1032 Scientific Reports JAVMA, Vol 238, No. 8, April 15, 2011
is known that to avoid the development of support-limb
laminitis, an increase in comfort on the affected limb
should be provided.2,3 In experimental4,5 and clinical
studies,6 the efficacy and safety of epidural analgesia in
the reduction of signs of severe hind limb pain in horses
has been demonstrated; as such, catheters are commonly
placed to allow repeated epidural administration.6,7 In
the forelimb, no local treatment is used clinically to pro-
vide postoperative analgesia.
Continuous regional anesthesia is used to con-
trol signs of pain in human8,9 and small animal pa-
tients.10–12 Clear, objective data support its use postop-
eratively, with a decrease in signs of pain, a decrease in
dose and frequency of administration of pain medica-
tion, and faster rehabilitation.8–12 In horses, a method
has been described experimentally for CPNB.13 The
CPNB technique in horses reduced hoof-withdrawal
lthough objective data are lacking that prove that the
effects of severe pain are detrimental for horses,1 it
Continuous peripheral neural blockade
to alleviate signs of experimentally induced
severe forelimb pain in horses
Ashlee E. Watts, dvm, dacvs; Alan J. Nixon, bvsc, ms, dacvs; Heidi L. Reesink, vmd;
Jonathan Cheetham, vetmb, phd, dacvs; Susan L. Fubini, dvm, dacvs; Andrea L. Looney, dvm, dacva
Objective—To investigate the efficacy and safety of a low-volume, single-catheter, continu-
ous peripheral neural blockade (CPNB) technique to locally deliver bupivacaine to alleviate
signs of severe forelimb pain resulting from experimentally induced tendonitis in horses.
Design—Randomized controlled experimental trial.
Sample—14 horses and 5 forelimbs from equine cadavers.
Procedures—Horses underwent collagenase-induced superficial digital flexor tendonitis
in the midmetacarpal region of 1 forelimb. To deliver analgesia, a closed-tip catheter was
placed from lateral to medial, approximately 12 cm distal to the accessory carpal bone,
between the suspensory ligament and accessory ligament of the deep digital flexor ten-
don. Success of catheter placement and anesthetic delivery was documented ex vivo in
5 forelimbs from equine cadavers. Effective analgesia in affected forelimbs of horses from
continuous (n = 7) versus intermittent (7) local anesthetic delivery (intermittent peripheral
neural blockade; IPNB) was compared over a 3-day period.
Results—Horses that received CPNB in the affected forelimb were less lame than
horses that received IPNB. A lower proportion of CPNB-treated horses had behavioral
and physiologic signs of pain, compared with IPNB-treated horses. Neither technique
completely blocked the sensation of pain or resulted in swelling in the distal portion of
the forelimb, vasodilation, or an increase in lameness. After removal, Staphylococcus
aureus was cultured from 1 catheter tip.
Conclusions and Clinical Relevance—For short-term treatment, CPNB was more effective
than IPNB for reduction in signs of severe pain in the distal aspect of the forelimb of horses.
(J Am Vet Med Assoc 2011;238:1032–1039)
response to both pressure and electrical stimulation.13
Despite its possible value, equine clinicians have been
wary of the technique. This may be because of vaso-
dilation and swelling that have occurred in the distal
aspect of the limb13 and concern that complete lack of
pain may lead to limb overuse and subsequent damage
to an already injured limb.1 A third concern, especially
important to the equine orthopaedist, is that catheters
and drug delivery devices may introduce infectious
The purpose of the study reported here was to
perform a randomized controlled experimental trial to
investigate the efficacy and safety of the CPNB tech-
nique13 in treating horses for signs of severe pain re-
sulting from experimentally induced tendonitis in the
distal aspect of the forelimb. Modifications to the tech-
nique were made to reduce the cost of instrumentation,
to lessen the likelihood of complete neural blockade of
From the Comparative Orthopaedics Laboratory, Department of Clin-
ical Sciences, College of Veterinary Medicine, Cornell University,
Ithaca, NY 14853.
Funded through private donations to the Comparative Orthopaedics
Laboratory at Cornell University.
The authors thank Michael Scimeca, Holly Sparks, and Charles Claywell
for technical assistance.
Address correspondences to Dr. Watts (firstname.lastname@example.org).
Accessory ligament of the deep digital
Continuous peripheral neural blockade
Intermittent peripheral neural blockade
Superficial digital flexor tendon
JAVMA, Vol 238, No. 8, April 15, 2011 Scientific Reports 1033
the distal portion of the forelimb, to minimize the risk
of nerve injury during catheter placement, and to allow
anesthetic delivery more proximally in the metacarpus.
We hypothesized that CPNB would be safe and would
result in lower lameness scores and fewer horses with
behavioral and physiologic signs of pain at rest.
