the veterinary journal - kasetsart...

The Veterinary Journal 196 (2013) 483–491

Contents lists available at SciVerse ScienceDirect

The Veterinary Journal

journal homepage: www.elsevier .com/ locate/ tv j l

Evaluation of clinical and electrocardiographic changes during the euthanasiaof horses

R. Buhl a,⇑, L.O.F. Andersen a, M. Karlshøj a, J.K. Kanters b,c

a Department of Large Animal Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Højbakkegård Allé 5, DK-2630 Taastrup, Denmarkb Department of Cardiology P, Gentofte University Hospital, DK 2900 Hellerup, Denmarkc Laboratory of Experimental Cardiology, Department of Biomedical Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Blegdamsvej 3B, DK-2200Copenhagen N, Denmark

a r t i c l e i n f o

Article history:Accepted 18 November 2012

Keywords:ECGEuthanasiaHorsePentobarbitalQT interval

1090-0233/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.tvjl.2012.11.016

⇑ Corresponding author. Tel.: +45 3533 2994.E-mail address: [email protected] (R. Buhl).

a b s t r a c t

The objective of this prospective field study was to investigate whether commonly used criteria for clin-ical death occurred at the same time as cardiac death, as determined by electrocardiography. Specific ECGchanges during euthanasia were also studied. Twenty-nine horses were euthanized with pentobarbital attwo different dose rates and 15 of the 29 horses also received detomidine hydrochloride for sedation. ECGwas recorded prior to and during euthanasia. Time to collapse, cessation of reflexes, heart sounds andasystole were recorded. ECG recordings were used to calculate RR intervals, PQ duration, QRS duration,distance from QRS complex to end of T wave corrected for HR (QTc interval), duration of T-wave frompeak to end (TpeakTend) and amplitudes of T wave (Tpeak) before and during euthanasia. Differencesbetween groups and ECG changes were evaluated using analysis of variance.

Clinical determination of death occurred before cardiac death (P < 0.05). Sedated horses took longer tocollapse than unsedated horses (P < 0.0001), but asystole occurred faster in sedated horses (P < 0.0001).No significant changes in QRS duration were observed, but RR, PQ, QTc, TpeakTend and Tpeak were influencedby both pentobarbital dose and sedation (P < 0.05–<0.0001). In conclusion, sedation prior to euthanasiaresulted in a shorter time to asystole and is therefore recommended for the euthanasia of horses. Impor-tantly, the results show that the clinical definition of death occurred significantly earlier than cardiacdeath (defined as asystole), which indicates that the clinical declaration of death in horses could be pre-mature compared to that used in humans.

� 2012 Elsevier Ltd. All rights reserved.

Introduction

Euthanasia means ‘good death’ and is defined as death withminimal pain and distress (USA-AVMA-Panel, 2007). Euthanasiaagents cause death by three basic mechanisms, namely, (1) hypox-ia, (2) direct depression of neurons necessary for life functions, or(3) physical disruption of brain activity necessary for life. Euthana-sia techniques should result in rapid loss of consciousness followedby cardiac or respiratory arrest and ultimately, the loss of brainfunction. Euthanasia of horses can be performed either by injec-tion, or using a penetrating bolt, or by a free bullet (Jones, 1992).

In human medicine, two different definitions on death are used,cardiorespiratory death and brain death. These terms are especiallyimportant when considering organs for transplantation and whento perform resuscitation. However, there is continuous debateabout these definitions and concerns are often raised about theaccuracy of the clinical diagnosis of death (Kerridge et al., 2002;

ll rights reserved.

Doig and Burgess, 2003; Zamperetti et al., 2003). In veterinarymedicine, the definition of death is not clear. The American Veter-inarian Medical Association states that death should be confirmedby examining the animal for cessation of vital signs, but no exactdefinition of death is specified.1

Among the most important vital signs of death are the absenceof a palpable or auscultable heartbeat, absence of the corneal re-flex, and respiratory arrest (Sinclair, 2001). Although apparentlysimple, the evaluation of these vital signs is not an easy task andin human medicine, pulse palpation has now been omitted in mod-ern cardiac arrest algorithms due to lack of sensitivity, even forexperienced personnel. In ECG-monitored human patients, asys-tole is a marker for the onset of death. Although death is alwaysassociated with asystole, short term asystole can be reversible. Inthe setting of euthanasia, it is likely that prolonged asystole is asso-ciated with death.

