Journal of Clinical Anesthesia (2007) 19, 53 56

Original contribution

Premedication with intravenous low-dose ketamine

suppresses fentanyl-induced cough
Chun-Chang Yeh MD (Assistant Professor)a, Ching-Tang Wu MD (Assistant Professor)a,
Billy K Huh MD, PhD(Clinical Associate Professor)b,
Meei-Shyuan Lee DPH(Associate Professor)c,
Shinn-Long Lin MD (Associate Professor)a,
Michael J Sheen MD (Staff Anesthesiologist)a, Chih-Shung Wong MD, PhD (Professor)a,*
a

Department of Anesthesiology, Tri-Service General Hospital and National Defense Medical Center, Neihu 114, Taipei,
Taiwan, Republic of China
b
Department of Anesthesiology, Duke University Medical Center, Durham, NC 27710, USA
c
School of Public Health, National Defense Medical Center, Taipei, Taiwan
Received 30 November 2005; revised 17 May 2006; accepted 18 May 2006

Keywords:
Ketamine;
NMDA receptor;
Fentanyl;
Cough

Abstract
Study Objective: To evaluate the effect of low-dose ketamine on fentanyl-induced cough.
Design: Prospective, randomized, double-blind, placebo-controlled clinical trial.
Setting: Medical center hospital.
Patients: 360 ASA physical status I-II patients aged 18 to 65 years, weighing between 40 and 80 kg,
and scheduled for elective surgery during general anesthesia.
Interventions and Measurements: Patients were randomly assigned to receive either ketamine 0.15
mg/kg or placebo (equal volume of 0.9% saline) given intravenously over 10 seconds, one minute
before administration of fentanyl (1.5 lg/kg IV, injected within 5 seconds), during induction of general
anesthesia. Any episode of cough was classified as coughing and the onset time of cough (the time of
the first episode of cough) was observed for one minute after fentanyl administration by a blinded
observer. Severity of coughing was graded based on the number of episodes of coughing (mild, 1-2;
moderate, 3-5; and severe, N5). Blood pressure, heart rate, and pulse oximetry oxygen saturation (Spo2)
were recorded before giving ketamine or 0.9% saline and 1 minute after fentanyl injections.
Main Results: After the intravenous injection of fentanyl bolus, patients in the placebo group showed significantly higher frequency cough than those in the ketamine pretreatment group (21.6% vs 7.2%, P b 0.05),
and onset time of the ketamine group was significantly longer than that of the control group (20 F 8 vs 15 F
10 seconds, P b 0.01). However, no difference in cough severity was observed between the two groups.
Conclusion: Low-dose ketamine (0.15 mg/kg IV) effectively reduces fentanyl-induced cough and delays
the onset time of cough.
D 2007 Elsevier Inc. All rights reserved.

* Corresponding author. Tel.: +1 886 2 87927128; fax: +886 2 87927127.

E-mail address: w82556@ndmctsgh.edu.tw (C.-S. Wong).
0952-8180/$ see front matter D 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jclinane.2006.05.021

54

1. Introduction
Fentanyl is often used as a premedicant during induction
of general anesthesia (GA) owing to its rapid onset, short
duration, intense analgesia, reduced cardiovascular depression, and low histamine release [1,2]. However, intravenous
(IV) administration of fentanyl elicits cough, with frequency
ranging from 28% to 65% [3-8]. Coughing is undesirable
during induction of anesthesia and is associated with
adverse effects from elevated intracranial, intraocular, and
intra-abdominal pressures [4-9]. Tweed and Dakin [9]
reported a case of explosive cough after peripheral IV
injection of fentanyl that produced multiple conjunctival and
periorbital petechiae. Prevention of fentanyl-induced cough
is an important clinical implication.
A number of studies have been conducted to reduce the
incidence of fentanyl-induced cough using readily available
anesthetic adjuncts (eg, propofol, lidocaine, ephedrine,
atropine, and midazolam, with varying success). Kamei et al
reported that excitatory amino acids and N-methyl-daspartate (NMDA) receptor antagonists are capable of
modulating the cough reflex [10]. Intravenous ketamine, an
NMDA antagonist, has well-known potent analgesic and
bronchodilatory effects [11-14]. In our experience, the incidence of fentanyl-induced cough seems to be attenuated with
ketamine. Thus, we conducted the present study to examine
the effect of low-dose ketamine on fentanyl-induced cough.

