Veterinary Anaesthesia and Analgesia 2019, 46, 28e35 https://doi.org/10.1016/j.vaa.2018.07.

007

RESEARCH PAPER

Rocuronium infusion: A higher rate is needed in

diabetic than nondiabetic dogs

Henning A Haga, Vanessa Bettembourg & Andreas Lervik

Faculty of Veterinary Medicine, Norwegian University of Life Sciences, Oslo, Norway

Correspondence: Henning A Haga, Faculty of Veterinary Medicine, Norwegian University of Life Sciences, Pb 369 Sentrum, 0102 Oslo,
Norway. E-mail: andreas.haga@nmbu.no

Abstract Conclusions and clinical relevance There is a

quite large individual variation in the infusion
Objective To determine the infusion rates that
rates of rocuronium needed to maintain a stable
maintain the train-of-four (TOF) ratio within
neuromuscular block in a varied population of
20e70% in dogs and compare the infusion rates
dogs. Of the variables investigated, diabetes melli-
between diabetic and nondiabetic dogs.
tus was the only one found to significantly influ-
Study design Prospective clinical study. ence the infusion rate of rocuronium.
Animals In total, 47 dogs scheduled for phaco-
emulsification were included with a median (80% Keywords diabetes, dog, infusion, neuromuscular
central range) bodyweight of 10.6 (5.7e35.5) kg blockade, rocuronium.
and age of 7 (1e11) years. Diabetes mellitus was
Introduction
previously diagnosed in nine dogs.
A centralized, immobile eye will facilitate phaco-
Methods After premedication using acepromazine
emulsification. General anaesthesia by itself is usually
and methadone, anaesthesia was induced by
not sufficient to achieve this in dogs, therefore, a
intravenous (IV) propofol and maintained by iso-
neuromuscular blocking agent (NMBA) may be
flurane and fentanyl or remifentanil. The TOF ratio
included. Centralization of the eye may be achieved
was monitored by stimulating the peroneal nerve
with a low dose of NMBA since the extraocular
and the response quantified by accelerometry.
muscles controlling eye movement are relaxed at a
Rocuronium 0.5 mg kg
1 was administered IV, and
low dose compared with other skeletal muscles (Lee
further infused to maintain the TOF ratio between
et al. 1998; Auer et al. 2007; Briganti et al. 2015),
20% and 70%. The infusion rates of rocuronium
however vigilant monitoring of respiratory function
were compared by the ManneWhitney test be-
is required. NMBAs may also be administered as an
tween diabetic and nondiabetic dogs, and the in-
infusion, which can be beneficial during prolonged
fluence of age, sex, bodyweight, body temperature,
surgeries since a stable effect of the NMBA may be
end-tidal carbon dioxide, end-tidal isoflurane con-
achieved. For monitoring purposes stimulation of a
centration, mean arterial blood pressure, pulse
motor nerve and registration of the corresponding
rate and time from induction and time from
muscle contraction may be used, commonly in a
rocuronium bolus to stable rocuronium infusion
train-of-four (TOF) pattern. The ratio between the
rate were investigated in a stepwise, forward
fourth and the first response is termed the TOF ratio.
regression model.
In dogs, the eye will centralize at a lower rocuronium
Results A stable infusion rate was found in 42 dose than what is needed to affect the TOF ratio, thus
dogs. A higher median (80% central range) infu- a decreased TOF ratio ensures a centralized eye
sion rate was found in diabetic [0.43 (0.35e0.50) (Briganti et al. 2015). Observing an effect upon the
mg kg
1 hour
1] compared with nondiabetic dogs TOF ratio may therefore be used as an end point for
[0.30 (0.20e0.50) mg kg
1 hour
1] (p ¼ 0.013). the infusion rate, ensuring a centralized eye. There
None of the other variables investigated were are studies using rocuronium as an infusion in dogs,
found to significantly influence the infusion rate. however a predetermined infusion rate was used