1 See: www.avma.org/issues/animal_welfare/euthanasia.pdf (Accessed July 25,2012).

http://dx.doi.org/10.1016/j.tvjl.2012.11.016

mailto:[email protected]

http://dx.doi.org/10.1016/j.tvjl.2012.11.016

http://www.sciencedirect.com/science/journal/10900233

http://www.elsevier.com/locate/tvjl

Fig. 1. ECG showing the different ECG variables. RR (ms), distance between two Rwaves; PQ (ms), distance from P to Q wave; QRS (ms), duration of QRS complex; QT(ms), distance from Q to the end of the T wave (Tend); TpeakTend (ms), duration from

484 R. Buhl et al. / The Veterinary Journal 196 (2013) 483–491

Several reports describe euthanasia of horses (Jones, 1992; Hub-bell et al., 1993; Knottenbelt et al., 1994; Steere, 1990; Sinclair,2001), but to our knowledge, no studies have investigated the elec-trical activity of the heart using systematic analysis of ECGs duringeuthanasia. Moreover, equine studies focusing on morphologicalchanges in electrocardiographic wave patterns and subsequentclinical changes are limited (Ayala et al., 2000; Matthews andHartsfield, 2004; Morgan et al., 2011) and most studies focus onthe assessment of arrhythmias. Information on normal ECG mor-phology and altered morphology during the process of death areessential in building a database of ECG changes associated withvarious pathophysiological conditions in horses. Such informationenables the equine clinician to recognize clinical patterns andintervene when significant ECG changes occur.

The purpose of the present study was to study ECG changes dur-ing the euthanasia of horses performed using sodium pentobarbitalat two different dose rates, with or without sedation. Additionally,temporal associations between commonly used clinical criteria fordeath, such as clinical examination, auscultation, cessation of re-flexes and collapse, and cardiac death as determined by ECG, wereinvestigated.

top of the T wave (Tpeak) until the T wave reaches baseline (Tend); Tpeak (mV),amplitude of the T wave.

Materials and methods

Animals

Twenty-nine horses (aged 1–30 years, bodyweight 160–703 kg) that wereeuthanized at University of Copenhagen were included in this project. The reasonsfor euthanasia included orthopaedic conditions, chronic incurable diseases andelective reasons. Horses euthanased due to cardiovascular problems, systemic ill-ness, or exhaustion and horses euthanased during anaesthesia were not included.Prior to euthanasia, a clinical examination was carried out and informed consentfor participation in the study was obtained from the horse owner. As the clinicalprocedures performed were standard procedures, permission was not requiredfrom the Danish Animal Experimentation Inspectorate.

Euthanasia procedure

Horses were randomly assigned into four groups based on differences in pento-barbital dose (high or low dose), and whether the horses were sedated with detom-idine hydrochloride (Domosedan, Orion Pharma) prior to euthanasia (Table 1). Anintravenous (IV) catheter (12 G, Intraflon, Vygon, E-vet) was placed in the jugularvein to avoid perivascular injection of drugs. Euthanasia was induced by adminis-tration of pentobarbital at 66.67 mg/kg IV for the high dose, as recommended byKnottenbelt (1995). The low dose was 44.44 mg/kg administered IV as a 200 mg/mL solution. Sedation was administered IV 5–10 min before euthanasia at a doserate of 0.01 mg/kg. An exception was made to random assignment to treatmentgroups for one nervous horse that needed immediate sedation. That horse was se-dated using detomidine (0.02 mg/kg IM).

Protocol

For all euthanasia procedures, one person registered the timing of events usinga stopwatch and another person administered the pentobarbital and recorded thecessation of clinical signs. During the injection procedure, the horse was restrainedby a third person, but a number of individuals performed this role throughout thestudy. The stop watch was started when the injection of pentobarbital was initiatedand split times were registered on the stop watch for injection time, time taken forthe horse to collapse (defined as when the horses were in lateral recumbency), timeto absence of auscultable heartbeat, time to asystole (defined as a flat baseline onECG) and time for loss of palpebral, corneal and perineal reflexes.

Table 1Distribution of the horses into four groups.

