This increase was notstatistically significant (p � 0.05) from baseline values(3.7 � 1.3 mmol/L). The experimental group showed signif-icantly higher levels of lactate when compared with controlsduring the initial 60-minute monitoring period (one-wayANOVA, p � 0.05). Lactate values were below baselinevalues at 24 hours and were similar to controls. Two of theexperimental animals had elevated baseline lactate levels(8.13 and 7.43 mmol/L). These were peak values seen forthese animals and they decreased to normal baseline levelsduring the initial 60-minute postdischarge period. The otherfour experimental animals had normal starting lactate levelsand showed slight increases after EMI exposure, which thenreturned to baseline.

EMI Discharge Did Not Affect Electrolyte LevelsPotassium and creatinine levels were not affected by

EMI discharge. Potassium values (Fig. 8) increased slightlyafter EMI exposure in all animals during the initial 60 min-utes (peak value � 3.6 � 0.1 mmol/L at t � 60 minutes).These increases were not statistically significant (p � 0.05)compared with baseline values (3.5 � 0.1 mmol/L). Allvalues were within normal limits for both of these parametersat all time points in all animals. Creatinine values (Fig. 9) didnot change significantly after EMI discharge. At no time didcreatinine values exceed normal levels in any animal (range0.6–1.2 mg/dL).

EMI Discharge Did Not Significantly AffectSerum Myoglobin

Mean myoglobin levels in the experimental group wereincreased when compared with the control group at all timepoints except 72 hours, but these differences were not signif-icantly different (one-way ANOVA, p � 0.05). Mean myo-

Fig. 5. CK-MB concentration during the 72-hour time course afterEMI discharge. CK-MB concentration (mean � SEM) did not risesignificantly at any time in either group. The control group showedsome variability with time, but the differences between the controland experimental groups were not significant (one-way ANOVA).

Fig. 6. Troponin-I (TnI) values during the 72-hour time courseafter EMI discharge. At the 24-hour time point, the TnI value for thecontrol group rose to 0.023 � 0.019 ng/mL and for the experimentalgroup to 0.040 � 0.031 ng/mL (mean � SEM), but TnI returned tobaseline values for both groups subsequently.

Fig. 7. Central venous pH over time for control and experimentalgroups. No significant changes in pH were seen. Similarly, thechanges in pCO2, lactate, and bicarbonate showed only smallchanges correlating with the observed changes in blood pH.

Fig. 8. Potassium ion concentrations during the 72-hour timecourse. Potassium levels did not change significantly after EMIdischarge. There were no significant differences between the exper-imental and control groups indicating a lack of significant muscleinjury.

Repeated Thoracic Discharges From a Stun Device

Volume 62 • Number 5 1139

globin levels did not exceed the upper limit of normal (54ng/mL). The highest single myoglobin level seen was 154ng/mL at 24 hours after EMI exposure. Two other animalshad peak levels of 101 ng/mL and 105 ng/mL at 60 minutesand 48 hours postdischarge, respectively. All elevated myo-globin values returned to normal levels by the next monitor-ing interval in these animals.

DISCUSSIONCardiac Effects

Case reports, autopsies, and retrospective analysis havefound EMI discharge to be associated with fatal ventricularfibrillation in humans, although the frequency of this com-plication is extremely low.10–13,41,42 In the present study, nochanges in cardiac rhythm were seen even after lengthy EMIexposures. At no time in the 72-hour monitoring period didany animal die. The experimental animals maintained a meanheart rate that decreased slightly from but returned rapidly tobaseline values. Other studies in anesthetized swine havereported that TASER X26 discharges resulted in acute onsetof tachycardia.21 This effect was also reported in studies ofhealthy human volunteers where the response was ascribed tointense pain associated with the discharge.20,18 The absenceof tachycardia in this experimental model suggests a deepplane of anesthesia that suppresses pain or may indicate thatthe MK63 device evoked less pain in this model than did theTASER X26 in anesthetized swine or conscious humans. Inany case, the MK63 device did not appear to directly interruptor capture cardiac rhythm.