Materials and Methods
Horses—Anatomic studies to document success
of injectate delivery were performed on forelimbs of
adult Thoroughbreds and Standardbreds (n = 5) that
had been euthanized for a different study. For the in
vivo trial, young adult Thoroughbred or Thoroughbred
crossbred horses that had undergone experimentally
induced tendonitis for another study were used (me-
dian age, 4 years [range, 3 to 8 years]; median body
weight of 535 kg [1,180 lb] with a range of 415 to 565 kg
[910 to 1,240 lb]). All horses were acclimated to hous-
ing and handling areas for a minimum of 10 days prior
to the beginning of this study. The study was approved
by the university’s Institutional Animal Care and Use
Ex vivo catheter placement and anatomic studies—
Limbs for anatomic studies were harvested at the level
of the proximal aspect of the antebrachium. Prior to
forelimb harvest, 2 large stay sutures were placed in a
cruciate pattern through the muscles and skin of the
caudal antebrachium to maintain normal anatomy of
the structures in the distal portion of the forelimb.
While the forelimb was maintained in an extended
weight-bearing position, a 20-gauge, closed-tip, flexible
polyamide nylon catheter from an epidural catheter kita
was placed approximately 12 cm distal to the accessory
carpal bone (Figure 1). The catheter kit included an epi-
Figure 1—Photographic sequence of catheter placement for bupivacaine delivery to the medial and lateral palmar nerves. Through
a vertical stab incision the introducer needle for catheter insertion is placed from lateral to medial, between the SL and ALDDFT ap-
proximately 12 cm distal to the accessory carpal bone (A). The catheter is inserted, and the introducer needle is carefully backed off the
catheter (B). The introducer is placed in the subcutaneous plane from the dorsolateral aspect of the metacarpus to the stab incision and
the proximal end of the catheter is placed retrograde through the introducer needle, to exit the skin at the dorsolateral aspect of the
metacarpus (C), and is secured with a porous tape butterfly and skin suture. The proximal adapter for slip tip or luer lock attachment is
placed on the catheter end (D). For continuous delivery of 0.5% bupivacaine, a battery-powered pump is filled (E) and secured to the
distolateral aspect of the antebrachium (F and G) within the bandage (H).
1034 Scientific Reports JAVMA, Vol 238, No. 8, April 15, 2011
dural needle (Tuohy tip) for catheter insertion and an
adapter for slip tip or Luer lock connection to the proxi-
mal end of the catheter. A small (stab) vertical skin inci-
sion was made laterally between the SL and the ALDDFT.
The introducer needle was inserted through the stab in-
cision, with the rounded side directed toward the palmar
surface. The lateral palmar fascia was punctured and the
introducer needle advanced, from lateral to medial, be-
tween the SL and ALDDFT, so that the distal end was
adjacent to or through the thinner medial palmar fascia.
After confirmation of correct needle placement by pal-
pation of the needle tip medially, beneath the skin, the
stylet was removed and the catheter inserted. The intro-
ducer needle was carefully removed, leaving the distal
end of the catheter deep to the medial palmar fascia and
the more proximal catheter holes deep to the lateral pal-
mar fascia. The introducer needle was passed from the
dorsolateral aspect of the metacarpus, subcutaneously, to
exit the existing stab incision. The catheter was passed
retrograde through the introducer needle to exit the skin
at the dorsolateral aspect of the metacarpus, where it was
secured via a single suture through the skin and butter-
fly made of porous, nonelastic tape. The catheter was
trimmed so that the proximal free end could reach the
middle of the antebrachium, and the clamp adapter was
attached. A single interrupted suture was placed through
the skin to close the stab incision.
In 2 forelimbs, 7 mL of new methylene blueb was
injected via the catheter, and the forelimbs were dis-
sected for identification of dyed structures. In 3 fore-
limbs, 7 mL of new methylene blue was injected via the
catheter, and the forelimb was frozen and sectioned in
1-cm increments in the transverse plane for identifica-
tion of dyed structures. When the palmar nerves were
not dyed blue, the distance between the nerve and dyed
tissue was measured in millimeters.
In vivo catheter placement for continuous versus
intermittent analgesia—Forelimb selection (left or
right) was made by a coin toss for the first horse and
then alternated for each subsequent horse in the tri-
al. Tendonitis induction was performed daily in pairs;
therefore, choice of CPNB or IPNB was made each day
by coin toss on the first horse, and the alternate treat-
ment was performed on the second horse of that day.
Horses were sedated with detomidine (0.01 mg/kg
[0.005 mg/lb], IV) and butorphanol (0.01 mg/kg, IV). Lo-
cal anesthesia was achieved through median nerve and ul-
nar nerve blockade with 5 mL (100 mg) of 2% lidocaine
and 5 mL (25 mg) of 0.5% bupivacainec at each site and
subcutaneous placement of 1 to 3 mL (20 to 60 mg) of li-
docaine at the stab incision and tunnel site. Catheters were
placed as described for anatomic studies with the addition
of 0.3 mL of heparin (1,000 U/mL) to the catheter lumen
after placement to confirm catheter patency and to prevent
intraluminal clot formation by contaminating blood during
collagenase injection for tendonitis induction.
Collagenase-induced core lesions were created in the
tensile region of the midmetacarpal region of the SDFT of
1 forelimb. The duration from local median nerve and ul-
nar nerve blockade to completion of tendonitis induction
was approximately 1 hour. After lesion induction, a 7-mL
bolus injection of bupivacaine was given via the catheter,
and phenylbutazone (4.4 mg/kg [2 mg/lb], IV) was ad-
ministered (t = 0 hours). Over the following 3 days, 10 mL
of 0.5% bupivacaine was delivered via the catheter, either
continuously via a battery-powered infusion diskd at the
rate of 0.14 mL/h or intermittently via manual injection
with bupivacaine boluses of 0.85 mL every 6 hours.