Pentobarbital Sedation with detomedine(0.01 mg/kg IV)

Group 1 (n = 9) High dose (66.67 mg/kg IV) PresentGroup 2 (n = 6) High dose (66.67 mg/kg IV) AbsentGroup 3 (n = 6) Low dose (44.44 mg/kg IV) PresentGroup 4 (n = 8) Low dose (44.44 mg/kg IV) Absent

Electrocardiographic monitoring

A modified chest lead ECG (Krutech Televet 100, Kruuse A/S) was recorded,monitored and stored telemetrically on a laptop. Hair coat was clipped to ensuregood skin contact and electrodes were attached on the saddle girth area, the redand yellow electrodes on the left and right side respectively, approximately20 cm from the dorsal thorax. The green electrode was placed 5 cm lateral to thesternum on the left and the black 0-electrode on the left side of the thorax,15 cm below the red electrode.

ECGs were recorded prior to sedation and euthanasia and during euthanasia.ECG monitoring during euthanasia commenced at the time that pentobarbital injec-tion was initiated. Monitoring ceased at 1200 s (20 min). As soon as asystole ap-peared and reflexes were absent, the animal was considered dead. In nine cases,perineal reflexes continued after asystole, but the animal was considered dead atthe time of asystole.

Measurements of ECG variables were performed directly on the computer after-wards when all horses had been euthanized. RR intervals, PQ duration, QRS dura-tion, distance from start of QRS complex to end of T wave (QT interval), durationof T-wave from peak to end (TpeakTend) and T wave amplitude (Tpeak; Fig. 1) weremeasured on lead II. These measurements were performed before euthanasia (time0) and repeated every minute for the first 5 min after the commencement of pento-barbital injection (time 1, 2, 3, 4, 5 min), followed by measurements at 10 and15 min after pentobarbital injection. If asystole occurred before 15 min, measure-ments stopped after asystole occurred. The QT interval was corrected for HR byBazett’s formula (Bazett, 1920):

QTB ¼QTffiffiffiffiffiffi

RRp

and was named QTc. All ECGs were manually checked for arrhythmias according topreviously described definitions (Buhl et al., 2010).

Statistics

Analysis of variance (ANOVA) using a general linear model was performed toevaluate the effect of pentobarbital dose (high/low) and sedation (present/absent)on time taken to collapse, auscultable heartbeat to cease, asystole to occur and re-flexes to be lost. The effect of dose (high/low), sedation (present/absent) and time(time 0, 1, 2, 3, 4, 5, 10 and 15 min) on RR, PQ, QRS, QTc, TpeakTend and Tpeak was alsoevaluated using ANOVA and a general linear model (PROC GLM, SAS 9.2, StatisticalAnalysis System). To compare differences between time 0 (before euthanasia) andafter euthanasia (time 1, 2, 3, 4, 5, 10 and 15 min), Dunnett’s test was used. As onlyone horse had electrocardiographic activity 15 min after pentobarbital injection,statistical analysis was not performed on data at 15 min after injection. P-valuesof <0.05 were considered statistically significant. Box-and-whisker plots were per-formed to illustrate the data (Graph Pad Prism version 5.04).

Results

Mean injection time was 17 s (range 6–45 s). The allocation ofhorses to the four groups is shown in Table 1.

Collapse

s

High dose w

ith se

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

20

40

60

80 Auscultable heart beat

s

High dose w

ith se

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

500

1000

1500

Asystole

s

High dose w

ithse

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

500

1000

1500Palpebral reflex

s

High dose w

ithse

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

100

200

300

Corneal reflex

s

High dose w

ithse

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

500

1000

1500Perineal reflex

s

High dose w

ith se

dation

High dose no se

dation

Low dose w

ith se

dation

Low dose no se

dation

0

500

1000

1500

Fig. 2. Graphical illustration of descriptive distribution of the clinical variable among the four groups of horses shown by box-and-whisker plots. The end of the whiskers isrepresenting minimum and maximum values, the rectangle represents the upper and lower quartiles, and the median is represented by the horizontal bar.

R. Buhl et al. / The Veterinary Journal 196 (2013) 483–491 485

Clinical findings during euthanasia

Clinical results for the four groups are shown in Fig. 2. The effectof pentobarbital dose (high or low) and sedation (present/absent)on time to collapse, auscultation of heart beat, time to asystoleand cessation of reflexes is shown in Table 2. Sedated horses tooksignificantly longer to collapse (31 ± 1 s) than unsedated horses

(23 ± 1 s; P < 0.0001; mean difference 8 s), but asystole occurredsignificantly faster in the sedated horses (518 ± 38 s in sedatedhorses vs. 740 ± 38 in unsedated horses; P < 0.0001; mean differ-ence 3 min 44 s). Clinical signs of death, indicated by collapse, ces-sation of auscultable heart beat and absence of palpebral andcorneal reflexes, occurred significantly faster than cardiac deathdefined as asystole (P < 0.05).

