J Neurol (2013) 260:470474

DOI 10.1007/s00415-012-6659-3

ORIGINAL COMMUNICATION

A randomized controlled trial of intranasal-midazolam

versus intravenous-diazepam for acute childhood seizures
Arpita Thakker Preeti Shanbag

Received: 9 June 2012 / Revised: 17 August 2012 / Accepted: 18 August 2012 / Published online: 16 September 2012
Springer-Verlag 2012

Abstract The objective of this study is to compare the

safety and efficacy of midazolam given intranasally with
diazepam given intravenously in the treatment of acute
childhood seizures. A randomized controlled study was
conducted in a pediatric emergency department in a tertiary
general hospital. Fifty children aged from 1 month to
12 years presenting with acute seizures of at least 10 min
duration were enrolled during a 12 month period. Intranasal midazolam (0.2 mg/kg) and intravenous diazepam
(0.3 mg/kg) were administered. The main outcome measures were interval between arrival at hospital and starting
treatment and interval between arrival at hospital and
cessation of seizures. Intranasal midazolam and intravenous diazepam were equally effective. Overall 18 of 27
seizures were controlled with midazolam and 15 of 23 with
diazepam. The mean interval between arrival at hospital
and starting treatment was significantly shorter in the
midazolam group [3.37 min (SD 2.46)] as compared to the
diazepam group [14.13 min (SD 3.39)]. The mean interval
between cessation of seizures and arrival at hospital was
significantly shorter in the midazolam group [6.67 min (SD
3.12)] as compared to the diazepam group [17.18 min (SD
5.09)]. The mean interval between control of seizures and
administration of the drug was shorter in the diazepam
group [2.67 min (SD 2.31)] as compared to the midazolam

A. Thakker (&)
Division of Child Neurology and Epilepsy,
Department of Pediatrics, Lokmanya Tilak Municipal Medical
College and General Hospital, B2-504, Gold Coin CHS,
Tardeo Road, Sion, Mumbai 400 034, India
e-mail: arpitathakker@gmail.com
P. Shanbag
Department of Pediatrics, Lokmanya Tilak Municipal Medical
College and General Hospital, Sion, Mumbai 400 034, India

123

group [3.01 min (SD 2.79)]. No significant side effects

were observed in either group. Seizures were controlled
more quickly with intravenous diazepam than with intranasal midazolam. Midazolam was as safe and effective as
diazepam. The overall interval between arrival at hospital
and cessation of seizures was shorter with intranasal
midazolam than with intravenous diazepam. The intranasal
route can be possibly used not only in medical centres, but
with appropriate instruction by the parents of children with
acute seizures at home.
Keywords Intranasal midazolam
Intravenous
diazepam
Seizures
Child

Introduction
Seizures are common in the pediatric age group and occur
in approximately 10 % of children, half of them occurring
before the age of 1 year. Acute onset of childhood seizures
require prompt medical attention, ventilator support and
appropriate oxygenation until they either stop spontaneously or are controlled by drugs.
In the acute treatment of seizures, diazepam is
undoubtedly the benzodiazepine most widely used [1].
However, it has a short duration of action. It is usually
given intravenously and it tends to accumulate, if repeated
doses are given with the possible rare complication of brain
stem depression leading to bradypnea or even respiratory
arrest. The introduction of a venous line may be difficult
particularly in children with generalized tonicclonic seizures [2, 3].
Midazolam, the first water-soluble benzodiazepine is
widely accepted as a parenteral anxiolytic and premedicant
[4]. Midazolam can be given intravenously, intramuscularly

J Neurol (2013) 260:470474

and rectally as well as via the nasal mucosa. It is popular as

a preanesthetic agent, because it is water soluble has a
rapid onset of action, with a relatively short duration of
action and is less likely than diazepam to accumulate. Its
safety and efficacy as an anticonvulsant drug given intramuscularly have been shown in several studies in animals
and humans (adults and children) [2, 3, 5, 6]. Midazolam
given intranasally as an anesthetic agent has been shown to
be safe and effective in children undergoing various diagnostic studies and minor surgical procedures [710].
Intranasal midazolam also suppresses epileptic activity and
improves the background of electroencephalograms in
children with epilepsy [11]. It has been seen that intranasal
midazolam is safe and effective for the management of
acute seizures in children [12].
In the present study, we aimed to compare midazolam
given intranasally with diazepam given intravenously in
the treatment of acute childhood seizures.