The forelimbs were bandaged with a compression
wrap over the metacarpus. The catheter and infusion
disk or injection port were incorporated within the
bandage along the distal and lateral aspect of the an-
tebrachium, so that observers could not differentiate
between groups (Figure 1). The horses were confined
individually to box stalls, and phenylbutazone (2.2 mg/
kg [1 mg/lb], IV) was administered that evening at t =
8 hours. The following morning, phenylbutazone (3.3
mg/kg [1.5 mg/lb], IV) was administered followed by
phenylbutazone (2.2 mg/kg, PO) every 12 hours for
3 days and then every 24 hours for 4 days. Twenty-
four hours after pump cessation for CPNB or the final
bupivacaine administration for IPNB, the catheter was
removed. Bacterial culture of the catheter tip was per-
formed if there was pain on palpation or swelling of the
catheter insertion site at the time of removal.
In vivo data collection—Ultrasonography was per-
formed 24 hours after lesion induction to confirm suc-
cessful formation of an SDFT core lesion. Bandage chang-
es and examination of the forelimbs were performed daily.
Sensitivity to digital palpation, swelling, and presence of
increased heat at the catheter insertion site were evalu-
ated at each bandage change, and forelimbs were digitally
photographed. In CPNB-treated horses, the continued
and normal function of the pump was confirmed every
12 hours, and in IPNB-treated horses, continued catheter
patency was confirmed at each bolus administration every
Horses were video recorded at 1 hour prior to and
6, 8, 12, 18, 24, 30, 36, 48, 60, 72, 84, and 96 hours
after tendonitis induction in the same pattern as fol-
lows: toward the stationary video camera, in a left-hand
quarter circle past the videographer, and then in a large,
right-hand semicircle back to their starting point, walk-
ing away from the videographer. At completion of the
study, video clips were digitized and placed in a random
sequence on a master compact disk. Two board-certified
surgery specialists (JC and SLF) and 1 surgery resident
(HLR) graded the randomized clips. Each observer was
blinded to treatment group and time point. Lameness
was graded on an 11-point simple descriptive scale (0
to 10; Appendix).
During video recording, lameness was scored by
the nonblinded investigator (data not used in statistical
analysis). Heart rate, respiratory rate, appetite, attitude,
posture, and the presence of sweat along the neck and
flank were evaluated in the stall prior to each lameness
evaluation and video recording. A notation for resting
behavioral or physiologic signs of pain was made in
the medical record if the horse had a heart rate > 48
beats/min, respiratory rate > 20 breaths/min, decrease
in appetite, dull or agitated mentation, or abnormal
sweating along the neck or flank or if it held the af-
fected forelimb abnormally (eg, pawing or pointing).
Additionally, if any variable was indicative of severe dis-
tress (heart rate > 60 beats/min, respiratory rate > 32
breaths/min, extreme agitation, persistent recumbency,
JAVMA, Vol 238, No. 8, April 15, 2011 Scientific Reports 1035
marked sweating, or repeated lifting of the forelimb) or
the lameness grade was > 7, horses received butorpha-
nol (0.01 to 0.02 mg/kg [0.005 to 0.009 mg/lb], IM, as
needed) and were excluded from statistical analysis be-
yond that time point. If additional analgesia did not im-
prove signs of severe pain, percutaneous median nerve
and ulnar nerve blockade were repeated.
At completion of this study, all horses were evalu-
ated for lameness, swelling, pain on digital palpation,
and other forelimb abnormalities every 12 hours un-
til the time of euthanasia for a different experiment.
Compressive bandages over the metacarpus were main-
tained for a total of 4 weeks. Horses were euthanized by
pentobarbital overdose at 3 (n = 2), 5 (2), 9 (8), and 17
(2) weeks after the completion of this trial. The treated
forelimb was collected, dissected, and grossly evaluated
for any abnormalities at the catheter site.
Statistical analysis—A summary lameness score was
created for each horse, at each time point, by calculation
of the median score assigned by the 3 blinded observers.
Lameness data were reported as a median (95% confidence
intervals) for the entire trial and median (range) for each
individual time point. The summary lameness score was
evaluated for interaction of treatment group and time by
2-way ANOVA. When lack of interaction was determined,
the main effects were compared in the repeated-measures
ANOVA, followed by individual evaluation at each time
point with the Wilcoxon rank sum test. Nominal data (yes
or no) for behavioral or physiologic signs of pain at rest
were tested for differences in proportion with the Fisher
exact test. Statistical analysis was performed by use of
commercially available software programs.e,f Between-
observer agreement was evaluated by determination of a
weighted (quadratic) κ coefficient.14,15 To test the hypoth-
esis that CPNB-treated horses would have lower lameness
scores and a lower proportion of behavioral signs indica-
tive of pain, 1-sided P values < 0.05 were considered sig-
nificant. Weighted κ coefficients > 0.8 were considered as
very good agreement.16
Ex vivo catheter placement and anatomic studies—
Catheter placement was successful in all equine cadaver
forelimbs. Lateral and medial subcutaneous swellings
occurred in all forelimbs following 7-mL bolus injection.