Table 2Repeated measurements analysis of variance of time (s) to collapse, time to noauscultable heart beat, time to asystole, time to cessation of palpebral, corneal andperineal reflexes. The results are given as means ± standard error (SE) and P-valueswere calculated from the ANOVA using the least square method.

Outcome Exposurevariable

Level Mean ± SE P-value

Collapse (s) Pentobarbitaldose

High 26 ± 1

Low 27 ± 1 0.36Sedation Absent 23 ± 1

Present 31 ± 1 <0.0001

Auscultable heartbeat(s)

Pentobarbitaldose

High 84 ± 25

Low 272 ± 29 <0.0001Sedation Absent 196 ± 27

Present 160 ± 27 0.35

Asystole (s) Pentobarbitaldose

High 653 ± 35


Present 518 ± 38 <0.0001

Palpebral reflex (s) Pentobarbitaldose

High 111 ± 7


Present 73 ± 7 <0.001

Corneal reflex (s) Pentobarbitaldose

High 161 ± 24


Present 145 ± 26 <0.05

Perineal reflex (s) Pentobarbitaldose

High 448 ± 47


Present 617 ± 50 <0.05


Electrocardiographic findings

In 4/29 cases, parts of post-euthanasia ECGs were missing be-cause of poor recordings, but in the remaining 25 cases, consis-tently high quality ECGs were obtained. The distribution of ECGvariables in the four groups is shown in Fig. 3 and the effects ofpentobarbital dose and sedation are shown in Table 3.

RR interval decreased significantly during the first 3 min ofeuthanasia (P < 0.05), but this was followed by an increase. MeanHR before euthanasia was 41 bpm, which increased to approxi-mately 60 bpm during the first 3 min of euthanasia, but then de-creased to 30–48 bpm at 10–15 min after the injectioncommenced. HR decreased in horses treated with high dose pento-barbital compared to those treated with low dose pentobarbital(P < 0.0001; Table 3).

Time taken from depolarization of the atria to the initiation ofventricular depolarization (represented by PQ interval) decreasedsignificantly during the first 5 min of euthanasia (P < 0.05–<0.0001) before an increase was observed. High dose pentobarbitaland sedation resulted in a significant increase in PQ interval(P < 0.0001 and P < 0.05, respectively; Table 3). An absence of Pwaves was observed in all horses after 5 min (Fig. 4). Sedation re-sulted in significant prolongation of QRS duration (P < 0.05; Ta-ble 3). QT duration corrected for heart rate (QTc) prolongedsignificantly during the first 3 min of euthanasia compared topre-euthanasia values (P < 0.05). High dose pentobarbital resultedin prolonged QT duration compared to low dose pentobarbital(P < 0.001; Table 3).

Before euthanasia, T wave morphology varied between horsesand a biphasic T wave was often observed. However, within few

seconds after the initiation of euthanasia, T waves became unipha-sic and positive in all horses (Fig. 5) and both TpeakTend and Tpeak

were significantly different from pre-euthanasia values (P < 0.05and P < 0.0001, respectively). High dose pentobarbital and sedationof horses significantly increased TpeakTend (P < 0.0001 and P < 0.001)and Tpeak amplitude (P < 0.0001 and P < 0.05; Table 3).

In two horses, ventricular premature complexes were observedduring post-euthanasia ECGs and one horse showed isolated supra-ventricular premature complexes. In all horses, gradual alterationsin the morphology of QRS and T complexes were observed, makingtheir classification as true arrhythmias difficult. Alterations, whichcan be characterized as ventricular tachycardia, were seen severaltimes in post-euthanasia ECGs, but could be due to either ventric-ular tachycardia or alterations of supraventricular beats (Fig. 4).

Discussion

In the present study, all ECGs obtained during euthanasiashowed that cardiac electrical activity continued after the clinicaldetermination of death. These observations are in agreement witha previously reported canine study (Evans et al., 1993). Since car-diac function continues beyond the loss of consciousness and thecessation of an auscultable heartbeat, the traditional clinical decla-ration of death in horses could be considered premature comparedto human standards. Euthanasia should result in cardiac death, butthis study has demonstrated that this occurs after clinical signs oflife are no longer present.

