Epileptic seizures are a common cause of pediatric emergency department visits. The vast majority of seizures cease within 5min; however, some are prolonged and may progress to status epilepticus (SE).1 SE, defined as continuous or recurring seizure activity lasting longer than 30min, is associated with major morbidity rate and carries a mortality rate of up to 20%.2 Prolonged seizure activity leads to failure of cerebral autoregulation, which consequently reduces cerebral blood flow and eventually results in cerebral hypoxia.2,3

Benzodiazepines (BZDs) have been used in the urgent care of seizures for over 40 years, and are considered first-line therapy for this purpose. The BZDs, which act on GABAA receptors, are effective in the treatment of various types of seizure, have a rapid onset of action after intravenous administration, excellent penetration into the central nervous system, and a good safety profile. Persistent generalized seizure activity increases benzodiazepine resistance; therefore, there is a consensus as to the need for immediate treatment of seizures.2,4–8 Lorazepam, diazepam, and midazolam are the three BZDs used in this setting. Lorazepam is the drug of choice due to its rapid onset of action and prolonged effect, but no parenteral formulations are available in Brazil. Consequently, intravenous diazepam, which provides seizure control in 85–90% of the cases,2,3 is the first-line drug of choice for acute treatment of seizures in Brazilian practice.

When venous access is challenging or cannot be obtained, the most common alternatives have been rectal diazepam and intranasal midazolam, with some studies suggesting intramuscular midazolam as an additional option.7,9 The difficulty of administering BZDs rectally or intranasally and the erratic absorption provided by these routes jeopardize anticonvulsant efficacy.5,6,8,10–13 Median time to achievement of intravenous access in children ranges from 5 to 7min.8,11 Difficulties in obtaining venous access delay treatment and, consequently, increase the risk of progression to status epilepticus. Thus far, no consensus has been established as to the best route for administration of BZDs in the event of failed intravenous access.1,5

The objective of this study was to compare the therapeutic efficacy of intramuscular midazolam (MDZ-IM) with that of intravenous diazepam (DZP-IV) in children admitted to the referral service of a pediatric emergency department with epileptic seizures.

This randomized clinical trial was carried out between August 2010 and August 2011 in a sample of children admitted to the pediatric emergency department of Santa Maria University Hospital (PSPed-HUSM) with epileptic seizures. PSPed is a tertiary referral center in Southern Brazil, staffed by dedicated pediatricians and pediatric residents, which treats approximately 4000 pediatric patients per month. The study was approved by the local Research Ethics Committee (protocol no. 0184.0.243.000-10) in accordance with the Declaration of Helsinki 1964 revised in 2008. In view of the nature of the study condition and to provide the best possible benefit in this urgent setting, the committee authorized immediate patient allocation and treatment without prior consent. Subjects’ parents or legal guardians were then notified of the trial, and those who agreed to the use of patient data for research purposes were asked to provide written informed consent.

The study protocol was presented by the authors and approved by the medical officers and nursing staff of the department, who received specific training in its use.

Inclusion criteria: children between the ages of 2 months and 14 years who were admitted to the emergency department with seizures, regardless of type or potential trigger, and in whom anticonvulsant medication was indicated and ordered by the attending physician. Children could be enrolled in the study more than once.

Exclusion criteria: children in whom venous access had been obtained in the prehospital setting and those with a known history of coagulopathy or hepatic and/or renal impairment.

Patients admitted to the pediatric emergency department with seizures were randomly allocated to one of the two treatment groups: (a) Intravenous diazepam (DZP-IV), 0.5mg/kg IV (maximum dose 10mg) at a concentration of 5mg/ml and with an application speed of 5mg/min, or (b) intramuscular midazolam (MDZ-IM), 0.5mg/kg IM (maximum dose 15mg) at a concentration of 5mg/ml.

Patients were randomized in blocks of 10. Five slips of paper marked “DZP-IV” and five slips marked “MDZ-IM” were placed into a brown paper envelope. At the time of admission, the nurse in charge of medication administration took a slip of paper from the envelope at random and administered the corresponding drug as standardized in the study protocol, after adjusting the dose for weight and age.

Identifying information and physiological parameters were recorded. The following variables were used as outcome measures: time from admission to drug administration (including time required to obtain intravenous access), time from drug administration to cessation of seizures, and total time from admission to cessation of seizures.