The palmar vasculature and nerves were not grossly
damaged in any forelimb. All forelimbs had dye pres-
ent within the palmar fascia, between the SL and the
ALDDFT, with the greatest amount of dye in this space
located abaxially, just deep to the palmar fascia, medi-
ally and laterally (Figure 2). Dye was present adjacent
to (5/5) and staining (5/5) the lateral palmar nerve and
adjacent to (5/5) and staining (1/5) the medial palmar
nerve. In each instance where the medial palmar nerve
was not stained, dye was within 2 mm of the nerve. In
all forelimbs, dye was present superficial to the palmar
fascia in the subcutaneous plane (laterally greater than
In vivo catheter placement—In 1 horse (horse 1),
initial placement of the catheter was not successful; af-
ter removal of the introducer needle, the catheter could
easily be advanced, as if it were in a cavity, and synovial
fluid was retrieved from the catheter. The catheter was
removed and replaced successfully 1 cm farther distally.
The following day, during ultrasonographic examina-
tion of the tendonitis lesion, it was observed that there
was anechoic fluid distending the distal part of the car-
pal canal. Otherwise, catheter placement was success-
ful in all forelimbs. There was subcutaneous swelling
(laterally greater than medially) in all forelimbs follow-
ing bolus injections of bupivacaine.
Hemorrhage during catheter placement occurred
in forelimbs of 2 horses (horses 7 and 8), and it was the
impression of the investigator that hemorrhage began
when the introducer needle was fully advanced to the
medial palmar fascia, suggesting that penetration of the
medial palmar vein had occurred. Hemorrhage caused
blood to back up into the catheter, but did not result in
swelling and did not dictate change in technique.
Induction of SDFT tendonitis was successful in
all horses. Diffuse swelling and warmth in the fore-
limb, centered along the palmar aspect of the middle
portion of the metacarpus, were present in all horses.
One horse (horse 8) required additional compression
bandaging to control swelling that extended proxi-
mally to the caudal aspect of the carpus from 30 to
48 hours following tendonitis induction. Light digi-
tal palpation along the catheter insertion site was tol-
erated in all horses, except for the horse (horse 14)
that required catheter replacement. There was no ap-
preciable change in swelling in the forelimb between
horses with excessive hemorrhage and those without.
No horses developed swelling of the forelimb below
the distal aspect of the metacarpus, either during
bupivacaine administration or after.
The catheter used to deliver CPNB was dislodged
with failure of the skin insertion site suture at t = 48
Figure 2—Photograph of a right forelimb in cross section from
a cadaver horse. A closed-tip catheter was placed ex vivo ap-
proximately 12 cm distal to the accessory carpal bone, from lat-
eral to medial, between the SL and ALDDFT. Seven milliliters of
new methylene blue was injected, and the limb was frozen and
sectioned in 1-cm increments. There is blue staining of the lat-
eral palmar nerve (white arrow) and blue staining adjacent to the
medial palmar nerve (gray arrow). DDFT = Deep digital flexor ten-
don. L = Lateral. M = Medial. MC2 = Second metacarpal bone.
MC4 = Fourth metacarpal bone.
1036 Scientific Reports JAVMA, Vol 238, No. 8, April 15, 2011
hours in 1 horse (horse 14) during a bandage change.
The catheter was replaced without difficulty, and the
same pump was attached to the new catheter. The fol-
lowing day, there was swelling, heat, and pain on palpa-
tion at the level of catheter insertion (laterally greater
than medially) and an increase in lameness (score,
4/10). Ultrasonographic examination revealed that the
swelling had a diffuse heteroechoic pattern and lack
of effusion within the carpal canal. The catheter was
removed, and bacterial culture of the catheter tip re-
sulted in growth of Staphylococcus aureus. The catheter
was not replaced, and the data from the horse were
excluded from statistical analysis from that time point
onward. The horse was treated with broad spectrum
antimicrobials for 3 days, and the heat, pain, and swell-
ing resolved. No lameness was observed throughout the
remainder of the study, and the horse received lameness
scores of zero at 72, 84, and 96 hours by the nonblinded
Two CPNB-treated horses (horses 10 and 12) had
cessation of pump action observed at 48 hours. The
pump was disconnected, the catheter was forcibly
flushed with 0.5 mL of heparin (1,000 U/mL), and the
pump was reconnected. For each of these horses at the
time of pump disconnection, approximately 1 mL of
bupivacaine flowed from the pump, equivalent to the
amount that should have been dispensed over the pre-
vious 7 hours. The summary lameness score (blinded)
for each of these 2 horses at 48 hours was not differ-
ent from each horse’s lameness score at adjacent time
points, 36 and 60 hours.