28
Rocuronium infusion in dogs HA Haga et al.
(Alderson et al. 2007). The infusion rates needed to
maintain a partial neuromuscular blockade in a
diverse dog population have not previously been
investigated.
Both rocuronium and vecuronium are NMBAs
with a steroid structure. The duration of action of
vecuronium is shorter in dogs with diabetes mellitus
when compared to nondiabetic dogs (Clark et al.
2012). The effect of atracurium, an NMBA with a
benzylisoquinolinium structure, however does not
seem to be affected by diabetes mellitus in dogs (Leece
& Clark 2017). Both clinical experience and an
earlier publication (Auer 2007) indicate that the ef-
fects of rocuronium may differ between diabetic and
nondiabetic dogs.
The primary aim of this study was, in a varied
population of clinical cases, to investigate the infusion
rate of rocuronium necessary to maintain neuro-
muscular blockade at a stable level as monitored by
the TOF ratio. A secondary aim was to investigate
whether a diagnosis of diabetes mellitus influenced
this infusion rate and a tertiary aim was to perform
an explorative analysis of the data to find other var-
iables that could influence the infusion rates.
Materials and methods
This research did not receive any specific grant from
funding agencies in the public, commercial or not-
for-profit sectors. The study was designed as a pro-
spective clinical study and was approved by the
ethical committee for clinical studies at the Norwe-
gian University of Life Sciences (reference number
14/04723). Informed, written owner consent was
obtained prior to anaesthesia. Recruitment period
was from April 2015 to March 2017. Dogs in which
the predetermined anaesthetic protocol was consid-
ered inappropriate were not included in the study.
Animals
Using commercially available software (JMP 11.0.0;
SAS Institute Inc., NC, USA), the number of dogs
needed to find a difference between diabetic and
nondiabetic dogs was calculated. A normal distribu-
tion was assumed, using an alpha of 0.05 and a po-
wer of 0.8, and aiming at finding a difference in
infusion rates between diabetic and nondiabetic dogs
of 1.5 standard deviations (assumed to be a clinically
relevant difference), nine dogs were needed in each
group. We also aimed at describing the infusion rates
needed in varied dog populations, so after including a
total of 47 dogs of which nine had been previously
© 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 46, 28e35
diagnosed with diabetes mellitus the study was
ended.
Upon admission to the hospital, the owners of the
dogs were interviewed about the health status of their
dogs including questions of comorbidity. A physical
examination was performed, but no standard panel of
blood analysis was performed in the dogs. The dogs
were hospitalized overnight on the day prior to sur-
gery and food was withheld for at least 12 hours prior
to anaesthesia. Diabetic dogs were fed a commercial
dog diet and administered their usual insulin dose in
the evening. At the morning of surgery, blood glucose
concentration was measured with a glucometer
(Accu-Chek Aviva; Roche Diagnostics Norge AS,
Norway). Insulin dosing was then determined based
on blood glucose level. Further management was
done at the discretion of the anaesthesiologist in
charge.
Anaesthesia
Acepromazine (Plegicil; Pharmaxim AB, Sweden)
0.01e0.02 mg kg
1 and methadone (Metadon NAF
10 mg mL
1; Sykehusapoteket Rikshospitalet, Nor-
way) 0.1e0.3 mg kg
1 were administered intra-
muscularly prior to placement of a 20 or 22 gauge
venous catheter (BD Venflon Pro; Becton Dickinson
Infusion Therapy AB, Sweden). Anaesthesia was
induced with propofol (Propofol-Lipuro; B. Braun
Melsungen AG, Germany) to effect by a slow, intra-
venous (IV) injection to allow endotracheal intuba-
tion with an armoured endotracheal tube. The dogs
were connected to an anaesthetic machine (Datex-
Ohmeda S/5 ADU; GE Health Care, Finland) using a
rebreathing circuit in dogs with bodyweight above 10
kg or a paediatric Mapleson D breathing circuit in
dogs less than 10 kg. Anaesthesia was maintained
with isoflurane in oxygen and air. The end-tidal iso-
flurane concentration (FE0 Iso) was adjusted at the
discretion of the anaesthetist to provide a surgical
plane of anaesthesia. Prior to surgery, hydrocortisone
(Solu-Cortef; Pfizer AB, Sweden) 5 mg kg
1 was
administered IV as a part of the ophthalmologic
treatment.
A balanced crystalloid solution (Ringer acetate;
Fresenius Kabi AB, Sweden) was administered IV
throughout anaesthesia at a rate of 5e10 mL kg
1
hour
1. During the period of neuromuscular blockade,
mechanical ventilation was performed with the
following initial settings: a respiratory rate of 12 breaths
minute
1 and a tidal volume of 10 mL kg
1. Ventila-
tion was then accommodated to the needs of the dog.
29