Respiratory arrest is also considered an important parameterwhen examining animals for vital signs (Sinclair, 2001). A limita-tion of this study is that respiratory rate and time to respiratory ar-rest were not recorded. For most horses, we observed thatrespiratory arrest occurred simultaneously with collapse and wasoften followed by one or two terminal gasps. These observationsare supported by the pharmacological action of pentobarbital,which has a rapid onset and acts primarily by depression of theCNS, including the respiratory centre, with secondary depressionof the cardiovascular system (Possner and Burns, 2009).

The discrepancy between clinical and electrocardiographicdetermination of death could be due to the development of pulse-less electrical activity (PEA), which is described in human medicineand is characterized by unconsciousness and lack of palpable pulseand other clinical signs in the presence of organized cardiac electri-cal activity, but absent or strongly reduced cardiac contractionsvisualized by echocardiography. Although the prognosis in thesepatients is poor, especially if ECG characteristics are abnormal(Aufderheide, 2007), correct resuscitation is sometimes successful.From the present study it is not possible to conclude whether thehorses demonstrate PEA, as we did not palpate peripheral pulses,examine cardiac contractility using echocardiography, or measureblood pressure, but theoretically PEA could occur in horses.

It is impossible to know whether the horses were able to regis-ter pain or distress from the commencement of the pentobarbitalinjection until cardiac death occurred up to 15 min later. It wouldtherefore be helpful to record brain activity during euthanasia untilcardiac death occurred. Brain activity can be measured using elec-troencephalography (EEG) and has been described in horses,although not during euthanasia (Haga and Dolvik, 2005; Williamset al., 2012). Chalifoux and Dallaire (1983) showed in dogs euthan-ased with carbon monoxide that EEG recordings were isoelectric4 min after euthanasia, which occurred earlier than the cessationof cardiac electrical activity measured by ECG (at 19 min). It wasconcluded that the dogs were unconscious before cardiac electricalactivity ceased, which indicated that brain activity was present fora maximum of 4 min after euthanasia, during which time the dogscould theoretically experience distress.

High dose with sedation

Time (min)

PQ (m

s)

0 1 2 3 4 5 100

100

200

300

400 High dose no sedation

Time (min)

PQ (m

s)

0 1 2 3 4 5 100

100

200

300

400

Low dose with sedation

Time (min)

PQ (m

s)

0 1 2 3 4 5 100

100

200

300

400Low dose no sedation

Time (min)

PQ (m

s)

0 1 2 3 4 5 100

100

200

300

400

PQ


Time (min)

RR

(ms)

0 1 2 3 4 5 100

2000

4000

6000High dose no sedation

Time (min)

RR

(ms)

0 1 2 3 4 5 100

2000

4000

6000


Time (min)

RR

(ms)

0 1 2 3 4 5 100

2000

4000

6000 Low dose no sedation

Time (min)

RR

(ms)

0 1 2 3 4 5 100

2000

4000

6000

RR

Fig. 3. Graphical illustration of descriptive distribution of development of the electrocardiographic variable among the four groups of horses over time shown by box-and-whisker plots. The end of the whiskers is representing minimum and maximum values, the rectangle represents the upper and lower quartiles, and the median is representedby the horizontal bar. PQ (ms), distance from P to Q wave; RR (ms), distance between two R waves; QRS (ms), duration of QRS complex; QTc (ms), distance from Q to the end ofthe T wave (Tend) corrected for HR; TpeakTend (ms), duration from top of the T wave (Tpeak) until the T wave reaches baseline (Tend); Tpeak (mV), amplitude of the T wave.