2. Materials and methods

This study was approved by the Tri-service General
Hospitals institutional review board. Written, informed
consent was obtained from all patients before they were
enrolled into this randomized, prospective, double-blind,
placebo-controlled study. Three hundred sixty ASA physical
status I and II patients aged between 18 and 65 years and
scheduled for elective surgery during GA, were randomly
assigned to one of two groups of 180 patients each, using
computer-generated random numbers. Based on a pilot
study, we found that the incidence of fentanyl-induced
cough was 25% (10 of 40 patients had cough after being
given fentanyl); a power analysis was performed using the
reduction of incidence as the primary variable. We
calculated a sample size so that a between-groups mean
difference in incidence of fentanyl-induced cough of 50%
reduction would permit a one-tailed type I error rate of a =
0.05 with a power of 80%. This analysis indicated that a
sample size of at least 152 patients per group was necessary.
Exclusion criteria were body weight exceeding 20% of the
ideal body weight, history of liver or kidney impairment,
depression, smoking, asthma, chronic cough, upper respiratory tract infection within one month of surgery, or use of
medications that could interfere with this study such as
angiotensin-converting enzyme inhibitors, antidepressants,
bronchodilators, or steroids.

C.-C. Yeh et al.

Before GA induction, venous access was established on
the dorsum of the hand, and an IV cannula was connected to
a T-connector for drug injection. All patients were monitored by electrocardiogram (ECG), noninvasive blood
pressure measurement, and pulse oximetry throughout the
study. Patients received either ketamine 0.15 mg/kg IV or
placebo (equal volume of 0.9% normal saline) administered
over 10 seconds, one minute before the fentanyl bolus
injection (1.5 lg/mg within 5 seconds). After the fentanyl
injection, the onset time (from the end of bolus administration to the beginning of coughing), frequency, and severity
of cough were observed for one minute and recorded by
another anesthesiologist who had no knowledge of the
premedication. Severity of cough was graded as mild (1-2),
moderate (3-5), and severe (N5) based on the number of
coughs within one minute of fentanyl injection. Systolic and
diastolic blood pressure (SBP and DBP), heart rate (HR),
and Spo2 were recorded before giving ketamine or normal
saline (T0), and one minute (T1) later after fentanyl
injection. Assisted mask ventilation with oxygen was
applied if desaturation was observed (Spo2 b90% or 5%
decreased from initial value). Some immediate side effects
related to fentanyl and ketamine, such as apnea and truncal
rigidity, were also recorded after injection.
The demographic characteristics of the patients between
groups were evaluated by Students t tests to compare means,
and by v 2 tests to compare the frequency of events. A logrank test was used to evaluate the difference of onset time to
cough between groups. General linear model repeatedmeasures analyses were applied to compare differences of
vital signs between groups before and after fentanyl injection.
A P value b.05 was taken as statistically significant.

3. Results
There were no differences between the two groups in
demographic data (Table 1). The hemodynamic data (BP,
HR, and Spo2) were also similar and there was no
significant difference between groups in baseline or after
fentanyl injection (Table 2). After the IV fentanyl bolus
injection, patients in the placebo group (39/180, 21.6%)
showed a significantly higher frequency of cough than in the
ketamine pretreatment group (13/180, 7.2%; P b 0.05), and
the onset time of cough was significantly longer in the
Table 1

Patient characteristics

Age (y)
Gender (M/F)
Weight (kg)
Height (cm)

Ketamine group
(n = 180)

Placebo group
(n = 180)

41.6 F 17.2
95/85
62.4 F 10.0
164 F 8

42.56 F 17.3
91/89
60.7 F 11.4
164 F 9

Values are means F SD. No statistical difference was observed in

this table.

Ketamine and fentanyl-induced coughing

Table 2
Group

Ketamine
Placebo

55

Changes in vital signs after treatment in both groups

Systolic blood pressure
(mmHg)

Diastolic blood pressure

(mmHg)

Heart rate (bpm)

Spo2 (%)

T0

T1

T0

T1

T0

T1

T0

T1

132.0 F 17.0
132.0 F 17.0

131.1 F 17.8
130.4 F 17.8

77.2 F 9.0
77.1 F 9.6

77.9 F 10.2
75.8 F 9.3

71.5 F 9.1
72.4 F 10.4

72.7 F 10.1
71.4 F 9.4

98.6 F 1.1
98.3 F 1.1

98.6 F 1.2
98.3 F 1.2

T0 indicates time before giving ketamine or normal saline injection; T1, 1 minute later after fentanyl injection.

ketamine group (20 F 8 seconds) than in the control group

(15 F 10 seconds, P b 0.01). However, there was no
difference between groups in cough severity (Table 3). None
of the patients suffered from hypoxemia (Spo2 b90%),
desaturation, apnea, truncal rigidity, or other adverse effects
after fentanyl injection in either group.