Intraoperative analgesia was provided by fentanyl
(Fentadon 50 mg mL
1; Eurovet Animal Health B.V.,
The Netherlands) or remifentanil hydrochloride
(Ultiva; GlaxoSmithKline Manufacturing S.p.A, Italy)
administered using a syringe driver (Alaris GH
Guardrails plus; BD, UK, or Graseby 3200; Medinor
ASA, Norway) at an initial infusion rate of 10 mg
kg
1 hour
1 IV but was later adjusted as needed.
Postoperative analgesia was provided by buprenor-
phine (Vetergesic vet.; Orion Pharma A/S Animal
Health, Norway) 20 mcg kg
1 administered IV to-
wards the end of surgery, prior to discontinuing the
infusion of fentanyl or remifentanil.
Monitoring and equipment
The dogs were placed in dorsal recumbency for sur-
gery. Capnography, electrocardiography, tempera-
ture, pulse oximetry and oscillometric noninvasive
blood pressure were monitored using a multiparam-
eter monitor (Datex-Ohmeda S/5) and noted on the
anaesthetic record every 5e10 minutes by the
anaesthetist. In dogs weighing less than 10 kg,
oscillometric noninvasive blood pressure was recor-
ded with a separate monitor (Cardell Veterinary
Monitor 9401 BP; Midmark, MI, USA).
For diabetic dogs, blood glucose was measured in
the morning prior to anaesthesia and at least every
60 minutes throughout anaesthesia. A glucose infu-
sion was initiated if deemed necessary at the discre-
tion of the anaesthesiologist in charge.
The level of neuromuscular blockade was moni-
tored using a peripheral nerve stimulator (TOF-
Watch S; Organon, Ireland). Electrical stimuli of 50
mA with 0.2 ms duration were delivered in a TOF
pattern at 2 Hz. The right hind limb was positioned to
allow free movement distal to the stifle. The stimu-
lating electrodes were attached to hypodermic nee-
dles placed subcutaneously over the peroneal nerve
at the level of the proximal fibula. The positive elec-
trode was placed 1e2 cm proximal to the negative
electrode. An acceleration transducer was secured to
the dorsal aspect of the lateral digit using adhesive
tape. The accelerometric measurement was cali-
brated prior to administering rocuronium using the
automatic calibration function of the monitor,
accepting a baseline TOF ratio of 90e100%, and the
TOF ratio was then quantified every 2e5 minutes.
Neuromuscular blockade and reversal
Rocuronium bromide (Esmeron 10 mg mL
1; MSD,
N.V., The Netherlands), further referred to as
30 © 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Rocuronium infusion in dogs HA Haga et al.
rocuronium, was diluted in saline solution
(Natriumklorid 9 mg mL
1; B. Braun Melsungen AG,
Germany) to achieve concentrations of 0.5, 1 or 2 mg
mL
1 for dogs weighing 10, 10e20 or above 20 kg,
respectively. Initially, a bolus of rocuronium 0.5 mg
kg
1 was administered IV and then once movement
response to TOF stimulation became apparent and
the TOF ratio increased above zero, the infusion was
started at 0.1 mg kg
1 hour
1 using a syringe driver
(Alaris GH Guardrails plus or Graseby 3200). The
infusion rate was further titrated in steps of 0.05e0.1
mg kg
1 hour
1, and at least 2 minutes passed be-
tween each step. When the TOF ratio was stable
within 20e70% for at least 15 minutes without
changing the rate, this was considered as the infusion
rate able to maintain the TOF ratio within 20e70%.
Details of the rocuronium administration and TOF
ratios were recorded in a separate registration form
throughout the procedure.
At end of surgery, the rocuronium infusion was
discontinued. When the TOF ratio had returned to
>90%, 10 mcg kg
1 glycopyrrolate (Robinul 0.2 mg
mL
1; Meda AB, Sweden) immediately followed by
50 mcg kg
1 neostigmine (Neostigmin NAF 0.5 mg
mL
1 NAF; Sykehusapoteket, Norway) was admin-
istered IV.
Statistics
Age, sex and bodyweight were gathered from the
anaesthetic record, as well as body temperature, end-
tidal carbon dioxide pressure (PE0 CO2), FE0 Iso, mean
arterial blood pressure (MAP) and pulse rate recorded
at the time point when a stable rocuronium infusion
rate was first found. A database was created and
analysed in commercially available statistical soft-
ware (JMP Pro 13.0.0; SAS Institute Inc.). The data
were plotted to evaluate distribution. Data where a
normal distribution was assumed are presented as
mean ± standard deviation, whereas for nonpara-
metric data median (80% central range) is given. The
infusion rates of rocuronium sufficient to maintain
the TOF ratio within the 20e70% range were
compared between diabetic and nondiabetic dogs
using the ManneWhitney test excluding one dog,
which was diagnosed with both diabetes mellitus and
hyperadrenocorticism. As a post hoc analysis, model
building was performed to investigate other variables
possibly influencing the rocuronium infusion rate.