Alternatively, sudden deaths associated with TASERdischarges in humans may result from direct or indirectdamage to the myocardium, which then leads to delayedarrhythmia.10–12 Two cardiac markers, CK-MB and TnI,were assayed in the present study to assess myocardial injury.There were no elevations in CK-MB after 80-second MK63

discharges and TnI showed small but insignificant rises. TnIis released from cardiac myocytes26–32 when their cell mem-branes are damaged. Free TnI then diffuses into the intersti-tial space and eventually into the blood.43 The high sensitivityand specificity of commercially available assays have madeTnI the gold standard for detecting myocardial injury. Re-lease of TnI from both human and swine cardiac myocytespeaks 18 to 24 hours after the injury and then graduallydecreases to normal during the course of the next severaldays.43,44

For the i-STAT TnI assay, the cutoff value for the upperlimit of normal is 0.030 �g/mL, which represents the97.5th percentile of healthy individuals without heart dis-ease (iSTAT manufacturer manual, Abbott Laboratories, Ab-bott Park, IL). Based on subsequent evaluation of a largenumber of patients, Apple et al.40 have determined the valuefor the 99th percentile of healthy individuals to be 0.040ng/mL. This value is based on complication rates from pa-tients presenting with symptoms of acute coronary syndrome.The TnI value considered diagnostic of myocardial infarctionis higher than the upper limit of normal and is determined byeach individual clinical laboratory.45 There are no publishedguidelines that establish levels of clinically significant TnIlevels in swine.

At its peak, the mean TnI value in the present studyreached the human upper limit of normal (0.040 ng/mL) atthe 24-hour monitoring point in the experimental group. Atthis time point, two animals (out of six) showed elevated TnIlevels (0.190 ng/mL and 0.050 ng/mL). For both animals, TnIreturned to the baseline level of 0.000 ng/mL at the subse-quent (48 and 72 hours) monitoring time points. TnI alsoincreased in one control animal at the 24-hour time point(0.080 ng/mL) but values similarly returned to baseline at thenext monitoring time.

The time-related pattern and magnitude of TnI increasesseen here were very different from that seen in humans orswine with myocardial injury.44,45 In models of myocardialischemia, TnI values peak at 18 to 24 hours and remainelevated for days to weeks thereafter. Because TnI valueshere decreased to zero within 48 hours, it is unlikely thatthese elevated TnI values signify myocardial injury such asthat seen in human heart attack or swine models of myocar-dial ischemia. The induction and prolonged anesthesia ses-sions (2–3 hours) employed on the first day of the experimentmay have caused some degree of cardiac stress that contrib-uted to the TnI elevation seen here. Anesthesia, especially atinduction, is a known cardiac stressor that results in anincreased risk of adverse cardiac events.46

Blood Gas DataIn rare circumstances, humans subjected to EMI dis-

charges have died within 24 hours of discharge exposure. Ithas been postulated that such “in-custody deaths” may resultfrom cardiac instability as a result of EMI-induced lacticacidosis.42,47 No evidence of severe acidosis was seen in the

Fig. 9. Creatinine concentrations seen during the 72-hour timecourse. Creatinine levels did not change significantly after EMIdischarge. There were no statistically significant differences be-tween the experimental and control groups indicating a lack ofmuscle injury.

The Journal of TRAUMA� Injury, Infection, and Critical Care

1140 May 2007

present study. There were minor physiologic fluctuations inacid-base status, but none of these changes were statisticallyor clinically significant. Similarly, the changes in pCO2, lac-tate, and bicarbonate showed only small, insignificant (p �0.05) changes that correlated with the observed changes inblood pH.

The results of the present study are largely at variancewith those of Jauchem et al.,21 which examined the effects ofdischarges from a TASER X26 in swine. In that study, severeacidosis (pH � 7.0) was seen immediately after EMI expo-sure and this was accompanied by dramatic hypercapnia(pCO2 � 100 mm Hg) and lactate elevation (�15 mmol/L).There are, however, a number of methodologic differencesbetween the present study and that of Jauchem et al.21 thatmay account for some or all of these disparities.