Materials and methods

The study was conducted in the pediatric emergency
department of the Lokmanya Tilak Municipal General
Hospital, a tertiary care public hospital in Mumbai, India.
Fifty children between the ages of 1 month and 12 years
who presented with acute motor seizures lasting for at least
10 min were eligible for inclusion in our study. Patients
were recruited over a 12-month period between January
2006 and December 2006. We excluded children with
established intravenous lines or those who had received
anticonvulsants before admission. Children with an ongoing motor seizure for at least 10 min were enrolled,
assuming that spontaneous cessation of such a prolonged
seizure is low;it is the least time it would take to reach the
hospital and most emergency physicians would initiate
anticonvulsive treatment after that time [13].
Subjects were randomized into two groups using a
random number table. Randomization was performed in
advance with a random number table by a pediatric assistant not involved in the study and treatment allocations
were sealed in opaque envelopes. Investigators were blind
to these allocations. Group A received intranasal midazolam (0.2 mg/kg) and Group B received intravenous diazepam (0.3 mg/kg). Midazolam solution (5 mg/ml) was
dripped by syringe into both nostrils in equal doses, and an
intravenous line was immediately introduced.
We recorded the following times: duration of seizure
before anticonvulsant medication, interval between arrival
at hospital and treatment with anticonvulsant medication,
time taken for securing an intravenous line, time taken for
cessation of seizure and recurrence after 60 min of initial
control, if any. Treatment was considered successful, if

471

seizures ceased within 5 min. Seizures that did not stop

within 5 min, but were controlled within 10 min were
defined as successful, but delayed control. Seizures which
did not stop within 10 min were defined as treatment
failures, and intravenous diazepam was given to the
midazolam group and phenobarbital followed by phenytoin
to the diazepam group. Seizures that were controlled with
midazolam or diazepam, but recurred within 60 min were
defined as recurrent seizures.
After seizures were controlled in the children, their
parents were asked to sign a consent form giving permission to enrol them in our study. The hospitals ethics
committee approved the study on the understanding that,
because midazolam is rapidly taken up by the intranasal
route, there would be no significant delay in treating
patients randomized to receive this drug and that if this
treatment failed an intravenous line would immediately be
introduced.
During seizure activity and for 60 min after control, the
children were followed by continuous cardio-respiratory
and pulse oximetry monitors. Vital signs were recorded
every 15 min. During seizure activity, high flow oxygen
was provided through a mask. All the children were
admitted to the pediatric ward for 24 h observation after
cessation of seizure. The results are presented as means
(standard deviations) for continuous data and proportions
for nominal data. The two groups were compared by the
independent sample t test or Fishers exact test.

Results
Fifty children were included in our study between January
2006 and December 2006. Intranasal midazolam was given
for 27 episodes of motor seizures and intravenous diazepam for 23 episodes. The two groups were comparable for
age and sex (Table 1).
There were no differences in the etiology, type and the
duration of seizure prior to being seen in the emergency
group (Tables 2, 3, 4).
Overall, 18 out of 27 seizures responded to initial
treatment with intranasal midazolam and 15 out of 23
responded to intravenous diazepam. However, as this difference was not statistically significant (P [ 0.05), both the
drugs were equally effective in stopping the seizures. Two
children (one with epilepsy and one with neurocysticercosis) had a delayed control in the midazolam group and
one child of pyogenic meningitis had delayed control in the
diazepam group. Eight treatment failures occurred, four in
each group. Two of the four treatment failures in the
midazolam group were controlled with intravenous diazepam, one with intravenous phenytoin and one with intravenous phenobarbital, after intravenous diazepam failed.

123

472

J Neurol (2013) 260:470474

Discussion

Table 1 Age and sex wise distribution

Group

Male

Female

Mean age in years (SD)

Midazolam: A

27

15

12

3.84 (2.93)

Diazepam: B

23

12

11

3.97 (3.33)

Table 2 Etiology of seizure

Etiology

Midazolam: A

Diazepam: B

Febrile convulsion

Seizure disorder

CNS infection

Hypocalcemia

Associated cerebral palsy

Congenital hydrocephalus

4
3

5
2

27

23

Type

Midazolam: A

Diazepam: B

Generalized tonicclonic

Total

Table 3 Type of seizure

14

16

Simple partial seizure

Complex partial seizure

Subtle convulsion

27

23

Total

Table 4 Duration of seizure prior to being seen in the emergency

department
Group

Mean time
in minutes (SD)

Midazolam: A

27

22.30 (16.55)

Diazepam: B

23

22.48 (12.11)

t test for equality

of means
t

df

Sig
(2-tailed)

-0.04

48

0.965

Three treatment failures in the diazepam group required

intravenous phenobarbital, one required both phenytoin
and phenobarbital. Six seizures recurred after initial treatment, three in each group.
Statistically significant differences were found between
the time of treatment and control of seizures (Table 5).
The mean interval for arrival at hospital to starting the
treatment and therefore cessation of seizures was significantly shorter in the midazolam group. The mean interval
for control of seizures after giving the drug was shorter in
the diazepam group. Only one child in the diazepam group
had clinical signs of respiratory depression.