First, the effect of age, sex, body weight, body tem-
perature, PE0 CO2, FE0 Iso, MAP, pulse rate, time from
induction of anaesthesia rate and time from
Elsevier Ltd. All rights reserved., 46, 28e35
Rocuronium infusion in dogs HA Haga et al.
rocuronium bolus to stable rocuronium infusion rate
was tested by analysis of variance (ANOVA). Variables
with a p < 0.20 were then included in the model
using a stepwise forward regression model. Model fit
was evaluated based on Akaike information criterion,
Bayesian information criterion and R2 adjusted. Only
variables with p < 0.10 were retained in the final
model, which was tested to ensure it fulfilled the as-
sumptions of ANOVA by visual inspection of residual
plot for normality and the Levene test for homoge-
neity of variance.
Results
All dogs recovered uneventfully from anaesthesia and
were discharged from the hospital the day after sur-
gery. The 0.5 mg kg
1 bolus dose of rocuronium
abolished all four TOF response in all dogs. In 42 of
the 47 dogs, a rocuronium infusion rate able to sta-
bilize the TOF ratio within the desired range was
found and this infusion rate centralized the eye in all
dogs. A total of 34 different breeds were represented
within these 42 dogs. For the five dogs where a stable
infusion rate was not established, the data were
excluded from further analysis. None of these five
dogs had a history of diabetes mellitus.
One dog had a history of diabetes mellitus and
hyperadrenocorticism. To stabilize the TOF ratio in
this dog, a rocuronium infusion rate of 0.7 mg kg
1
hour
1 was needed. This dog was excluded from the
statistical analysis. Demographic data split into dia-
betic and nondiabetic dogs are given in Table 1,
whereas clinical data are given in Table 2. Plasma
glucose level the morning prior to surgery was 19.5 ±
9.7 mmol L
1 in diabetic dogs, and in six of these
eight dogs, insulin was administered preoperatively.
A glucose infusion was administered to one dog
intraoperatively.
When compared using the ManneWhitney test,
the diabetic dogs needed a higher rocuronium infu-
sion rate to maintain the TOF ratio within 20e70%
(p ¼ 0.013; Table 2).
Table 1 Characterization of the diabetic and nondiabetic dogs that were included in the final analysis of rocuronium
infusion rate necessary to maintain train of four within the 20e70% range
Parameter
Sex (n)
Age (years)
Bodyweight (kg)
Number of breeds represented (n)
Age and bodyweight are given as median (80% central range).
© 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 46, 28e35
When screening for effects using ANOVA, age (p <
0.0004) and bodyweight (p ¼ 0.13) were the only
variables with a p < 0.20 and included initially in the
analysis. An increasing infusion rate with increasing
age was found. However, the age distribution was
quite different in diabetic versus nondiabetic dogs and
the effect of age was opposite in the two groups. An
age-matched database was created and then age
failed ANOVA significance testing, and thus was not
included in the final model. In the final model, only
diabetes and weight were included, and the infusion
rate of rocuronium was found to be higher in diabetic
than in nondiabetic dogs (F1,39 ¼ 5.1; p ¼ 0.03),
whereas weight did not influence rocuronium infu-
sion rate (F1,39 ¼ 2.1; p ¼ 0.15). The relationship
between rocuronium infusion rate and bodyweight in
the diabetic and nondiabetic dogs is shown in Fig. 1.
Discussion
In diabetic dogs, a significantly higher median
rocuronium infusion rate (0.43 mg kg
1 hour
1) was
needed compared with nondiabetic dogs (0.30 mg
kg
1 hour
1). The quite wide central ranges
(Table 2) also illustrate that the infusion rates need to
be individualized if a specific level of neuromuscular
blockade is desired. In a previous study in dogs, a
rocuronium bolus of 0.5 mg kg
1, similar to the
current study, was administered and an infusion at
0.2 mg kg
1 hour
1 immediately started (Alderson
et al. 2007). The TOF registered from the facial
nerve was lost in 21 of 22 dogs but was regained in
three dogs within less than 60 minutes. The current
findings support that a rocuronium infusion rate of
0.2 mg kg
1 hour
1 in most dogs is not sufficient to
maintain a deep level of neuromuscular blockade.
The significantly higher median infusion rate
found in the diabetic dogs supports the finding by
Auer (2007) who found that of a total of 64 dogs, the
three dogs in which rocuronium had the shortest
duration of action also had diabetes mellitus. For
vecuronium, another steroid NMBA, the duration of
Diabetic Nondiabetic
6 females, 2 males 15 females, 18 males
9 (5e12) 6 (1e11)
12 (5.5e21.1) 10.6 (5.3e36.6)
8 25
31
 Rocuronium infusion in dogs HA Haga et al.

