SHORT COMMUNICAT ION

The use of electrical stimulation to guide epidural and

intrathecal needle advancement at the L5-L6 intervertebral

space in dogs

Pablo E Otero*, Natali Verdier*, Martin R Ceballos*, Lisa Tarragona*, Myriam Flores* & Diego A Portela†

*Anaesthesiology Department, College of Veterinary Medicine, University of Buenos Aires, Buenos Aires, Argentina

†Department of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY, USA

Correspondence: Pablo E Otero, Anaesthesiology Department, College of Veterinary Medicine, University of Buenos Aires, Av. Chorroar�ın

280, C1427CWO Buenos Aires, Argentina. E-mail: potero@fvet.uba.ar

Abstract

Objective To determine the minimal electrical

threshold (MET) necessary to elicit appropriate

muscle contraction when the tip of an insulated

needle is positioned epidurally or intrathecally at the

L5-6 intervertebral space (phase-I) and to determine

whether the application of a fixed electrical current

during its advancement could indicate needle entry

into the intrathecal space (phase-II) in dogs.

Study design Prospective, blinded study.

Animals Thirteen (phase-I) and seventeen (phase-

II) dogs, scheduled for a surgical procedure where

L5-6 intrathecal administration was indicated.

Methods Under general anesthesia, an insulated

needle was first inserted into the L5-6 epidural space

and secondly into the intrathecal space and the MET

necessary to obtain a muscular contraction of the

pelvic limb or tail at each site was determined (phase-

I). Under similar conditions, in dogs of phase-II an

insulated needle was inserted through the L5-6intervertebral space guided by the use of a fixed

electrical current (0.8 mA) until muscular contrac-

tion of the pelvic limb or tail was obtained. Intrathe-

cal needle placement was confirmed by either free

flow of cerebrospinal fluid (CSF) or myelography.

Results The current required to elicit a motor

response was significantly lower (p < 0.0001) when

the tip of the needle was in the intrathecal space

(0.48 � 0.10 mA) than when it was located epi-

durally (2.56 � 0.57). The use of a fixed electrical

stimulation current of 0.8 mA resulted in correct

prediction of intrathecal injection, corroborated by

either free flow of CSF (n = 12) or iohexol distribu-

tion pattern (n = 5), in 100% of the cases.

Conclusion and clinical relevance Nerve stimula-

tionmay be employed as a tool to distinguish epidural

from intrathecal insulated needle position at the L5-6intervertebral space in dogs. This study demonstrates

the feasibility of using an electrical stimulation test to

confirm intrathecal needle position in dogs.

Keywords dog, electrical threshold, epidural, intra-

thecal, nerve stimulation test.

Introduction

Intrathecal anesthesia has some advantages over

epidural anesthesia, including faster onset and offset

of action, lower systemic absorption of drugs and the

possibility to use different weight solutions to target

the metameric region to be blocked (Sarotti et al.

2013). However, the difficulty in performing the

technique may be compounded by the lack of a free

flow of cerebrospinal fluid (CSF) that can be used to

confirm correct needle placement. In dogs, Sarotti

et al. (2011) reported a failure rate of 26% during

the first stage of learning curve, which improved to

10% in experienced hands.

1

Veterinary Anaesthesia and Analgesia, 2014 doi:10.1111/vaa.12137

Nerve stimulation has been used to confirm

epidural or intrathecal needle placement in humans,

pigs (Tsui et al. 2004, 2005), rabbits, cats (Otero

et al. 2012, 2013) and dogs (Garcia-Pereira et al.

2010). Moreover, it has been proposed to confirm

intrathecal injection in the absence of CSF in the

needle hub (Tsui 2006; Otero et al. 2012). Although

the minimal electrical threshold (MET) necessary to

elicit pelvic limb or tail muscle contraction when the

tip of an insulated needle is located in the epidural

space at the lumbosacral region has been reported in

dogs (Garcia-Pereira et al. 2010), to our knowledge

approach through a lumbar intervertebral space has

not been studied.

The purpose of the two stage study reported here

was to determine the MET necessary to elicit

appropriate muscle contraction when the tip of an

insulated needle is positioned epidurally and intra-

thecally at the L5-6 intervertebral space (phase I) and

to determine whether the application of a fixed

electrical current during the advancement of an

insulated needle could be used to confirm the needle

tip introduction into the intrathecal space (phase II)

in dogs.