123

Midazolam given intranasally is as safe and as effective as

diazepam given intravenously in the management of acute
seizures in children.
Midazolam, a 1, 4 benzodiazepine agent of the group of
1, 2-unrelated benzodiazepines, is a water soluble compound. Its solubility is pH dependent, below pH 4 it is
freely water-soluble as the ring opens. At physiological pH
of plasma, the ring closes and the drug becomes lipid
soluble and rapidly penetrates the blood brain barrier to
exert its action [14, 15]. Biotransformation of midazolam
occurs in the liver, which involves hydroxylation by
hepatic microsomal oxidative mechanisms. Intranasal
midazolam has the advantage of rapid administration of the
drug into the systemic circulation from an area rich in
blood supply, without the disadvantage of passing through
the portal circulation. Midazolam with high hepatic clearance has much a higher systemic availability following
nasal administration.
As a result of the popularity of intranasal midazolam as
a sedative agent for minor surgical interventions and
diagnostic procedures, there is considerable information on
its use in young children. No significant complications
have been reported, when it is given through the intranasal
route [16]. Therefore, it seemed pertinent to investigate the
use of intranasal midazolam in the management of acute
seizures, especially in children, where the introduction of
an intravenous line is frequently unsuccessful during the
seizures.
To our knowledge there are only three controlled studies
comparing intranasal midazolam with intravenous diazepam for seizures in children [13, 17, 18]. In our study we
found that 18 out of 27 seizures stopped within 5 min of
intranasal midazolam (54.5 %), whereas 15 out of 23 seizures stopped within 5 min of intravenous diazepam
(45.5 %). However, this difference was not statistically
significant (P [ 0.05), i.e., midazolam is as effective as
diazepam in the treatment of acute childhood seizure.
Lahat et al. [13] also reported that intranasal midazolam
and intravenous diazepam are equally effective in children
with prolonged febrile seizures. In his study 23 out of 26
seizures responded to intranasal midazolam and 24 of 26
responded to intravenous diazepam.
We found that the mean interval between arrival at
hospital and starting treatment was significantly shorter in
the midazolam group 3.37 min (SD 2.46) than the diazepam group 14.13 min (SD 3.39). Hence, intranasal
midazolam is easier and faster to administer than intravenous diazepam. The mean interval between cessation of
seizures and arrival at hospital was significantly shorter in
the midazolam group 6.67 min (SD 3.12) than the diazepam group 17.18 min (SD 5.09). This difference is due to

J Neurol (2013) 260:470474

473

Table 5 Duration of time interval (in minutes) for giving the drug, for seizure control and for response to treatment in the study groups
Time in minutes
Midazolam: A

t test for equality of means

Diazepam: B

df

Sig
(2-tailed)

Interval between giving drug and arrival at hospital

3.37 (2.46)

14.13 (3.39)

-12.89

48

0.00

Interval between cessation of seizure and giving drug

3.01 (2.79)

2.67 (2.31)

0.34

48

0.05

Interval between cessation of seizure and arrival at hospital

6.67 (3.12)

17.18 (5.09)

0.10

41

0.00

Values are means (standard deviation)

the difficulty in introduction of an intravenous line in an

actively convulsing child, hence, the delay in administration of the drug in the diazepam group. However, the mean
interval between control of seizures after giving the drug
was faster in the diazepam group 2.67 min (SD 2.31) than
in the midazolam group 3.01 min (SD 2.79). Lahat et al.
[13] also showed that the mean time to cessation of seizures after giving the drug was faster in the diazepam
group 2.5 min (SD 1.9) than the midazolam group 3.1 min
(SD 2.2). Mahmoudian and colleagues [17] showed that the
mean time to seizure control after giving the drug was
significantly shorter in the diazepam group 2.94 min (SD
2.62) than in the midazolam group 3.58 min (SD 1.68), not
counting the time required to insert the intravenous line. In
his study, he also concluded that if the time taken for
introduction of an intravenous line is added to the time to
cessation of seizures, treatment with intranasal midazolam
may work faster than treatment with intravenous diazepam.
Mittal et al. [18] found that the mean time to control the
seizure by intravenous diazepam, after arrival in hospital
was maximum in the age group 01 year and minimum in
the age group above 6 years, this was attributed to difficulty in establishing an intravenous access in younger age
groups.
There were no recurrences of seizures within 60 min of
treatment in other studies. In our study six children had
recurrence of seizures after initial control, but these were
equally distributed in both the groups. The incidence of
treatment failures was also equal in both groups. Hence,
intranasal midazolam is as effective as intravenous diazepam in the treatment of acute childhood seizures.
In our study, one child in the diazepam group had bradypnea after rapid administration of intravenous diazepam.
None of the children in the midazolam group had any
adverse effects. However, Lahat et al., reported that none
of the children in either group had clinical signs of respiratory distress or bradycardia. The safety and efficacy of
midazolam has been shown by several clinical studies in
epileptic adults and children [2, 3, 6]. Midazolam given
intravenously or intramuscularly is not associated with
respiratory changes although, there are reported associations with hypertension, bradycardia and hypoxia in adults