Table 2 Data describing the effect of a 0.5 mg kg
1 bolus of rocuronium followed by the infusion rate necessary to
maintain the train-of-four ratio within the 20e70% range in 42 dogs undergoing phacoemulsification under general
anaesthesia

Variables All dogs (n ¼ 42) Diabetic dogs (n ¼ 8) Nondiabetic dogs (n ¼ 33)

Rocuronium (mg
1 kg
1 hour
1) 0.35 (0.20e0.50) 0.43 (0.35e0.50) 0.30 (0.20e0.50)
Time from bolus to start of CRI (minutes) 19 (11e28) 18 (6e36) 19 (15e30)
TOF ratio at start of CRI 14 (1e43) 15 (9e51) 14 (2e20)
TOF ratio at stable CRI 43 (24e64) 51 (23e65) 42 (24e64)
Time from induction to stable CRI (minutes) 105 (71e155) 100 (69e157) 105 (72e157)
Time from bolus to stable CRI (minutes) 50 (28e106) 48 (14e107) 50 (28e109)
Body temperature ( C) 36.3 ± 0.9 36.8 ± 1.0 36.2 ± 0.8
PE0 CO2 (kPa) 6.5 (5.1e7.0) 6.7 (6.4e7.8) 6.3 (5.1e6.9)
PE0 CO2 (mmHg) 49 (38e53) 50 (48e59) 47 (38e52)
FE0 Iso (%) 1.2 ± 0.1 1.2 ± 0.1 1.2 ± 0.1
MAP (mmHg) 71 ± 11 69 ± 8 72 ± 12
Pulse rate (beats minute
1) 75 ± 10 73 ± 12 76 ± 10

The data are presented for all dogs that were included, and further split into diabetic and nondiabetic dogs. Nonparametric data are presented as median
(80% central range), and normally distributed data as mean ± standard deviation. CRI, continuous rate infusion; FE0 Iso, end-tidal isoflurane; MAP,
mean arterial pressure; PE0 CO2, end-tidal carbon dioxide; TOF, train of four.