Materials and methods

Animals

This study was approved by the Institutional Animal

Care and Use Committee, College of Veterinary

Medicine, University of Buenos Aires (Project num-

ber 2012/14). Thirteen (phase I) and seventeen

(phase II) dogs, admitted to the Veterinary Teaching

Hospital for different surgical procedures during

which intrathecal anesthesia was scheduled as part

of the analgesic protocol, were enrolled. Written

informed consent was obtained from the owners

prior to participation in the study.

Anesthesia and preparation of animals

Dogs were premedicated with acepromazine (0.02

mg kg�1; Inadrin; Laboratorios Richmond, Argen-

tina) plus morphine (0.3 mg kg�1; Analmorph; Dr.

Gray S.A.C.I., Argentina) by intramuscular (IM)

injection. Fifteen minutes after premedication, a

cannula was placed in a cephalic vein and anesthesia

was induced with propofol (1–5 mg kg�1; Propovet;

Laboratorios Richmond) administered intravenously

to effect, followed by tracheal intubation. Anesthesia

was maintained with isoflurane (Forane; Abbott

Laboratories, Argentina) in oxygen using a circle

breathing system allowing spontaneous breathing.

Throughout anesthesia, heart rate, respiratory rate,

non-invasive arterial blood pressure, end-tidal car-

bon dioxide concentration, end-tidal isoflurane con-

centration, esophageal temperature, hemoglobin

oxygen saturation and lead II electrocardiogram

were continuously monitored (VET 420F; Goldway

US, NY, USA). When anesthesia was stable, the hair

over L3-S1 vertebrae was clipped and the skin

aseptically prepared. Animals were placed in sternal

recumbency, with the pelvic limbs flexed at the hips.

The use of a cradle allowed flexion of the lumbar spine

to facilitate epidural and intrathecal access through a

paramedian approach at the L5-6 intervertebral

space, as previously described (Otero & Campoy

2013).

Phase-I (Epidural and intrathecal MET

determination)

A 30° short bevel 22-gauge, 50 mm insulated

needle without its extension line (SonoPlex Stim

cannula; Pajunk GmbH, Germany) was inserted into

the epidural space, using the loss of resistance

technique with air (limited to 0.03 mL kg�1). When

the epidural space was identified, the needle hub was

observed for 60 seconds for absence of CSF. Then, a

second investigator connected the needle to a

peripheral nerve stimulator (NS) (MultiStim

SWITCH, Pajunk, Dyna Medical Corp, UK). The

negative lead of the NS was attached to the insulated

needle, while the positive lead was placed on the skin

over the caudal aspect of the thigh. The NS was set to

a frequency of 2 Hz, pulse width of 0.1 ms, and the

current was gradually increased from 0 mA and by

0.01 mA until pelvic limb or tail motor response was

elicited. When these muscle twitches were evident,

the delivered stimulating current was recorded as

METepi. The NS was then turned off and the needle

was advanced until CSF was obtained. A separate

investigator turned on the NS for MET determina-

tion. The procedure was abandoned if CSF was not

present. The delivered stimulating current necessary

to elicit pelvic limb or tail motor response was

recorded as METint as described above. Before needle

withdrawal, a solution of 0.5% hyperbaric bupiva-

caine (Caina-G Hiperb�arica; Dr. Gray S.A.C.I.) con-

taining 0.1 mg mL�1 of morphine (Duramorph; Dr.

Gray S.A.C.I.) in a total volume of 0.05 mL kg�1

was administered intrathecally to all dogs in prep-

aration for the surgical procedure.

© 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia2

Neuraxial electrical stimulation in dogs PE Otero et al.

Phase-II (Electrical stimulation test)

After analyzing the results of phase I, a new group of

seventeen dogs was enrolled. Anesthesia and instru-

mentation were performed as in phase I. For this

group a 30° short bevel 24-gauge 25 mm or a

22-gauge 50 mm insulated needle without their

extension lines were used. After the needle pene-

trated the skin, a separate investigator connected the

NS which, based on the phase I results, was set to a

fixed current of 0.8 mA, frequency of 2 Hz and pulse

width of 0.1 ms. The needle was advanced into the

intrathecal space. As soon as muscle twitching was

detected, needle advancement was stopped and the

current was lowered by 0.01 mA to determine the

threshold for motor activity (METphase-II). The NS

was then turned off and if CSF was present in the hub

of the needle, 0.05 mL kg�1 of the same bupiva-

caine and morphine solution used in phase I was

injected. If CSF was absent (dry tap), the anesthetic

solution was injected followed by 1–2 mL of iohexol

and a radiograph of the lumbar area was obtained to

confirm intrathecal injection.