Other than in 1 horse (horse 14), catheters were not
dislodged. At catheter removal, catheters maintained
their in situ orientation with an acute bend where they
had a transition from the dorsolateral subcutaneous
tunnel to the space between the ALDDFT and SL. No
gross abnormalities were detected in any forelimbs after
completion of the study during the 3 to 17 weeks prior
to euthanasia. During this period, digital pulses were
regularly detectable medially and laterally, skin turgor
and tone were within normal limits, and there was no
swelling or pain on palpation of the distal aspect of the
affected forelimb. On postmortem examination, no ab-
normalities were detected on gross anatomic dissection
of the catheter site.
In vivo data collection—Starting at approximately
t = 6 hours, 1 horse (horse 5) developed a large colon
impaction that caused an increase in heart rate and re-
spiratory rate, decreased appetite, dull mentation, signs
of abdominal pain, and abdominal distention; this horse
required medical treatment during the first 3 days of the
trial. Because detomidine and butorphanol were repeat-
edly administered to control signs of abdominal pain,
data from the horse were excluded from all statistical
analysis. Otherwise, no horses had signs of distress or se-
vere lameness (score > 7) that required rescue analgesia
at any time point.
While at rest in their stalls, 3 IPNB-treated horses
(horses 2, 6, and 13) received a pain assessment of yes
on 7 total occasions, all within the first 24 hours: horse
6 for pawing, horse 2 for pawing and tachycardia (60
beats/min), and horse 13 for tachycardia (52 beats/
min). All other IPNB-treated horses and all CPNB-
treated horses (other than the horse with severe colic)
received a pain assessment of no at all time points, re-
sulting in a significantly (1-tailed P = 0.013) higher pro-
portion of IPNB-treated horses with behavioral scores
that indicated signs of pain.
There was no significant (2-sided P = 0.831) inter-
action of time and treatment (CPNB vs IPNB) on sum-
mary lameness scores. When the main effects were test-
ed, summary lameness scores of CPNB-treated horses
(median, 1 [95% confidence interval, 1.3 to 1.9]) were
significantly (1-tailed P < 0.001) lower than that of
IPNB-treated horses (median, 2 [95% confidence inter-
val, 1.9 to 2.6]) during the course of the study. When
each time point was individually tested, CPNB-treated
horses had significantly lower lameness scores than
IPNB-treated horses at 8 hours (median, 2 [range, 0 to
4] vs median, 4 [range, 3 to 5]; 1-tailed P = 0.014) and
12 hours (median, 2 [range, 0 to 3] vs median, 3 [range,
2 to 5]; 1-tailed P = 0.037]). At 18 hours, CPNB-treated
horses had lower lameness scores (median, 1.5 [range,
0 to 2]) than those of IPNB-treated horses (median, 3
[range, 1 to 5]), but this difference was not significant
(1-tailed P = 0.055). There was no significant difference
in lameness scores between CPNB- and IPNB-treated
horses at all other time points (0, 6, 24, 30, 36, 48, 60,
72, 84, and 96 hours; Figure 3).
Weighted κ coefficients were > 0.8 for all observers
(SLF and JC [weighted κ = 0.804], SLF and HLR [weight-
ed κ = 0.843], and JC and HLR [weighted κ = 0.850]).
Experimental assessment of signs of pain in horses
is often done with an avoidance reaction such as hoof-
withdrawal response to heat, pressure, and electrical
stimulation.13,17–22 Although hoof-withdrawal response
is useful, especially in testing a nerve block, it may not
represent clinically relevant pain.23 Induction of sy-
novitis through intra-articular injection of autologous
Figure 3—Median (range) lameness scores (summary score from
3 blinded observers on 11-point scale; see Appendix) of 14 horses
receiving CPNB (n = 7) or IPNB (7). Both groups received 10 mL
of 0.5% bupivacaine injected over 3 days, and lameness was
evaluated for 4 days after collagenase tendonitis induction. Indi-
vidual analysis at each time point revealed significantly (*1-tailed
P < 0.05) lower lameness score in CPNB-treated horses at t =
8 hours and t = 12 hours, compared with that of IPNB-treated
horses. Error bars represent minimum and maximum scores.
JAVMA, Vol 238, No. 8, April 15, 2011 Scientific Reports 1037
blood,24 amphotericin,4,25 interluekin-1β,26 and polyvi-
nyl alcohol foam27 and through induction of tendonitis
by injection of enzymes28 has been performed experi-
mentally in horses to assess clinically relevant signs of
pain. The effect of experimentally induced synovitis in
horses is generally graded at a walk and trot and may
not adequately mimic signs of severe pain in horses.
Collagenase-induced tendonitis was selected to test
peripheral neural blockade in our study, as it results in
lameness at a walk and marked behavioral and physi-
ologic responses to pain that are relatively short lasting
(up to 48 hours).
Assessment of signs of pain is difficult in nonver-
bal patients and is especially difficult in horses,29 which
may be in part because of the importance of concealing
potential weakness in a prey animal.1 Therefore, a strin-
gent protocol for identification of pain was used wherein
any abnormal resting variable in heart rate, respiratory
rate, attitude, appetite, presence of sweat, or abnormal
stance (pointing or pawing with the limb) would result
in the same categorization of behavioral and physiologic
signs of pain as yes. This rigorous method worked well,
allowing detection of differences in signs of pain between
CPNB- and IPNB-treated horses.