Figure 1 Rocuronium infusion rate necessary to maintain the train-of-four ratio within the 20e70% range along the y axis
and bodyweight along the x axis. Nondiabetic dogs are symbolized by open circles and diabetic dogs are symbolized by filled
triangles. The solid line illustrates the linear regression curve for the nondiabetic dogs, whereas the stippled line illustrates the
linear regression curve for the diabetic dogs.

action was shorter in diabetic than in nondiabetic
dogs (Clark et al. 2012). In diabetic compared with
nondiabetic dogs, vecuronium has a shorter duration
of action and for rocuronium a higher maintenance
infusion rate is needed, whereas in humans the
opposite seems to be the case.
A delayed recovery after vecuronium was found in
diabetic versus nondiabetic human patients (Saitoh
et al. 2003, 2005) and the same was observed
when rocuronium was used (Armendariz-Buil et al.
2014). There are, however, also reports finding no
effect of diabetes upon onset or recovery from
rocuronium in humans (Alper et al. 2010). These
studies in humans have been based on patients with
type 2 diabetes, and diabetes in humans is known to
affect motor nerves and skeletal muscle as well as
increasing the electrical current needed for supra-
maximal electrical stimulation. The reason for a
different effect of rocuronium and vecuronium in
diabetic dogs is unknown, but possible causes has

32 © 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 46, 28e35
Rocuronium infusion in dogs HA Haga et al.
been discussed in relation to vecuronium in previous
publications (Clark et al. 2012; Leece & Clark 2017).
The diagnosis of diabetes mellitus in the current study
was given in different veterinary clinics than where
the study was performed. Ideally, the diagnosis
should have been confirmed prior to inclusion, but
this was not done because of economical and time
constraints.
In the current study, the 80% central range of
28e106 minutes for time to stabilize the rocuronium
infusion rate was quite wide, and a longer lasting
infusion could possibly lead to a lower infusion rate
being needed because of active metabolites or satu-
ration of peripheral tissues. However, we found no
association between time to stable infusion rate and
the actual infusion rate needed. When studied in cats,
rocuronium was mainly found to be excreted un-
metabolized, 54% through hepatic and 9% through
renal excretion (Khuenl-Brady et al. 1990). A
possible active metabolite of rocuronium is 17-
desacetyl rocuronium, however this metabolite has
only 5% of the activity of the parent compound (Muir
et al. 1989) and was not found in an experimental
study in cats (Khuenl-Brady et al. 1990).
In the current study, when evaluated statistically,
no association between time to stabilization of infu-
sion rate and the necessary infusion rate was
discovered. Alderson et al. (2007) state that in their
study increased duration of infusion increased the
time to return of spontaneous ventilation and return
of T4. They further hypothesized that this may be
caused by accumulation of rocuronium in the muscle
tissues of the dogs referring to a study by Ezzine and
Varin (2005). However, in the paper by Alderson
et al. (2007), the time to return of spontaneous
ventilation and time to return of T4 were taken from
the start of infusion, so naturally a longer infusion
time will result in a longer time to return of sponta-
neous ventilation and return of T4. Further, Ezzine
and Varin (2005) studied a rocuronium bolus of
0.3 mg kg
1 and continuous rate infusion of 3.6 mg
kg
1 hour
1. This infusion rate is 10-fold higher than
the current median infusion rate, possibly explaining
why they found accumulation. In a study where
neuromuscular blockade was maintained by incre-
mental boluses of rocuronium, no evidence of accu-
mulation was observed (Dugdale et al. 2002), further
supporting the view that accumulation of rocuro-
nium does not seem to be a problem when used in
clinical doses.
In humans, both age and hypothermia will influ-
ence the effect of NMBA (Naguib et al. 2015). Sex,
© 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 46, 28e35
bodyweight and cardiovascular status are variables
that possibly could have an influence. However, in
the current study no influence of these variables was
found. Volatile agents are known to increase the
potency of neuromuscular blocking drugs by a
pharmacodynamic mechanism (Naguib et al. 2015).
However, no effect of FE0 Iso was found in the current
study, a likely reason being that FE0 Iso was main-
tained within a quite narrow range, making it un-
likely to discover an effect of the volatile agent.
Corticosteroids are known to antagonize the effects
of NMBA (Naguib et al. 2015). In the current study,
hydrocortisone was administered to all dogs preop-
eratively. This may have influenced the infusion rates
used. Interestingly in the one dog with hyper-
adrenocorticism an infusion rate of 0.7 mg kg
1
hour
1 was needed, which was the highest infusion
rate used in the study.
There were several limitations to this study. Some
of the data were collected from the anaesthetic re-
cords, whereas data regarding the rocuronium
administration and TOF stimulation were collected in
a dedicated form. Time to disappearance and reap-
pearance of response to TOF stimulation was not
registered with high precision. However, since the
aim of the study was to find an infusion rate main-
taining the TOF ratio within a specified range, regis-
tration of an accurate onset and offset time or
analysing these was not prioritized. One of the
anaesthesiologists always set up the infusion, TOF
stimulation and accelerometry. Response to TOF
stimulation may be quantified using different tech-
nologies of which mechanomyography may be
considered the gold standard (Hemmerling & Le
2007). However, the set up is somewhat complex
and in the clinical setting accelerometry is considered
an acceptable method. Different neuromuscular units
may be used for monitoring the TOF ratio, differing in
sensitivity to NMBA (Sakai et al. 2017). However, in
the current study, access limited which nerves could
be used when monitoring potentially small dogs in
dorsal recumbency during intraocular surgery. In the
current study, the peroneal nerve was the most
accessible nerve and thus used. For all dogs the TOF
monitor was set to deliver 50 mA. Ideally, prior to
monitoring the TOF ratio, the individual milliampere
setting eliciting a supramaximal stimulation should
have been found. However, this is less crucial as
illustrated by previous studies in humans since a ratio
was used (Brull et al. 1990; Helbo-Hansen et al.
1992). The diagnosis in the dogs with diabetes mel-
litus was not further investigated, and the inclusion
33