Since Purkinje fibers are widespread within the equine myocar-dium, ECG tracings have mainly been used in horses for the diag-

nosis of arrhythmias (Hamlin and Smith, 1965; van Loon andPatteson, 2010). We propose that ECG measurements such as


Time (min)

QR

S (m

s)

0 1 2 3 4 5 100

50

100

150


Time (min)

QR

S (m

s)

0 1 2 3 4 5 100

50

100

150

200


Time (min)

QR

S (m

s)

0 1 2 3 4 5 100

50

100

150


Time (min)

QR

S (m

s)

0 1 2 3 4 5 10

100

150

200

QRS


Time (min)

QT c (

ms)

0 1 2 3 4 5 100

200

400


Time (min)

QT c (

ms)

0 1 2 3 4 5 100

200

400

600


Time (min)

QT c (

ms)

0 1 2 3 4 5 100

200

400


Time (min)

QT c (

ms)

0 1 2 3 4 5 100

200

400

600

QTc

Fig. 3. (continued)


QRS duration, QT interval and T wave morphology should beroutinely measured on equine ECGs, as this study clearlydemonstrates significant changes can occur in equine ECGmorphology.

High dose pentobarbital caused significant prolongation of PQ,QTc, TpeakTend and Tpeak interval in this study, which indicates thatpentobarbital causes dose-related depression of cardiac function.The QT prolongation reported here could be due to several factors


Time (min)

T peak

T end

(ms)

0 1 2 3 4 5 100

50

100

150


Time (min)

T peak

T end

(ms)

0 1 2 3 4 5 100

50

100

150

200


Time (min)

T peak

T end

(ms)

0 1 2 3 4 5 100

50

100

150


Time (min)

T peak

T end

(ms)

0 1 2 3 4 5 100

50

100

150

200

TpeakTend


Time (min)

T peak

(mV)

0 1 2 3 4 5 100

1

2


Time (min)

T peak

(mV)

0 1 2 3 4 5 100

1

2

3


Time (min)

T peak

(mV)

0 1 2 3 4 5 100

1

2


Time (min)

T peak

(mV)

0 1 2 3 4 5 100

1

2

3

Tpeak

Fig. 3. (continued)


(Al-Khatib et al., 2003). Pentobarbital is known to produce a dose-related CNS depression, including respiratory inhibition (Possnerand Burns, 2009), which could cause QT prolongation secondaryto loss of cardiac function, as seen in a study of dogs with marked

myocardial depression (Manders and Vatner, 1976). Furthermore,pentobarbital is known to block the cardiac hERG (human Ether-à-go-go-Related Gene) potassium channel, which is the main tar-get for drug-induced QT prolongation (Kanters et al., 2006). This

Table 3Repeated measurements analysis of variance of the electrocardiographic measure-ments PQ, RR, QRS, QTc, TpeakTend and Tpeak. The results are given as means ± standarderror (SE) and P-value was calculated from the ANOVA with least square method.

Outcome Exposure variable Level Mean ± SE P-value

PQ (ms) Pentobarbital High 206 ± 9Low 158 ± 8 <0.0001

Sedation Absent 190 ± 8Present 173 ± 9 <0.05

RR (ms) Pentobarbital High 1464 ± 91Low 1101 ± 93 <0.0001

Sedation Absent 1256 ± 89Present 1310 ± 95 0.56

QRS (ms) Pentobarbital High 120 ± 4Low 115 ± 4 0.27


QTc (ms) Pentobarbital High 444 ± 9Low 403 ± 9 <0.0001

Sedation Absent 423 ± 9Present 425 ± 9 0.81

TpeakTend (ms) Pentobarbital High 134 ± 3Low 113 ± 4 <0.0001


Tpeak (mV) Pentobarbital High 1.26 ± 0.10Low 0.83 ± 0.10 <0.0001

Sedation Absent 0.94 ± 0.10Present 1.15 ± 0.11 <0.05

Ventricu

Fig. 4. ECG obtained during euthanasia showing example of ventricular premature beats.

Start injection of pentobarbital

Biphasic T wave

Fig. 5. ECG obtained from start of injection of pentobarbital. Note the clear changes of T wspeed 25 mm/s, gain 10 mm/mV.


is supported by the prolongation of the terminal part of theT-wave, the TpeakTend interval, which was prolonged in a dosedependent way by pentobarbital. Despite the QT prolongation,pentobarbital is not considered arrhythmogenic due to its rela-tively homogenous blockade of the myocardial wall (Kanterset al., 2006). However, the present study shows increased T-waveamplitude, which is a sign of increased myocardial dispersionand possible hypoxia (Hunt, 2002; Antzelevitch, 2006), which canbecome a proarrhythmic substrate. This effect was probably notsubstantial in the present study, as we found only few arrhythmiasduring euthanasia.