In contrast to generalized assessment of signs of
pain, lameness is easily graded, and it is widely accept-
ed that horses with lower grades of lameness have less
pain. To improve consistency of lameness characteriza-
tion between observers30 and allow detection of differ-
ent degrees of lameness at a walk, an 11-point simple
descriptive scale was developed (Appendix). The
simple descriptive scale allowed very good agreement
between all observers with a weighted κ coefficient >
0.8.14–16 Although force plate data would have increased
objectivity of lameness data, it was not used because it
requires the horse to be walked over the plate several
times, which is not appropriate for horses with severe
lameness. An in-shoe pressure measurement system has
been described and may have been useful in obtaining
more objective data,31 but was not available to the in-
There were limitations associated with video re-
cording and grading in this study. The number of left
and right forelimbs that received CPNB (4 left; 3 right)
or IPNB (3 left; 4 right) was slightly unbalanced. Al-
though investigators felt the increase in lameness dur-
ing turning was similar for the limb on the inside versus
the outside of the circle, it may have led to a decreased
ability to equally judge left and right forelimb lameness.
Next, throughout the trial, only the treated limb was
bandaged. Although this is unlikely to have changed
assessment of most lameness grades, it may have re-
sulted in increased lameness detection of horses with
absent or mild lameness; however, any effect of ban-
daging would have been equal between the 2 groups, if
not more pronounced in CPNB-treated horses, where
lameness grades were lower.
The CPNB technique described by Driessen et al13 in-
volves placement of 2 catheters, 1 medially and the second
laterally, deep to the palmar fascia, alongside the medial
and lateral palmar nerves in a proximal to distal direc-
tion. Proximal skin entry is approximately at the level of
the carpometacarpal joint, and the catheter is advanced
distally 8 cm for the medial catheter and 11.5 cm for the
lateral catheter. After catheter placement, a 5-mL bolus
dose of bupivacaine is given followed by an infusion of
0.5% bupivacaine at 4 mL/h. Minor modifications to this
technique13 were made for several reasons. First, by use of
1 catheter and 1 pump, the cost of instrumentation was
reduced. Second, to reduce the likelihood of complete
nerve block, a substantially decreased dose of bupivacaine
was used.32 Although anesthetic infusions were performed
for only 3 days and vasculature diameter in the distal as-
pect of the forelimb was not measured, another possible
benefit of a lower bupivacaine dose is reduced likelihood
of swelling and vasodilation in the distal aspect of the
limb.13 Third, by placing catheters transversely between
the ALDDFT and SL, dorsal to the medial and lateral pal-
mar nerves rather than adjacent and parallel to them, the
risk of contacting the nerves with the introducer needle
or catheter was reduced, thereby reducing the risk of in-
jury to the nerves and subsequent paresthesia or neuroma
formation. Finally, modification in technique allowed for
more proximal placement of the catheter exit holes along
The use of IPNB was selected as a control rather
than a true negative control, such as a continuous infu-
sion of saline (0.9% NaCl) solution. Without planned
local analgesia of some form, it was expected that many
or all horses in the control group would meet the re-
quirements outlined for rescue analgesia. In short,
IPNB was expected to be effective, but less so than
CPNB. This was confirmed, as rescue analgesia was not
required in either group, suggesting that IPNB may also
be effective for pain reduction. Superior pain control of
CPNB may be because continuous analgesia provides
better pain relief8,9,33 and prevents wind-up, where there
is centrally mediated sensitization to pain because of
repeated painful stimulation at the peripheral nerves
leading to increased sensation of pain with the same
Two horses had pump failure that was detected at
48 hours and appeared to have begun at approximately
41 hours. Neither horse had a change in lameness score
(lameness score of 1) at 48 hours, compared with the
lameness score at 36 and 60 hours. Although this could
be interpreted as lack of a CPNB effect, it is more likely
a combination of diminishing lameness at 48 hours
with experimentally induced tendonitis and the posi-
tive effect of reduced wind-up34 secondary to continu-
It is possible that the lameness reduction of anes-
thetic infusion was due to local effects of bupivacaine
bathing the tendon rather than a reduction in nerve
transmission. Because the catheters were approximate-
ly 4 cm proximal to the lesion and bupivacaine was in-
fused at only 0.14 mL/h, the authors feel this is unlikely
in the CPNB group. Additionally, bupivacaine was de-
posited between the ALDDFT and the SL and would
have to diffuse past the palmar nerves, prior to reaching
the SDFT. In IPNB-treated horses, bolus injections of
0.85 mL of bupivacaine were administered via the cath-
eter, and if a local effect could have occurred, it would
be more likely in this group.
Swelling in the distal portion of the limb is a re-
ported complication of CPNB13 and was not detected
1038 Scientific Reports JAVMA, Vol 238, No. 8, April 15, 2011
in the horses of the study presented here. However, the
experimentally induced tendonitis may have made it
difficult to detect catheter- or infusion-related swelling,
as forelimb swelling, centered on the palmar aspect of
the middle portion of the metacarpus, is a consequence
of tendonitis. Additionally, because of expected swell-
ing, all horses were bandaged with a supportive wrap
to the metacarpal region, which may have minimized
catheter- or anesthetic-related swelling.