and children. These changes were however, mild and

transient. No patients had to be intubated or mechanically
ventilated [19, 20].
Traditional rectal, oral and intramuscular premedicants
have been helpful, but each has its own disadvantage and
limitations [2127]. Rectal diazepam is established as a
standard rescue or emergency treatment for seizure or
status epilepticus; however, the rectal route of administration has not been universally accepted [2125]. To
determine, if an alternative route of administration of a
benzodiazepine was equally effective, De Haan et al. [24]
has compared a novel midazolam HCl concentrated nasal
spray (MDZ-n) with diazepam rectal solution (DZP-r) in
the treatment of prolonged seizures in a residential adult
epilepsy center. MDZ-n was equal to DZP-r with respect to
efficacy and side effects in the suppression of seizure
exacerbations. The nasal spray was preferred to the rectal
solution by 16 of 21 caregivers and patients conjointly.
Bhattacharya et al. [25] has also shown that intranasal
midazolam is preferable to rectal diazepam for treatment of
acute seizures in children.
Comparing the efficacy of buccal midazolam versus
rectal diazepam, Nakken and Lossius [26] have shown that
in the treatment of serial seizures or status epilepticus of
residential adult patients, buccal midazolam appeared
equally effective as rectal diazepam with little or no
adverse effects and also showed that the buccal route was
socially more acceptable. Mcintyre et al. [27] has shown in
his study that buccal midazolam was more effective than
rectal diazepam for children presenting to hospital with
acute seizures and was not associated with an increased
incidence of respiratory depression. He also mentioned that
buccal administration is preferred over the nasal route due
to local irritation of the nasal mucosa and less reliable
absorption in the presence of concurrent upper respiratory
tract infection.
Intramuscular administration is painful, absorption may
be erratic, but it is easy to administer. Silbergleit et al. [28]
has compared intramuscular versus intravenous therapy for
prehospital status epilepticus. He has shown in his study
that intramuscular midazolam is as safe and effective as
intravenous lorazepam for prehospital seizure cessation in

123

474

J Neurol (2013) 260:470474

children and adults with status epilepticus. He has also

shown that intramuscular midazolam is easy to administer
and improves the speed of treatment administered by
emergency medical services personnel. The median time to
active treatment were 1.2 min in the intramuscular
midazolam group and 4.8 min in the intravenous lorazepam group, with corresponding median times from active
treatment to cessation of convulsions of 3.3 and 1.6 min.
The possibility of administration of intranasal midazolam not only in medical centers, but with appropriate
instruction, by the parents of children with acute seizures at
home, is also worth consideration. However, further studies
in the home setting are needed on a larger series of children
before a definite recommendation can be made.
Acknowledgments We thank our Dean, Dr Sandhya Kamath, and
Head of Department, Dr Mamta V Manglani, for giving us the permission to publish this article.
Conflicts of interest

None to declare.

Ethical standard This Randomised Controlled Trial has approval

of Ethics Committee, Staff and Research Society, Lokmanya Tilak
Municipal Medical College and General Hospital, Mumbai, India.