criteria were wide, therefore the material is quite
varied. However, the data reflect the clinical popu-
lation that undergo phacoemulsification in our hos-
pital. An advantage of this is that the results then
may be generalized to this population rather than to a
narrower defined population of otherwise healthy
individuals.
In these dogs, unfortunately no preanaesthetic
standard analysis of blood was performed, and data
for the concentration of blood glucose are available in
only eight dogs. Because of this lack of data, the
overall effect of blood glucose on the infusion rate of
rocuronium could not be investigated. Respiratory
and metabolic acid base status may influence the
effect of NMBA (Funk et al. 1980). No influence of
PE0 CO2 was found upon the rocuronium infusion rate,
however this does not preclude an influence as the
acid bases status of these dogs was unknown.
The two groups, diabetic and nondiabetic dogs,
were not uniform (Table 1). The nondiabetic group
consisted of more dogs, more males than females,
several breeds, lower median age and larger range of
bodyweights. However, when building the statistical
model, diabetes still came out as the single variable
influencing rocuronium infusion rate, strongly sug-
gesting that diabetes has an effect in its own right.
This is also illustrated in Fig. 1 where bodyweight, the
other variable coming closest to be included in the
final model, is visualized together with the effect of
diabetes.
In conclusion, there is a quite large individual
variation in the infusion rate of rocuronium needed
to maintain neuromuscular block at the level of a
TOF ratio between 20% and 70% in dogs. Diabetic
dogs required a higher infusion rate of rocuronium
than nondiabetic dogs. None of the other clinical
variables investigated were part of the final regression
model, therefore, the effect of diabetes mellitus seems
to be dominant in this investigation.
Acknowledgements
The authors would like to acknowledge the veter-
inary technicians who helped to collect the data and
Janicke Nordgreen for statistical advice. The study
was funded by the Norwegian University of Life Sci-
ences; no external funding was provided.
Authors’ contribution
HAH: study design, collecting data, data manage-
ment, data analysis, preparation of manuscript. VB:
collecting data, data management, preparation of
34 © 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Rocuronium infusion in dogs HA Haga et al.
manuscript. AL: study design, collecting data, prep-
aration of manuscript.
Conflict of interest statement
Authors declare no conflict of interest.
References
Alderson B, Senior JM, Jones RS et al. (2007) Use of
rocuronium administered by continuous infusion in
dogs. Vet Anaesth Analg 34, 251e256.
Alper I, Ulukaya S, Makay O et al. (2010) The pharma-
codynamic effects of rocuronium during general
anesthesia in patients with type 2 diabetes mellitus.
Minerva Anestesiol 76, 115e119.
Armendariz-Buil I, Lobato-Solores F, Aguilera-Celorrio L
et al. (2014) Residual neuromuscular block in type II
diabetes mellitus after rocuronium: a prospective
observational study. Eur J Anaesthesiol 31, 411e416.
Auer U (2007) Clinical observations on the use of the