No attempt was made to document complete loss
of pain perception, sensation, or proprioception, and
given the low dose of anesthetic used, it was not ex-
pected to occur. However, 3 CPNB-treated horses and 1
IPNB-treated horse received a summary lameness score
of zero after they had lameness in the first 24 hours
of the study, when lameness was expected to be more
pronounced. Although it is tempting to interpret this
as possible complete lack of pain, 1 blinded observer
had assigned a score of 1 in each horse, suggesting that
some pain was still perceived, albeit mild.
Hemorrhage during catheter placement occurred
in 2 horses. In each instance, hemorrhage began dur-
ing palpation of the introducer needle beneath the skin
medially and was assumed to be a result of laceration
to the medial palmar vein. On anatomic specimens, the
medial vein was not injured, but it was not surprising
that it was punctured in vivo, as veins are the dorsal-
most structure in the distal limb (relative to the artery
and nerve) and the medial vein lies directly in the path
of needle insertion between the SL and ALDDFT. Al-
though not ideal, the investigators felt this was a mild
complication with minimal to no clinical consequence:
no swelling, pump cessation, or difficulty of intermit-
tent injections occurred in horses that had hemorrhage.
Two complications occurred that were potentially
serious. One catheter was advanced into the distal as-
pect of a distended carpal canal during catheter place-
ment. This was immediately recognized; the catheter
was removed and successfully replaced farther distally,
and no additional complications occurred. Another
catheter had positive bacterial culture results for S au-
reus after removal. It should be noted that this catheter
had been accidentally removed at 24 hours and then
replaced at the same level through the same stab inci-
sion. The horse was treated with broad-spectrum anti-
microbials for 3 days, and no additional complications
In conclusion, this study demonstrates the efficacy
of peripheral neural blockade by local catheter deliv-
ery for the short-term reduction of severe signs of pain
over 3 days in 14 horses with experimentally induced
tendonitis of a forelimb. Continuous infusion of 0.5%
bupivacaine at 0.14 mL/h resulted in significantly lower
lameness scores and a lower proportion of horses with
behavioral and physiologic signs of pain than infusion
of a proportional volume of bupivacaine every 6 hours.
This was achieved with a very low dose of local anes-
thetic that did not cause total loss of pain sensation.
The technique was easily applied, and catheters were
not dislodged with normal limb movements. The de-
scribed technique is appropriate for clinical use when
a short-term reduction in distal limb pain is desired
and should be tested for safety and efficacy prior to
longer-term applications. The clinician should consid-
er strongly the risk of microbial contamination to the
tissue surrounding the catheter when selecting horses
for CPNB and may consider prophylactic antimicrobi-
als while catheters are in place to minimize the risk of
catheter-related soft tissue infection.
a. Perifix Continuous Epidural Anesthesia Set, B. Braun Medical
Inc, Bethlehem, Pa.
1% methylene blue, Taylor Pharmaceuticals, Decatur, Ill.
Marcaine, Hospira Inc, Lake Forest, Ill.
Infu-Disk, MILA International Inc, Erlanger, Ky.
Statistix 9, Analytical Software, Tallahassee, Fla.
MedCalc Statistical Software, version 11.1.1, MedCalc Software,
1. Taylor PM, Pascoe PJ, Mama KR. Diagnosing and treating pain
in the horse. Where are we today? Vet Clin North Am Equine
Peloso JG, Cohen ND, Walker MA, et al. Case-control study of
risk factors for the development of laminitis in the contralateral
limb in Equidae with unilateral lameness. J Am Vet Med Assoc
Baxter GM, Morrison S. Complications of unilateral weight
bearing. Vet Clin North Am Equine Pract 2008;24:621–642.
Sysel AM, Pleasant RS, Jacobson JD, et al. Efficacy of an epi-
dural combination of morphine and detomidine in alleviating
experimentally induced hindlimb lameness in horses. Vet Surg
Goodrich LR, Nixon AJ, Fubini SL, et al. Epidural morphine and
detomidine decreases postoperative hindlimb lameness in horses
after bilateral stifle arthroscopy. Vet Surg 2002;31:232–239.
Martin CA, Kerr CL, Pearce SG, et al. Outcome of epidural cath-
eterization for delivery of analgesics in horses: 43 cases (1998–
2001). J Am Vet Med Assoc 2003;222:1394–1398.
Goodrich LR. Strategies for reducing the complication of or-
thopedic pain perioperatively. Vet Clin North Am Equine Pract
Singelyn FJ, Deyaert M, Joris D, et al. Effects of intravenous pa-
tient-controlled analgesia with morphine, continuous epidural
analgesia, and continuous three-in-one block on postoperative
pain and knee rehabilitation after unilateral total knee arthro-
plasty. Anesth Analg 1998;87:88–92.
Capdevila X, Barthelet Y, Biboulet P, et al. Effects of periop-
erative analgesic technique on the surgical outcome and dura-
tion of rehabilitation after major knee surgery. Anesthesiology
10. Wenger S, Moens Y, Jaggin N, et al. Evaluation of the analge-
sic effect of lidocaine and bupivacaine used to provide a bra-
chial plexus block for forelimb surgery in 10 dogs. Vet Rec
11. Wenger S. Brachial plexus block using electrolocation for pan-
carpal arthrodesis in a dog. Vet Anaesth Analg 2004;31:272–275.