First, the devices studied differed somewhat in theirwaveforms, although both produce an EMI response. TheTASER X26 used by Jauchem et al.21 delivers DC pulses ata voltage of about 50 kV, with a pulse duration of 140 �sec,a frequency of 19 Hz, and power of 0.36 J/pulse (www.taser.com). The MK63 device used here also delivers DCpulses but at a lower voltage (�19 kV), with pulse durationsof 15 �sec at 65 Hz and 0.08 J/pulse (C. Hathcock, PhD,Avocet Polymer Technologies, personal communication).Second, in the present study the MK63 device was dischargedfor 80 seconds with one 10-second rest interval after the first40-second discharge. Jauchem et al.21 administered dis-charges in 5-second intervals alternating with 5-second restsfor 3 minutes, accumulating a total of 90 seconds of EMIexposure. Differences in either the device design or the dis-charge pattern might account for some of the discrepanciesbetween the studies. Third, different anesthetic agents andmethods of physiologic support were used. Jauchem et al.21

did not mechanically ventilate the intubated animals duringthe study, whereas the swine in the present study were ven-tilated while anesthetized, except during the 80-seconds EMIdischarge. As a result, it is unclear what role respiratorydepression from the propofol/telazol anesthesia and bu-prenorphine analgesia or the lack of respiratory support in theJauchem et al.21 study may have played in the reportedacid-base abnormalities.

The animals used in the present study did not show anyacute or delayed changes in electrolytes. Potassium levels canbe used to assess the extent of skeletal muscle injury and alsothe potential for electrolyte-associated cardiac arrhythmia. Inthe present study, no significant changes in electrolytes wereseen. Jauchem et al.21 reported significant acute increases inpotassium levels. Again, differences in experimental designor devices studied may have contributed to the discrepanciesin electrolyte levels seen in these studies.

In summary, no evidence of acute arrhythmia, myocar-dial damage, severe acidosis, or electrolyte/biochemical ab-normalities were seen in the present study. In this swinemodel, prolonged discharges from the MK63 device pro-duced no significant or harmful physiologic changes. Be-

cause previous animal studies of the TASER X26 showedsome dramatic physiologic changes,21 the present findingsmay be a result of the unique wave form and pulse powergenerated by the MK63 device, differences in the electrodespacing for the MK63 as compared with the TASER X26, ordifferences between the model systems. Further studies areneeded to distinguish among these possibilities, to evaluatethe short- and long-term physiologic effects of these devices,and to elucidate the mechanism by which these devices trig-ger EMI. This knowledge will be instrumental in developingfuture guidelines and treatment protocols for the growingnumber of individuals exposed to EMI.

Study LimitationsFor ethical reasons, anesthesia (ketamine � xylazine)

was used in this swine model. Anesthesia precludes painperception, which is one of the two principal effects of EMIdischarges in conscious humans. Pain perception would un-doubtedly alter some of the responses seen here. The numberof animals used here was also limited, but this was counter-balanced by the high animal-to-animal reproducibility of theresults. Finally, in the field, EMI devices are used to subduecombative individuals in a state of greatly increased sympa-thetic activity. In many cases, these persons are under theinfluence of alcohol or other drugs. Under those conditions,the effects of EMI discharge may vary from those seen here.

ACKNOWLEDGMENTSWe would like to thank the staff of the Animal Facility for their

assistance and Raphael Lee, MD, DSc, Anita Mehta, PhD, and CaseyHathcock, PhD for helpful discussions. We also thank Avocet PolymerTechnologies for supporting these studies.

REFERENCES1. Koumbourlis AC. Electrical injuries. Crit Care Med. 2002;30:S424–

S430.2. Fish RM. Electric injury, part I: Treatment priorities, subtle

diagnostic factors, and burns. J Emerg Med. 1999;17:977–983.3. Oltman CL, Clark CB, Kane NL, et al. Coronary vascular

dysfunction associated with direct current shock injury. Basic ResCardiol. 2003;98:406–415.

4. Lee RC. Injury by electrical forces: pathophysiology, manifestations,and therapy. Curr Probl Surg. 1997;34:677–764.

5. Kalkan T, Demir M, Ahmed AS, et al. A dynamic study of thethermal components in electrical injury mechanism for betterunderstanding and management of electric trauma: An animal model.Burns. 2004;30:334–340.