References
1. (1993) Treatment of convulsive status epilepticus: recommendations of the Epilepsy Foundation of Americas Working Group
in status epilepticus. JAMA 270:85449
2. Chamberlain JM, Altieri MA, Futterman C et al (1997) A prospective, randomized study comparing intramuscular midazolam
with intravenous diazepam for the treatment of seizures in children. Pediatr Emerg Care 13:9294
3. Lahat E, Aladjem M, Eshel G et al (1992) Midazolam in treatment of epileptic seizures. Pediatr Neurol 8:215216
4. Conner JT, Katz RL, Pagano RP et al. (1978) PO21-3981 for
intravenous surgical pre-medication and induction of anesthesia.
Anesth Analg 57:15
5. De Jong RH, Bonin JD (1981) Benzodiazepines protect mice
from local anesthetic convulsions and death. Anesth Analg
60:385389
6. Jawad S, Oxley J, Wilson J et al (1988) A pharmacodynamic
evaluation of midazolam as an antiepileptic compound. J Neurol
Neurosurg Psychiatr 49:10501054
7. Latson LA, Cheatham JP, Gumbiner CH et al (1991) Midazolam
nose drops for outpatient echocardiography sedation in infants.
Am Heart J 121:209210
8. Lacon A, Reddy VG (1994) Nasal midazolam and ketamine for
pediatric sedation during computerized tomography. Acta Anesthesiol Scand 38:259261
9. Wilton NCT, Leight J, Rosen DR et al (1988) Preanesthetic
sedation of preschool children using intranasal midazolam.
Anesthesiology 69:972

123

10. Saint-Maurice C, Landais A, Delleur MM et al (1990) The use of

midazolam in diagnostic and short surgical procedures in children. Acta Anesthisiol Scand 94(suppl 92):3941
11. ORegan ME, Brown JK, Clarke M (1996) Nasal rather than
rectal benzodiazepines in the management of acute childhood
seizures. Develop Med Child Neurol 38:10371045
12. Lahat E, Goldman M, Barr J et al (1998) Intranasal midazolam
for childhood seizures. Lancet 352:620
13. Lahat E, Goldman M, Barr J et al. (2000) Comparison of intranasal midazolam with intravenous diazepam for treating febrile
seizures in children: prospective randomized study. BMJ 8386
14. Greenblatt DJ, Arendt RM, Abernethy DR et al (1983) In vitro
quantitation of benzodiazepine lipophilicity: relation to in vivo
distribution. Br J Anesth 55:985989
15. Arendt RM, Greenblatt DJ, de Jong RH et al (1983) In vivo
correlates of benzodiazepine cerebrospinal fluid uptake pharmacodynamic action and peripheral distribution. J Pharmacol Exp
Ther 227:98106
16. Rey E, Delaunay G, Pons IM et al (1991) Pharmacokoinetics of
midazolam in children: comparative study of intranasal and
intravenous administration. Eur J Clin Pharmacol 41:355357
17. Mahmoudian T, Zadeh MM (2004) Comparison of intranasal
midazolam with intravenous diazepam for treating acute seizures
in children. Epilepsy Behav 5:253255
18. Mittal P, Manohar R, Rawat AK (2006) Comparative study of
intranasal midazolam and intravenous diazepam sedation for
procedures and seizures. Indian J Pediatr 73:975978
19. Koul RL, AIthala GR, Chacko A et al (1997) Continuous
midazolam infusion as treatment of status epilepticus. Arch Dis
Child 76:445448
20. Ghialin S, Van Ruckevorsel-Harmant K, Harmant J et al (1988)
Midazolam in the treatment of epileptic seizures. J Neurol Neurosurg Psychiatry 51:732735
21. Scott RC, Nevile BGR, Besag FMC et al (1997) Nasal rather than
rectal benzodiazepines in the management of acute childhood
seizures. Develop Med Child Neurol 39:137138
22. Lombroso CT (1989) Intermittent home treatment for status and
clusters of seizures. Epilepsia 30(suppl 2):1114S
23. Seigler RS (1990) The administration of rectal diazepam for
acute management of seizures. J Emerg Med 8:155159
24. De Haan GJ, Van Der Geest P, Doelman G et al (2010) A
comparison of midazolam nasal spray and diazepam rectal
solution for the residential treatment of seizure exacerbations.
Epilepsia 51:478482
25. Bhattacharya M, Kalra V, Gulati S (2006) Intranasal midazolam
vs rectal diazepam in acute childhood seizures. Pediatr Neurol
34:355359
26. Nakken KO, Lossius MI (2011) Buccal midazolam or rectal
diazepam for treatment of residential adult patients with serial
seizure or status epilepticus. Acta Neurol Scand 124:99103
27. Mcintyre J, Robertson S, Norris E et al (2005) Safety and efficacy
of buccal midazolam versus rectal diazepam for emergency
treatment of seizures in children: a randomized controlled trial.
Lancet 366:205210
28. Silbergleit R, Durkalski V, Lowenstein D, Conwit R, Pancioli A,
Palesch Y, NETT Investigators (2012) Intramuscular versus
intravenous therapy for prehospital status epilepticus. N Engl J
Med 366(7):591600