4. Discussion
The major finding in the present study is that pretreatment with ketamine 0.15 mg/kg IV reduced the frequency of
fentanyl-induced cough from 21.6% to 7.2% and delayed
the onset time of cough.
In our previous series, the frequency of fentanyl-induced
cough in the placebo group was 18% at the fentanyl dose of
1.5 lg/kg, with an injection time of less than 5 seconds [15];
the frequency of fentanyl-induced cough in our control
group was lower than the previously reported range of 28%
to 65% [3-8]. We propose three explanations. First, the
patients current smoking status may provide a protective
effect for fentanyl-induced cough [15]. Second, the lower
frequency was probably related to a smaller dose of fentanyl
(1.5 lg/kg in 5 seconds). Third, the frequency of cough
seems to be higher with rapid injection of fentanyl; Lin et al
[5] reported up to 65% frequency of cough following
fentanyl (2.5 lg/kg) administration within two seconds. The
threshold of fentanyl-induced cough may be readily reached
with a rapid injection. Our previous study also found that
fentanyl injection time is another factor that affects cough
frequency [15].
A number of theories have been proposed on the
mechanism of fentanyl-induced cough, such as inhibition
of central sympathetic outflow. The inhibition results in a
predominance of vagal tone, which in turn can provoke
cough and reflex bronchoconstriction [3,4,6] before inducing vocal cord spasm or supraglottic obstruction, or by
stimulating the rapidly adapting (irritant) receptors or
C-fiber receptors (J receptors) [3-7,9]. Furthermore, opioid
receptors have been identified in the trachea, bronchi, and
alveolar walls. Hence, fentanyl is thought to trigger airway
smooth muscle constriction and subsequently stimulate the
rapidly adapting receptors, leading to deformation of the
tracheobronchial wall, which results in the cough.
The presence of NMDA receptors also has been demonstrated in the larynx, lung, and airways, and activation of

these receptors can trigger airway constriction [16-18].

Therefore, the direct bronchodilating effect of ketamine may
be attributed to its NMDA receptor antagonism effect
[11,17,19,20]. Moreover, Ambalavanar et al [21] indicated
that ketamine suppresses the laryngeal adductor effect in
cats. However, Sato et al [17] reported that ketamine relaxed
the tracheal smooth muscle contraction caused by histamine
release independent of NMDA receptors. The inhibitory
effect of ketamine on bronchomotor tone is probably due to
interruption of Ca2+ influx. On the other hand, Durieux [22]
reported that ketamine effectively blocked muscarinic
signaling, which may explain the peripheral anticholinergic
(bronchodilation) effect. Brown and Wagner [23] also
demonstrated the presence of a local vagal (bronchodilatory)
effect of ketamine. However, other studies found that
atropine, a vagolytic agent, did not attenuate the fentanyl
cough reflex, suggesting that the cough reflex is independent of the vagal efferent pathway [4,7]. The mechanism
behind the antitussive effect of b 2-adrenoceptors is not clear.
Bronchomotor tone could be augmented by central vagal
discharge. Fentanyl is known to stimulate the vagal nucleus
[7], but ketamine is known to have an inhibitory effect on
the central vagal pathway. Haxhiu et al [24] reported that
applying an NMDA antagonist to the neurons located near
the ventral surface of the medulla increased parasympathetic
preganglionic activity to airway smooth muscles and
prevented NMDA receptormediated tracheal tone. Our
study demonstrates that ketamine pretreatment not only
decreased frequency of cough to 7.2%, but also delayed the
onset time of cough, Therefore, ketamine is believed to
counteract the action of fentanyl by preserving sympathetic-

Table 3 Incidence and onset time of coughing and severity of

cough after fentanyl injection
Group

Incidence of
cough (%)

Onset (s) Severity of cough (%)

Mild

Ketamine 13/180 (7.2)* 20 F 8* 9 (69)

Placebo 39/180 (21.6) 15 F 10 26 (67)

Moderate Severe
4 (31)
9 (23)

0
4 (10)

The dose of fentanyl was 1.5 lg/kg at an injection duration of less than
5 seconds. Severity of cough (number of coughs observed in one
minute): mild (1-2), moderate (3-5), and severe (N5). The onset time
(median F SD) was the median value of patients who coughed.
* Denotes no significant differences compared with the value
before treatment in both groups.

56
parasympathetic balance, resulting in suppression of fentanyl-induced cough and delaying cough onset.
In conclusion, perioperative fentanyl injection causes a
high frequency of cough, which may interrupt smooth
anesthesia induction and endotracheal intubation. Our study
demonstrated that ketamine, at a low dose (0.15 mg/kg),
effectively reduced fentanyl-induced cough when given
one minute before fentanyl injection during GA induction.

Acknowledgments
The authors thank Dr J. A. Lin for support in conducting
this study.