Heart rate and pulse oximetry were monitored in all patients throughout treatment. Vital signs were recorded on admission and every 5min thereafter until discharge or transfer. Airway suctioning, supplemental oxygen, or tracheal intubation were provided as necessary or when ordered by the attending physician. Immediate adverse drug reactions were assessed in the first 10min after administration of MDZ-IM or DZP-IV. The emergency department where the study was performed is equipped with crash carts containing all the necessary equipment and medications for treatment of cardiorespiratory instability.

Treatment was considered successful when cessation of seizures was achieved within 5min of administration of a single dose of study drug (DZP or MDZ). Requirement of a second dose of study drug or additional medications was defined as treatment failure.

Inability to obtain venous access within 4min was defined as “failed intravenous access.” This 4-min cutoff was chosen on the basis of mean time to establish intravenous access by Pediatric Emergency nursing staff during emergency care.

Statistical analysis: continuous variables were expressed as means and standard deviations. Comparisons were performed using Student's t-test (for normally distributed variables) or the Mann–Whitney U test (in case of wide variability). Categorical variables were expressed as percentages and comparisons were performed with the chi-square or Fisher's exact tests.

Of the 180 children admitted to the Pediatric Emergency Department at Santa Maria University Hospital (Brazil) with a chief complaint of convulsive seizures, 144 were postictal on arrival. The remaining 36 children were enrolled in the study. Of these, 16 were allocated to the MDZ-IM group and 20 to the DZP-IV group. All children allocated to the midazolam group completed the study. Of those allocated to the diazepam arm, four (20%) were excluded due to failed intravenous access, and the remaining 16 completed the study (Fig. 1). The parents or guardians of all 32 children who completed the study had provided informed consent for participation.

Figure 1.

Study flowchart.

(0.06MB).

There were no significant between-group differences in age, weight, sex, seizure etiology, vital signs, or blood glucose levels on admission. In both groups, most children were under the age of 5 (Table 1).

In 14 patients in each group (87.5%), seizure activity ceased after administration of a single BZD dose, which shows that both intramuscular midazolam and intravenous diazepam are effective anticonvulsants. Two patients in each group (12.5%) required a second dose of study drug or additional medications for seizure control (treatment failure).

In the MDZ-IM group, active treatment was instituted significantly sooner than in the DZP-IV group (2.8 versus 7.4min; p=0.001). Total time to cessation of seizures was also significantly shorter in the MZD-IM group (time between admission and seizure cessation, 7.3 versus 10.6min; p=0.006) (Table 2). However, time from drug administration to cessation of seizures was significantly shorter in the DZP-IV group (3.3 versus 4.4min; p=0.001).

As noted above, two patients in each group (12.5%) were considered to have failed treatment due to persistence of seizures more than 5min after BZD administration. The between-group differences in time to active treatment and time to cessation of seizures held true regardless of inclusion or exclusion of the patients that failed treatment (Table 2).

All children in both groups exhibited cyanosis and psychomotor agitation on admission and required supplemental oxygen by nasal cannula or non-rebreather mask. One child in the DZP-IV group progressed to SE of 40-min duration, requiring phenytoin for seizure control, and one child had a seizure duration of 6min, requiring only an additional dose of IV diazepam. In the MDZ-IM group, one child had a seizure lasting 10min, requiring intravenous diazepam and phenytoin. Additionally, one child in the MDZ group had SE that lasted 55min and did not respond to intravenous diazepam. In this case, cessation was achieved after rectal diazepam and an additional dose of intramuscular midazolam.

One child (6.25%) in the MDZ group required intubation, artificial ventilation, and ICU admission due to respiratory failure on admission to the ED (SaO2=64%). Two children (12.5%) in the DZP group required intubation and ICU admission (one due to severe head trauma and one due to respiratory failure with SaO2=80%).

All children were monitored until cessation of seizures or discharge. There were no significant between-group differences in vital signs (Fig. 2).

Figure 2.

Vital signs (systolic and diastolic blood pressure, heart rate, respiratory rate) of children in the IM midazolam and IV diazepam groups, at 5min, 10min, and discharge or transfer.

(0.29MB).

Possible immediate adverse drug reactions (those occurring within 10min of study drug administration) included hypotension in one child (6.3%) in the MDZ-IM group, two children (12.6%) with hyperactivity and salivation, one (6.3%) with nausea, and two (12.6%) with vomiting in the DZP-IV group. Nevertheless, the between-group difference in adverse reactions did not reach statistical significance (Table 2).