- Email: [email protected]
Diazepam and intubation in emergency treatment of seizures in children
ORIGINAL CONTRIBUTION diazepam; seizures, pediatric
Diazepam and Intubation in Emergency Treatment of Seizures in Children Study objectives: This study was undertaken to determine the incidence of endotracheal intubation after the use of diazepam compared with phenobarbital or phenytoin in emergency treatment of seizures in children. Design: The records of all children (98) were reviewed in a case-control fashion. A logistic regression model was used to determine whether there was an association between diazepam administration and intubation, adjusting for all other covariates (age, weight, convulsion time before first anticonvulsant was given, response latency, diagnosis, and therapy). Setting: All children were treated in an emergency department and then transported to a tertiary pediatric center by a pediatric transport team. Type of participants: All patients were children, w~th a median age of 2.7 years flange, 0.17 to 15.3 years). Interventions: None. Measurements and main results: Only the use of diazepam was found to be significantly associated with intubation after adjusting for all other covariates (adjusted odds ratio, 49.4; P < .001). In the comparison of diazepam versus phenobarbital or phenytoin used as the first anticonvuIsan& median response latency did not differ (27.0 vs 32.5 minutes, P > .83). Conclusion: A significant association was found between diazepam use and intubation. Response latency was not shorter when diazepam was used as the initial anticonvulsant compared with phenobarbital or phenytoin. [Orr RA, Dimand R J, Venkataraman ST, Karr VA, Kennedy KJ: Diazepam and intubation in emergency treatment of seizures in children. Ann Emerg Med September 1991;20:1009-1013.]
INTRODUCTION Children are often seen in the emergency departments of nonpediatric hospitals for the treatment of seizures before they are referred to a pediatric center. Diazepam has been recommended for the initial treatment of seizures 1-9 because of its presumed rapid onset of action compared with other antieonvulsants. The emergency physician, who is responsible for treating the acute onset of seizures, thus may choose diazepam as the initial drug. Apnea and respiratory depression are well-known side effects of diazepaml,6, lo-13 that may necessitate intubation for both airway protection and assisted ventilation. These conditions demand personnel who are able to recognize respiratory compromise and are skilled in the management of the pediatric airway. It has been shown that prehospital personnel have difficulty in airway management in children. 14 However, many emergency medical services crews carry diazepam for the management of seizures in children. The incidence of apnea and respiratory depression requiring intubation after the use of diazepam in this setting is not known. All patients who were transported to our institution by the interhospital transport team from other EDs and had a diagnosis of seizures requiring anticonvulsant therapy were studied in a case-control fashion. The goals of this study were to determine the incidence of endotracheal intubation in these patients, to identify factors associated with intubation, and to test the hypothesis that diazepam results in a greater incidence of intubation while offering no better control in stopping seizures than phenobarbital or phenytoin.
20:9 September 1991
Annals of Emergency Medicine
Richard A Orr, MD* Robert J Dimand, MD* Shekhar T Venkataraman, MD* Vaterie A Karrt Kathleen J KennedylPittsburgh, Pennsylvania From the Departments of Anesthesiology/ Critical Care Medicine and Pediatrics, University of Pittsburgh School of Medicine;* and the Transport Team, Department of Nursing, Children's Hospital ef Pittsburgh,t Pittsburgh, Pennsylvania. Received for publication May 23, 1989. Revisions received September 18, 1990, and February 14, 1991. Accepted for publication March 25, 1991. Presented in part at the University Association for Emergency Medicine Annual Meeting in Cincinnati, Ohio, May 1988. Address for reprints: Richard A Orr, MD, Department of Anesthesiology/Critical Care, University of Pittsburgh School of Medicine, Children's Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, Pennsylvania 15213.
1009/107
DIAZEPAM & INTUBATION Orr et al
lOO 80
0--0 e--I
~
,9
- - / e /
~i
60
°
e~
o,.~/o /
20
I
I
<10
10 29
I
I
30 49 50-69
I
I
70 119
>120
RespDnseLatency(rnin)
F I G U R E . Cumulative frequency of i n t u b a t e d versus n o n i n t u b a t e d patients for response latency.