Complications such as prolonged neural deficit or

pain around the injection site were noted until

the dogs were discharged from the hospital

24–48 hours after surgery.

Statistical analysis

After the normality distribution of data was con-

firmed (Shapiro-Wilk test), statistical significance

between METepi and METint was determined using a

paired t-test. Differences were judged to be signifi-

cant when p < 0.05. Data distributed normally are

presented as mean � SD values.

Accuracy of the electrical stimulation test was

measured by positive predictive value (PPV). PPV

were calculated as the number of times the intra-

thecal space was reached within two SD of the

METint divided by the total number of accesses

performed. Confidence limits (95%) were calculated

according to the Clopper–Pearson method.

Results

Thirteen dogs ASA I-III (9 non-spayed females and 4

non-neutered males) weighing 16.4 � 8.4 kg and

aged 4.0 � 2.4 years and seventeen non-spayed

females dogs ASA I-II weighing 12.6 � 8.8 kg and

aged 4.4 � 3.8 years were included in phase I and

II, respectively.

All animals in phase I had two distinguishable

threshold currents as the needle was advanced

(Fig. 1). The first mean current to elicit the motor

response (METepi) was 2.56 � 0.57 mA, where no

CSF was present. The second mean current (METint)

was 0.48 � 0.10 mA and CSF was present in all

cases but one, which was excluded from data

analysis. There was a significant difference between

the epidural and intrathecal space threshold cur-

rents (p < 0.0001).

The MET values registered in phase II were

0.43 � 0.09 mA. The use of the electrical stimula-

tion test resulted in correct prediction of intrathecal

injection, corroborated by either free flow of CSF

(n = 12) or iohexol distribution pattern (n = 5), in

100% of the cases. Positive predictive value (PPV) for

successful intrathecal injection was 100%. The 95%

confidence interval extends from 80% to 100%.

In all cases the injection of local anesthetic

produced a successful spinal block. No complications

occurred in any of the animals treated.

Discussion

In this study, the mean threshold current required to

elicit a motor response when an insulated needle

was positioned epidurally was significantly higher

than that required when the needle tip was placed

intrathecally (2.56 � 0.57 mA and 0.48 � 0.10

mA, respectively). This demonstrates that a fixed low

electrical stimulation current can distinguish epidu-

ral from intrathecal insulated needle position at the

L5-6 intervertebral space in dogs.

Based upon the observed stimulating currents and

the lack of CSF when MET values were >1 mA, we

Figure 1 Minimal electrical thresholds (MET) necessary to

produce the motor response in the epidural space (METepi)

and intrathecal space (METint and METphase-II) in dogs.

*Significant differences with METint (p < 0.0001).

© 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia 3

Neuraxial electrical stimulation in dogs PE Otero et al.

concluded that, if loss of resistance was present,

these needles were most likely in the epidural space.

Confirmation of intrathecal injections was per-

formed by obtaining CSF in the hub of the needle

or by typical iohexol distribution pattern within the

subarachnoid space.

The results of this study are in agreement with

those reported in other species. In pigs (Tsui et al.

2004) and humans (Tsui et al. 2005) the METs

tested in lumbar segments of the spinal cord are

>1 mA in the epidural space and <1 mA in the

intrathecal space. Garcia-Pereira et al. (2010)

reported in dogs a MET value of 0.3 mA when the

needle was placed in the lumbosacral epidural space.

In contrast, METepi in this study was 2.56 � 0.57

mA. Based on the fact that the lesser the distance

between the nerve and the tip of the stimulating

device, the lower the MET needed to stimulate it

(Ford et al. 1984), the difference found between

these epidural MET values could be explained by

closeness of the needle to the centrally running

nerve roots at the lumbosacral level.

Air was preferred over saline when performing

loss of resistance for this study to avoid confusion

between injected saline and CSF. The amount of air

injected was limited to 0.03 mL kg�1 to minimize

hindrance to electrical conduction.

In this study, the METint was 0.48 � 0.10 mA

and did not overlap the METepi (2.56 � 0.57 mA).

Our METint values were similar to those reported in

humans (Tsui et al. 2005), pigs (Tsui et al. 2004),

rabbits (Otero et al. 2012) and cats (Otero et al.