Combined with detomidine, which induces bradycardia, hypo-tension and reduced cardiac contractility (Sarazan et al., 1989;Buhl et al., 2007), we expected the effect of pentobarbital and sub-sequent collapse to be delayed in the sedated horses due to cardio-vascular depression and reduced cardiac output by detomidine.This was confirmed and is in agreement with Knottenbelt et al.(1994). Due to the depressive effect of detomidine on the cardio-vascular system, we also suspected sedation would prolong timeto asystole, but surprisingly, asystole appeared almost 4 min ear-lier in horses receiving sedation compared to unsedated horses.An explanation for this could be that detomidine induces hypoten-sion, which decreases cardiac output more rapidly and thereforeprecipitates asystole. Among clinicians there is a tendency not tosedate horses prior euthanasia since it prolongs the time to col-lapse. However, our study shows that sedation with detomidine re-duces the time to cardiac death and therefore sedation should berecommended.

lar premature beats

P waves (stars) are not present consistently. Paper speed 25 mm/s, gain 10 mm/mV.

Collapse of horse

Positive T wave

ave morphology from biphasic to highly positive and with increased duration. Paper


Conclusions

Clinical definition of death ceases significantly earlier than car-diac death as defined by asystole. Although sedated horses tookapproximately 8 s longer to collapse than unsedated horses, asys-tole occurred almost 4 min earlier in sedated horses, which indi-cates a synergistic effect of pentobarbital and detomedine.Furthermore, a significant dose-dependent effect of pentobarbitalwas shown for ECG variables, with marked QT prolongation andchanges in T-wave morphology. These ECG results provide impor-tant information about the ECG changes in horses during death andare essential for future studies of ECG changes in diseased horses.High dose pentobarbital combined with sedation resulted in fastercardiac death and is therefore preferable when euthanasing horses.This study raises the important ethical question of whether horsesbeing euthanased may experience pain or distress during the per-iod immediately before cardiac death and whether they could the-oretically recover from unconsciousness during this terminalphase, both of which are unacceptable. However, as horses shouldbe deeply anaesthetised during the procedure, it is unlikely thatthey do experience pain. Further studies need to be conducted toanalyze and fully document brain and cardiac activity during theprocess of death in horses.

Conflict of interest

None of the authors of this paper has a financial or personalrelationship with other people or organisations that could inappro-priately influence or bias the content of the paper.

Acknowledgements

The authors would like to thank the animal welfare organiza-tion Hestens Værn for economical support for destruction of horsesafter euthanasia. Funding was obtained from the Fraenkel Founda-tion and the Danish Council for Strategic Research.

References

Al-Khatib, S.M., LaPointe, N.M., Kramer, J.M., Califf, R.M., 2003. What cliniciansshould know about the QT interval. Journal of the American Medical Association289, 2120–2127.

Antzelevitch, C., 2006. Cellular basis for the repolarization waves of the ECG. Annalsof the New York Academy of Sciences 1080, 268–281.

Aufderheide, T.P., 2007. Etiology, electrophysiology and myocardial mechanics ofpulseless electrical activity. In: Cardiac Arrest. The Science and Practice ofResuscitation Medicine, Second Ed. Cambridge University Press, Cambridge,UK, pp. 426–446.

Ayala, I., Gutierrez-Panizo, C., Benedito, J.L., Prieto, F., Montes, A., 2000. Morphologyand amplitude values of the electrocardiogram of Spanish-bred horses ofdifferent ages in the Dubois leads system. Veterinary Research 31, 347–354.

Bazett, H.C., 1920. An analysis of the time-relations of electrocardiograms. Heart 7,350–353.

Buhl, R., Ersbøll, A.K., Larsen, N.H., Eriksen, L., Koch, J., 2007. The effects ofdetomidine, romifidine or acepromazine on echocardiographic measurementsand cardiac function in normal horses. Veterinary Anaesthesia and Analgesia34, 1–8.

Buhl, R., Meldgaard, C., Barbesgaard, L., 2010. Cardiac arrhythmias in clinicallyhealthy showjumping horses. Equine Veterinary Journal 42, 196–201.

Chalifoux, A., Dallaire, A., 1983. Physiologic and behavioral evaluation of COeuthanasia of adult dogs. American Journal of Veterinary Research 44, 2412–2417.

Doig, C.J., Burgess, E., 2003. Brain death: Resolving inconsistencies in the ethicaldeclaration of death. Canadian Journal of Anaesthesia 50, 725–731.