References
[1] Bovill JG, Sebel PS, Stanley TH. Opioid analgesics in anesthesia:
with special reference to their use in cardiovascular anesthesia.
Anesthesiology 1984;61:731 - 55.
[2] Grell FL, Koons RA, Denson JS. Fentanyl in anesthesia: a report of
500 cases. Anesth Analg 1970;49:523 - 32.
[3] Agarwal A, Azim A, Ambesh S, et al. Salbutamol, beclomethasone or
sodium chromoglycate suppress coughing induced by IV fentanyl.
Can J Anaesth 2003;50:297 - 300.
[4] Lui PW, Hsing CH, Chu YC. Terbutaline inhalation suppresses
fentanyl-induced coughing. Can J Anaesth 1996;43:1216 - 9.
[5] Lin CS, Sun WZ, Chan WH, Lin CJ, Yeh HM, Mok MS. Intravenous
lidocaine and ephedrine, but not propofol, suppress fentanyl-induced
cough. Can J Anaesth 2004;51:654 - 9.
[6] Pandey CK, Raza M, Ranjan R, et al. Intravenous lidocaine
suppresses fentanyl-induced coughing: a double-blind, prospective, randomized placebo-controlled study. Anesth Analg 2004;99:
1696 - 8.
[7] Phua WT, Teh BT, Jong W, Lee TL, Tweed WA. Tussive effect of a
fentanyl bolus. Can J Anaesth 1991;38:330 - 4.
[8] Bohrer H, Fleischer F, Werning P. Tussive effect of a fentanyl bolus
administered through a central venous catheter. Anaesthesia 1990;45:
18 - 21.

C.-C. Yeh et al.

[9] Tweed WA, Dakin D. Explosive coughing after bolus fentanyl
injection. Anesth Analg 2001;92:1442 - 3.
[10] Kamei J, Tanihara H, Igarashi H, Kasuya Y. Effects of N-methyl-daspartate antagonists on the cough. Eur J Pharmacol 1989;168:153 - 8.
[11] Corssen G, Gutierrez J, Reves JC, Huber FC. Ketamine in the anesthetic
management of asthmatic patients. Anesth Analg 1972;51:588 - 96.
[12] Lundy PM, Gowdey CW, Colhoun EH. Tracheal smooth muscle
relaxant effect of ketamine. Br J Anaesth 1974;46:333 - 6.
[13] Himmelseher S, Durieux ME. Ketamine for perioperative pain
management. Anesthesiology 2005;102:211 - 20.
[14] Elia N, Tramer MR. Ketamine and postoperative paina quantitative
systematic review of randomized trials. Pain 2005;113:61 - 70.
[15] Lin JA, Yeh CC, Lee MS, Wu CT, Lin SL, Wong CS. Prolonged
injection time and light smoking decrease the incidence of fentanylinduced cough. Anesth Analg 2005;101:670 - 4.
[16] Said SI, Berisha HI, Pakbaz H. N-Methyl-d-aspartate receptors
outside the central nervous system: activation causes acute lung
injury that is mediated by nitric oxide synthesis and prevented by
vasoactive intestinal peptide. Neuroscience 1995;65:943 - 6.
[17] Sato T, Hirota K, Matsuki A, Zsigmond EK, Rabito SF. The role of the
N-methyl-d-aspartate receptor in the relaxant effect of ketamine on
tracheal smooth muscle. Anesth Analg 1998;87:1383 - 8.
[18] Robertson BS, Satterfield BE, Said SI, Dey RD. N-Methyl-d-aspartate
receptors are expressed by intrinsic neurons of rat larynx and
esophagus. Neurosci Lett 1998;244:77 - 80.
[19] Cheng EY, Mazzeo AJ, Bosnjak ZJ, Coon RL, Kampine JP. Direct
relaxant effects of intravenous anesthetics on airway smooth muscle.
Anesth Analg 1996;83:162 - 8.
[20] Sato T, Hirota K, Matsuki A, Zsigmond EK, Rabito SF. The relaxant
effect of ketamine on guinea pig airway smooth muscle is epitheliumindependent. Anesth Analg 1997;84:641 - 7.
[21] Ambalavanar R, Purcell L, Miranda M, Evans F, Ludlow CL.
Selective suppression of late laryngeal adductor responses by Nmethyl-d-aspartate receptor blockade in the cat. J Neurophysiol
2002;87:1252 - 62.
[22] Durieux ME. Inhibition by ketamine of muscarinic acetylcholine
receptor function. Anesth Analg 1995;81:57 - 62.
[23] Brown RH, Wagner EM. Mechanisms of bronchoprotection by
anesthetic induction agents: propofol versus ketamine. Anesthesiology
1999;90:822 - 8.
[24] Haxhiu MA, Deal EC, Norcia MP, van Lunteren E, Cherniack NS.
Effect of N-methyl-d-aspartate applied to the ventral surface of the
medulla on the trachea. J Appl Physiol 1987;63:1268 - 74.