muscle relaxant rocuronium bromide in the dog. Vet J
173, 422e427.
Auer U, Mosing M, Moens YP (2007) The effect of low
dose rocuronium on globe position, muscle relaxation
and ventilation in dogs: a clinical study. Vet Oph-
thalmol 10, 295e298.
Briganti A, Barsotti G, Portela DA et al. (2015) Effects of
rocuronium bromide on globe position and respiratory
function in isoflurane-anesthetized dogs: a comparison
between three different dosages. Vet Ophthalmol 18,
89e94.
Brull SJ, Ehrenwerth J, Silverman DG (1990) Stimulation
with submaximal current for train-of-four monitoring.
Anesthesiology 72, 629e632.
Clark L, Leece EA, Brearley JC (2012) Diabetes mellitus
affects the duration of action of vecuronium in dogs.
Vet Anaesth Analg 39, 472e479.
Dugdale AH, Adams WA, Jones RS (2002) The clinical
use of the neuromuscular blocking agent rocuronium
in dogs. Vet Anaesth Analg 29, 49e53.
Ezzine S, Varin F (2005) Interstitial muscle concentra-
tions of rocuronium under steady-state conditions in
anaesthetized dogs: actual versus predicted values. Br J
Anaesth 94, 49e56.
Funk DI, Crul JF, Pol FM (1980) Effects of changes in
acid-base balance on neuromuscular blockade pro-
duced by ORG-NC 45. Acta Anaesthesiol Scand 24,
119e124.
Helbo-Hansen HS, Bang U, Nielsen HK et al. (1992) The
accuracy of train-of-four monitoring at varying stim-
ulating currents. Anesthesiology 76, 199e203.
Hemmerling TM, Le N (2007) Brief review: Neuromus-
cular monitoring: an update for the clinician. Can J
Anaesth 54, 58e72.
Khuenl-Brady K, Castagnoli KP, Canfell PC et al. (1990) The
neuromuscular blocking effects and pharmacokinetics of
Elsevier Ltd. All rights reserved., 46, 28e35
Rocuronium infusion in dogs HA Haga et al.
ORG 9426 and ORG 9616 in the cat. Anesthesiology 72,
669e674.
Lee DD, Meyer RE, Sullivan TC et al. (1998) Respiratory
depressant and skeletal muscle relaxant effects of low-
dose pancuronium bromide in spontaneously breath-
ing, isoflurane-anesthetized dogs. Vet Surg 27,
473e479.
Leece EA, Clark L (2017) Diabetes mellitus does not affect
the neuromuscular blocking action of atracurium in
dogs. Vet Anaesth Analg 44, 697e702.
Muir AW, Houston J, Green KL et al. (1989) Effects of a
new neuromuscular blocking agent (Org 9426) in
anaesthetized cats and pigs and in isolated nerve-
muscle preparations. Br J Anaesth 63, 400e410.
Naguib ML, Lien CA, Meistelman C (2015) Pharma-
cology of neuromuscular blocking drugs. In: Miller’s
Anesthesia. Miller RD (ed.). Elsevier, Philadelphia.
pp. 958e994.
© 2018 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 46, 28e35
Saitoh Y, Kaneda K, Hattori H et al. (2003) Monitoring of
neuromuscular block after administration of vecuro-
nium in patients with diabetes mellitus. Br J Anaesth
90, 480e486.
Saitoh Y, Hattori H, Sanbe N et al. (2005) Delayed re-
covery of vecuronium neuromuscular block in diabetic
patients during sevoflurane anesthesia. Can J Anaesth
52, 467e473.
Sakai DM, Martin-Flores M, Romano M et al. (2017)
Recovery from rocuronium-induced neuromuscular
block was longer in the larynx than in the pelvic limb
of anesthetized dogs. Vet Anaesth Analg 44,
246e253.
Received 14 November 2017; accepted 3 July 2018.
Available online 19 September 2018
35