2013). The information on needle tip position and

the different thresholds registered provided the

framework for applying a current of fixed frequency

and intensity (2 Hz, and 0.8 mA) during intrathecal

needle insertion. Based on these results, with 95%

confidence, the intensity of the electrical current

needed to perform the technique should be between

the lower METepi [1.42 mA (mean �2SD)] and the

higher METint [0.68 mA (mean +2SD)]. In this

study, every motor response obtained with 0.8 mA

during needle insertion indicated that the stimulat-

ing needle was inside the intrathecal space. More-

over, the use of the electrical stimulation test

resulted in 100% of intrathecal injections. However,

since the observed values were given under the

specific settings for the small sample studied, the

lower 95% confidence limit for the 100% PPV

reported here can still be as low as 80%. Further

studies with a larger sample size are needed to

determine the advantages of nerve stimulation when

used to confirm intrathecal needle placement.

Although the prevalence of dry tap has not been

reported in dogs, Sarotti et al. (2011, 2013) referred

a procedural failure rate, defined as absence of CSF in

the needle for unspecified cause, of 10% and 7%,

respectively. It should be noted that the occurrence

of dry tap in this study was 29.4%. Possible causes of

dry taps in this study are sternal recumbency

position with the lumbar spine curved dorsally and

the needle positioned vertically that may have acted

in opposition to CSF spontaneous free flow. Addi-

tionally, the lack of stylet could have resulted in a

blocked needle.

In summary, at the L5-6 intervertebral space in

dogs, significantly lower electrical currents are

required to elicit a muscular response when the tip

of an insulated needle is located in the intrathecal

space than when it is placed epidurally. Additionally,

the present study demonstrated the feasibility of

using an electrical stimulation test with a fixed

current of 0.8 mA to confirm intrathecal needle

position in dogs.

Acknowledgements

This work was financed by UBACyT program,

University of Buenos Aires.

References

Ford DJ, Pither C, Raj PP (1984) Comparison of insulated

and uninsulated needles for locating peripheral nerves

with a peripheral nerve stimulator. Anesth Analg 63,

925–928.

Garcia-Pereira FL, Hauptman J, Shih AC et al. (2010)

Evaluation of electric neurostimulation to confirm

correct placement of lumbosacral epidural injections in

dogs. Am J Vet Res 71, 157–160.

Otero PE, Campoy L (2013) Epidural and spinal anesthesia.

In: Small Animals Regional Anesthesia and Analgesia.

Campoy L, Read M (eds). Wiley-Blackwell, New York,

USA. pp. 227–260.

Otero PE, Portela DA, Brinkyer JA et al. (2012) Use of

electrical stimulation to monitor lumbosacral epidural

and intrathecal needle placement in rabbits. Am J Vet

Res 73, 1137–1141.

Otero PE, Zaccagnini AS, Fuensalida SE et al. (2013) Use of

electrical nerve stimulation to monitor lumbosacral

epidural needle placement in cats. Vet Anaesth Analg

doi:10.1111/vaa.12107.

Sarotti D, Rabozzi R, Corletto F (2011) Efficacy and side

effects of intraoperative analgesia with intrathecal

© 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia4

Neuraxial electrical stimulation in dogs PE Otero et al.

bupivacaine and levobupivacaine: a retrospective study

in 82 dogs. Vet Anaesth Analg 38, 240–251.

Sarotti D, Rabozzi R, Franci P (2013) A retrospective study

of efficacy and side effects of intrathecal administration of

hyperbaric bupivacaine and morphine solution in 39

dogs undergoing hind limb orthopaedic surgery. Vet

Anaesth Analg 40, 220–224.

Tsui BC (2006) Verifying spinal needle location in the

presence of a “dry tap”. Can J Anaesth 53, 424–425.

Tsui BC, Wagner A, Finucane B (2004) The threshold

current in the intrathecal space to elicit motor response is

lower and does not overlap that in the epidural space: a

porcine model. Can J Anaesth 51, 690–695.

Tsui BC, Wagner AM, Cunningham K et al. (2005) Can

continuous low current electrical stimulation distinguish

insulated needle position in the epidural and intrathecal

spaces in pediatric patients? Paediatr Anaesth 15, 959–

963.

Received 5 August 2013; accepted 30 October 2013.

© 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia 5

Neuraxial electrical stimulation in dogs PE Otero et al.