6. Luce EA. Electrical burns. Clin Plast Surg. 2000;27:133–143.7. Lee RC, Zhang D, Hannig J. Biophysical injury mechanisms in

electrical shock trauma. Annu Rev Biomed Eng. 2000;2:477–509.8. Danielson JR, Capelli-Schellpfeffer M, Lee RC. Upper extremity

electrical injury. Hand Clin. 2000;16:225–234.9. Battershill P, Naughton B, Laur D, et al. TASER technology review

and interim recommendations. Available at www.cprc.org/docs/bcopcc_final.pdf. Accessed March 4, 2007.

10. McDaniel WC, Stratbucker RA, Nerheim M, Brewer JE. Cardiacsafety of neuromuscular incapacitating defensive devices. PacingClin Electrophysiol. 2005;28 Suppl 1:S284–S287.

Repeated Thoracic Discharges From a Stun Device

Volume 62 • Number 5 1141

11. Bleetman A, Steyn R, Lee C. Introduction of the Taser into Britishpolicing. Implications for UK emergency departments: an overviewof electronic weaponry. Emerg Med J. 2004;21:136–140.

12. Bleetman A, Steyn, R. The advanced TASER: A medical review.Available at www.taser.com/documents/tasersubmit.pdf. AccessedMarch 20, 2007.

13. Roy OZ, Podgorski AS. Tests on a shocking device–the stun gun.Med Biol Eng Comput. 1989;27:445–448.

14. Bozeman WP. Withdrawal of TASER electroshock devices: toomuch, too soon. Ann Emerg Med. 2005;46:300–301.

15. Ordog GJ, Wasserberger J, Schlater T, Balasubramanium S.Electronic gun (Taser) injuries. Ann Emerg Med. 1987;16:73–78.

16. O’Brien DJ. Electronic weaponry–a question of safety. Ann EmergMed. 1991;20:583–587.

17. Koscove EM. The TASER weapon: a new emergency medicineproblem. Ann Emerg Med. 1985;14:1205–1208.

18. Levine SD, Sloane C, Chan T, Vilke G, Dunford J. Cardiacmonitoring of subjects exposed to the TASER. Acad Emerg Med2005;12:71.

19. Lakkireddy D, Wallick D, Ryschon K, et al. Effects of cocaineintoxication on the threshold for stun gun induction of ventricularfibrillation. J Am Coll Cardiol. 2006;48:805–811.

20. Ho JD, Miner JR, Lakkireddy DR, Bultman LL, Heegaard WG.Cardiovascular and physiologic effects of conducted electricalweapon discharge in resting adults. Acad Emerg Med. 2006;13:589–595.

21. Jauchem JR, Sherry CJ, Fines DA, Cook MC. Acidosis, lactate,electrolytes, muscle enzymes, and other factors in the blood of Susscrofa following repeated TASER exposures. Forensic Sci Int. 2005;161:20–30.

22. Webster JG, Will JA, Sun H, et al. Can TASERs directly causeventricular fibrillation. IFBME Proc. 2006;14:3307–3310.

23. Nanthakumar K, Billingsley IM, Masse S, et al. Cardiacelectrophysiological consequences of neuromuscular incapacitatingdevice discharges. J Am Coll Cardiol. 2006;48:798–804.

24. Swindle MM. Surgery, Anesthesia, and Experimental Techniques inSwine. Ames: Iowa State University Press; 1998.

25. Bollen PJA, Hansen AK, Rasmussen HJ. Important biologicalfeatures. In: Suckow MA, ed. The Laboratory Swine. Boca Raton:CRC Press; 2000:1–14.

26. Kost GJ, Kirk JD, Omand K. A strategy for the use of cardiac injurymarkers (troponin I and T, creatine kinase-MB mass and isoforms,and myoglobin) in the diagnosis of acute myocardial infarction. ArchPathol Lab Med. 1998;122:245–251.

27. Apple FS, Murakami MM, Quist HH, et al. Prognostic value of theOrtho Vitros cardiac troponin I assay in patients with symptoms ofmyocardial ischemia. Risk stratification using European Society ofCardiology/American College of Cardiology recommended cutoffvalues. Am J Clin Pathol. 2003;120:114–120.