Evans, A.T., Broadstone, R., Stapleton, J., Hooks, T.M., Johnston, S.M., McNeil, J.R.,1993. Comparison of pentobarbital alone and pentobarbital in combinationwith lidocaine for euthanasia of dogs. Journal of the American VeterinaryMedical Association 203, 664–666.

Haga, H.A., Dolvik, N.I., 2005. Electroencephalographic and cardiovascular variablesas nociceptive indicators in isoflurane-anaesthetized horses. VeterinaryAnaesthesia and Analgesia 32, 128–135.

Hamlin, R.L., Smith, C.R., 1965. Categorization of common domestic mammals basedupon their ventricular activation process. Annals of the New York Academy ofSciences 127, 195–203.

Hubbell, J.A.E., Muir, W.W., Gaynor, J.S., 1993. Cardiovascular effects of thoraciccompression in horses subjected to euthanasia. Equine Veterinary Journal 25,282–284.

Hunt, A.C., 2002. T wave alternans in high arrhythmic risk patients: Analysis in timeand frequency domains: A pilot study. BMC Cardiovascular Disorders 2, 6.

Jones, R.S., 1992. Euthanasia in horses. Equine Veterinary Education 4, 154–157.Kanters, J.K., Graff, C., Andersen, M.P., Hardahl, T., Toft, E., Christiansen, M., Bloch

Thomsen, P.E., Struijk, J.J., 2006. Long QT syndrome genotyping byelectrocardiography: Fact, fiction, or something in between? Journal ofElectrocardiology 39, 119–122.

Kerridge, I.H., Saul, P., Lowe, M., McPhee, J., Williams, D., 2002. Death, dying anddonation: Organ transplantation and the diagnosis of death. Journal of MedicalEthics 28, 89–94.

Knottenbelt, D.C., 1995. Euthanasia of horses – Alternatives to the bullet. In Practice17, 462–464.

Knottenbelt, D.C., Jones, R.S., Brazil, T.J., Proudman, C.J., Edwards, S.R., Harrison, L.J.,1994. Humane destruction of horses with a mixture of quinalbarbitone andcinchocaine. Veterinary Record 134, 319–324.

Manders, W.T., Vatner, S.F., 1976. Effects of sodium pentobarbital anesthesia on leftventricular function and distribution of cardiac output in dogs, with particularreference to the mechanism for tachycardia. Circulation Research 39, 512–517.

Matthews, N.S., Hartsfield, S.M., 2004. Arrhythmogenic dose of epinephrine inisoflurane or sevoflurane anaesthetized horses. Journal of Equine VeterinaryScience 24, 110–114.

Morgan, R.A., Raftery, A.G., Cripps, P., Senior, J.M., McGowan, C.M., 2011. Theprevalence and nature of cardiac arrhythmias in horses following generalanaesthesia and surgery. Acta Veterinaria Scandinavia 53, 3–8.

Possner, L.P., Burns, P., 2009. Injectable anesthetic agents. In: VeterinaryPharmacology and Therapeutics, Ninth Ed. Wiley-Blackwell, Iowa, USA, pp.265–300.

Sarazan, R.D., Starke, W.A., Krause, G.F., Garner, H.E., 1989. Cardiovascular effects ofdetomidine a new a2-adrenoceptor agonist in the conscious pony. Journal ofVeterinary Pharmacology and Therapeutics 12, 378–388.

Sinclair, L., 2001. Humane euthanasia. Veterinary Technician 22, 370–374.Steere, J.H., 1990. Equine euthanasia. Journal of the American Veterinary Medical

Association 197, 555–556.Van Loon, G., Patteson, M., 2010. Electophysiology and Arrhythmogensis. In:

Cardiology of the Horse, Second Ed. Saunders Elsevier, St. Louis, MO, USA, pp.59–75.

Williams, D.C., Aleman, M., Tharp, B., Fletcher, D.J., Kass, P.H., Steffey, E.P., Lecouteur,R.A., Holliday, T.A., 2012. Qualitative and quantitative characteristics of theelectroencephalogram in normal horses after sedation. Journal of VeterinaryInternal Medicine 26, 645–653.

Zamperetti, N., Bellomo, R., Ronco, C., 2003. Defining death in non-heart beatingorgan donors. Journal of Medical Ethics 29, 182–185.

the veterinary journal - kasetsart...

Documents