12. Wolfe TM, Bateman SW, Cole LK, et al. Evaluation of a local an-
esthetic delivery system for the postoperative analgesic manage-
ment of canine total ear canal ablation—a randomized, controlled,
double-blinded study. Vet Anaesth Analg 2006;33:328–339.
13. Driessen B, Scandella M, Zarucco L. Development of a technique
for continuous perineural blockade of the palmar nerves in the
distal equine thoracic limb. Vet Anaesth Analg 2008;35:432–448.
14. Landis JR, Koch GG. The measurement of observer agreement
for categorical data. Biometrics 1977;33:159–174.
15. Cohen J. Weighted kappa: nominal scale agreement with pro-
vision for scaled disagreement or partial credit. Psychol Bull
16. Dawson B, Trapp RG. Basic and clinical biostatistics. 4th ed. New
York: Lange Medical Books/McGraw-Hill, 2004.
17. Pippi NL, Lumb WV. Objective tests of analgesic drugs in po-
nies. Am J Vet Res 1979;40:1082–1086.
18. Kalpravidh M, Lumb WV, Wright M, et al. Analgesic effects
JAVMA, Vol 238, No. 8, April 15, 2011 Scientific Reports 1039
of butorphanol in horses: dose-response studies. Am J Vet Res
19. Kalpravidh M, Lumb WV, Wright M, et al. Effects of butorpha-
nol, flunixin, levorphanol, morphine, and xylazine in ponies.
Am J Vet Res 1984;45:217–223.
20. Kamerling SG, Weckman TJ, DeQuick DJ, et al. A method for
studying cutaneous pain perception and analgesia in horses.
J Pharmacol Methods 1985;13:267–274.
21. Kamerling SG, Dequick DJ, Weckman TJ, et al. Differential ef-
fects of phenylbutazone and local anesthetics on nociception in
the equine. Eur J Pharmacol 1984;107:35–41.
22. Haussler KK, Hill AE, Frisbie DD, et al. Determination and use
of mechanical nociceptive thresholds of the thoracic limb to as-
sess pain associated with induced osteoarthritis of the middle
carpal joint in horses. Am J Vet Res 2007;68:1167–1176.
23. Matthews NS. Equine analgesics: what do we have and what do
we need?, in Proceedings. 55th Annu Conv Ame Assoc Equine
24. Judy CE, Galuppo LD. Evaluation of iatrogenic hemarthrosis of the
metacarpophalangeal joint as a method of induction of temporary
reversible lameness in horses. Am J Vet Res 2005;66:1084–1089.
25. Bussieres G, Jacques C, Lainay O, et al. Development of a compos-
ite orthopaedic pain scale in horses. Res Vet Sci 2008;85:294–306.
26. DePuy T, Howard R, Keegan K, et al. Effects of intra-articular
botulinum toxin type A in an equine model of acute synovitis: a
pilot study. Am J Phys Med Rehabil 2007;86:777–783.
27. Cornelissen BP, Rijkenhuizen AB, van den Hoogen BM, et al.
Experimental model of synovitis/capsulitis in the equine meta-
carpophalangeal joint. Am J Vet Res 1998;59:978–985.
28. Chambers JP, Livingston A, Waterman AE, et al. Analgesic ef-
fects of detomidine in thoroughbred horses with chronic tendon
injury. Res Vet Sci 1993;54:52–56.
29. Raekallio M, Taylor PM, Bennett RC. Preliminary investiga-
tions of pain and analgesia assessment in horses administered
phenylbutazone or placebo after arthroscopic surgery. Vet Surg
30. Fuller CJ, Bladon BM, Driver AJ, et al. The intra- and inter-assessor
reliability of measurement of functional outcome by lameness scor-
ing in horses. Vet J 2006;171:281–286.
31. Judy CE, Galuppo LD, Snyder JR, et al. Evaluation of an in-shoe
pressure measurement system in horses. Am J Vet Res 2001;62:23–28.
32. Muir WW, Hubbell JAE, Bednarski RM, et al. Local anesthetic
drugs and techniques. In: Handbook of veterinary anesthesia. 4th
ed. St Louis: Mosby Elsevier, 2007;51–55.
33. Liu SS, Salinas FV. Continuous plexus and peripheral nerve blocks
for postoperative analgesia. Anesth Analg 2003;96:263–267.
34. Woolf CJ. Recent advances in the pathophysiology of acute
pain. Br J Anaesth 1989;63:139–146.
Score Lameness characteristics
Uneven gait; rare head nod
Uneven gait; occasional head nod (or consistent nod only on turning)
Consistent head nod, mild (or occasional severe head nod)
Consistent head nod, moderate
Consistent head nod, severe; heel on ground throughout stride
Consistent head nod, severe; heel touches ground briefly every step
Consistent head nod, severe; heel touches ground occasionally
Consistent head nod, severe; heel touches ground rarely
Consistent head nod, severe; heel never touches ground
Swinging limb; toe never touches ground
Simple descriptive scale for grading lameness observed at a walk.