28. Apple FS, Christenson RH, Valdes R Jr, et al. Simultaneous rapidmeasurement of whole blood myoglobin, creatine kinase MB, andcardiac troponin I by the triage cardiac panel for detection ofmyocardial infarction. Clin Chem. 1999;45:199–205.

29. Ng SM, Krishnaswamy P, Morissey R, et al. Ninety-minuteaccelerated critical pathway for chest pain evaluation. Am J Cardiol.2001;88:611–617.

30. Ng SM, Krishnaswamy P, Morrisey R, et al. Mitigation of theclinical significance of spurious elevations of cardiac troponin I insettings of coronary ischemia using serial testing of multiple cardiacmarkers. Am J Cardiol. 2001;87:994–999.

31. Apple FS, Murakami M, Panteghini M, et al. International survey onthe use of cardiac markers. Clin Chem. 2001;47:587–588.

32. Apple FS, Wu AH. Myocardial infarction redefined: role of cardiactroponin testing. Clin Chem. 2001;47:377–379.

33. Perry SV. The regulation of contractile activity in muscle. BiochemSoc Trans. 1979;7:593–617.

34. Sinclair Research. Clinical Chemistry Values of Sinclair Pigs.Columbia, MO: Sinclair Research, 1991.

35. Parsons AH, Wells RE. Serum biochemistry of healthy Yucatanminiature pigs. Lab Animal Sci. 1986;36:428–430.

36. Brechbuhler T, Kaeslin M, Wyler F. Reference values of variousblood constituents in young minipigs. J Clin Chem Biochem. 1984;22:301–304.

37. Waagstein LM, Wennberg E, Waagstein F, Haljamae H. Hypertonicsaline without or with dextran-70 in the treatment of experimentalacute myocardial ischemia and reperfusion. Crit Care Med. 1999;27:605–616.

38. Rixen D, Raum M, Holzgraefe B, et al. Local lactate and histaminechanges in small bowel circulation measured by microdialysis in pighemorrhagic shock. Shock. 2002;18:355–359.

39. Hicks TA, McGlone JJ, Whisnant CS, et al. Behavioral, endocrine,immune, and performance measures for pigs exposed to acute stress.J Anim Sci. 1998;76:474–483.

40. Apple FS, Ler R, Chung AY, Berger MJ, Murakami MM. Point-of-care i-STAT cardiac troponin I for assessment of patients withsymptoms suggestive of acute coronary syndrome. Clin Chem. 2006;52:322–325.

41. Schlosberg M, Levin J, Batliwalla S, Daniels J. Stun Gun Fallacy:How the Lack of Taser Regulation Endangers Lives. Available athttp://aclunc.org/issues/criminal_justice/police_practices/asset_upload_file384_3131.pdf. Accessed March 20, 2007.

42. Amnesty International. United States of America Excessive andLethal Force? Amnesty International’s concerns about deaths andill-treatment involving police use of TASERS. http://www.amnestyusa.org/countries/usa/Taser_report.pdf. Accessed March 4, 2007.

43. Zipes DP, Libby P, Bonow RO, Braunwald E. Braunwald’s HeartDisease: A Text Book of Cardiovascular Medicine. Philadelphia:Saunders; 2005:1158–1161.

44. Feng YJ, Chen C, Fallon JT, et al. Comparison of cardiac troponin I,creatine kinase-MB, and myoglobin for detection of acute ischemicmyocardial injury in a swine model. Am J Clin Pathol. 1998;110:70–77.

45. Kontos MC, Shah R, Fritz LM, et al. Implication of different cardiactroponin I levels for clinical outcomes and prognosis of acute chestpain patients. J Am Coll Cardiol. 2004;43:958–965.

46. Townsend CM, Beauchamp DR, Evers MB, Mattox KL, SabistonDC. Sabiston Textbook of Surgery. Philadelphia: Saunders; 2004:425–426.

47. Stratton SJ, Rogers C, Brickett K, Gruzinski G. Factors associatedwith sudden death of individuals requiring restraint for exciteddelirium. Am J Emerg Med. 2001;19:187–191.

The Journal of TRAUMA� Injury, Infection, and Critical Care

1142 May 2007