MATERIALS A N D M E T H O D S The records of all children w i t h a h i s t o r y of seizures w h o were transported to our hospital by our transport t e a m between October 1986 and September 1987 were reviewed. Children were i n c l u d e d in the s t u d y if they had seizures w i t n e s s e d by personnel in the referring ED and t h e y received at least one anticonvulsant. Excluded were children who had seizures not w i t n e s s e d by the referring ED or w h o were i n t u b a t e d before a w i t n e s s e d seizure or a n t i c o n v u l s a n t therapy. For each child, the data recorded included age, weight, convulsion t i m e (minutes) before the first dose of anticonvulsant, elapsed t i m e from i n i t i a t i o n of t h e r a p y to cessat i o n of c l i n i c a l c o n v u l s i v e m o v e ments (response latency), treatment, p r e s e n c e or a b s e n c e of i n t u b a t i o n , and diagnosis. Diagnostic categories were defined as prior seizure disorder (included patients with a confirmed diagnosis of e p i l e p s y ) , s e i z u r e w i t h f e v e r (inc l u d e d p a t i e n t s w h o s e s e i z u r e was associated with fever but did not have evidence of central nervous syst e m [CNS] infection or a history of a s e i z u r e disorder), m e t a b o l i c c a u s e s (included patients whose seizure was r e l a t e d to a m e t a b o l i c d e r a n g e m e n t ( s u c h as h y p o g l y c e m i a or h y p o calcemia), CNS infection (included p a t i e n t s w h o s e s e i z u r e w a s associated w i t h a d o c u m e n t e d CNS infection such as m e n i n g i t i s or encephalitis), a n d i d i o p a t h i c or t u m o r (inc l u d e d p a t i e n t s w h o s e s e i z u r e was associated w i t h an intracranial lesion and those in w h o m a cause for the seizure was never determined). Diagnosis was assigned on review of the records. Children were often t r e a t e d w i t h an anticonvulsant in the ED before a 108/1010
TABLE l. Comparison of intubated ~ t h nonintubated patients
--//O
Inlubatedpafierls Nonntubatedpaden!s
No. of patients Age (yr) Age range (yr) Diagnosis Prior seizure disorder Seizure with fever Metabolic CNS infection Tumor Idiopathic
Convulsion time before anticonvulsant (min)
Intubated
Nol Intubated
(%)
(%)
32 4.5
22 1.3
0.5 - 15.3
0.2 - 10.0
5 (15.6) 10 (31.3) 2 (6.3) 5 (15.6) 1 (3.1) 9 (28.1)
6 5 4 3 0 4
< .01"
> .50t
1 - 180
1
30.0 145
> .25*
Range (min)
Response latency (rain)
40.0 1 360
15.0 1 150
> .30*
8 (25.0) 19 (59.4) 5 (15.6) 27 (84.4) 13 (40.6) 5 (15.6) 24 (75.0) 3 (9.4) 5 (15.6) 12 (37.5) 16 (50.0) 4 (12.5)
1 (4.5) 6 (27.3) 15 (68.2) 7 (31.8) 9 (40.9) 14 (63.6) 6 (27.3) 2 (9.1) 14 (63.6) 12 (54.5) 9 (40.9) 1 (4.5)
0.23 0.06 - 1.50
0.140 0.11 - 2.30
Range (min)
16.5
(27.3) (22.7) (18.2) (13.6) (0.0) (18.2)
P
Drug therapy Diazepam only Diazepam plus phenobarbital, phenytoin, or both No diazepam Ever used diazepam Ever used phenytoin Ever used phenobarbital First used diazepam First used phenytoin First used phenobarbital Used only one drug Used two drugs Used three drugs
Total dose of diazepam for patients receiving diazepam (mg/kg) Range (mg/kg)
< .001, < .0001§ > .90§ > .09§ < .002*
> .35t
> .50*
Results of continuous variables are expressed as median and range• *Mann-Whitney rank sum test; tFreeman Halton test; :x 2 test for independent samples; 9Fisher's exact test, two-tailed.
diagnosis was d e t e r m i n e d . N o n e of the " n o n i n t u b a t e d " children required b a g - v a l v e - m a s k v e n t i l a t i o n at a n y time. Adequacy of the initial dose of a n t i c o n v u l s a n t (mg/kg) was determ i n e d for each patient based on the r e c o m m e n d a t i o n of t h e A m e r i c a n A c a d e m y of Pediatrics C o m m i t t e e on Drugs. * 15 Statistical analyses were performed using t h e M a n n - W h i t n e y r a n k s u m test to compare c o n t i n u o u s variables between two groups and the Fisher's e x a c t test, F r e e m a n - H a l t o n test~ 16 and X2 for categorical variables where appropriate. P < .05 was considered significant. A univariate logistic regression *The minimum recommended dose for each anticonvulsant was considered an adequate initial dose: diazepam, 0.3 mg/ kg; phenobarbital, 10 mg/kg; phenytoin, 10 mg/kg. Annals of Emergency Medicine
m o d e l 17 was used to test the association b e t w e e n i n t u b a t i o n and each of the continuous v a r i a b l e s (age, weight, convulsion time, response latency, n u m b e r of drugs, and drug doses) as well as b e t w e e n i n t u b a t i o n and categorical variables (diagnosis and drug therapy). T h e u n a d j u s t e d o d d s r a t i o s a n d 95% c o n f i d e n c e l i m i t s were calculated for each indep e n d e n t variable separately. A stepwise logistic regression analysis t h e n was performed to fit the m o s t parsim o n i o u s model. A m u l t i v a r i a t e logistic regression model was used to determine w h e t h e r there was an association between diazepam administration.and i n t u b a t i o n , a d j u s t i n g for all o t h e r covariates, including adequacy of the initial dose of anticonvulsant. A statistically significant association w i t h i n t u b a t i o n was d e f i n e d as t h a t for w h i c h the 95% confidence i n t e r v a l 20:9 September 1991
DIAZEPAM & INTUBATION Orr et al
TABLE 2. Comparison of patients: Diazepam first versus phenobarbital or phenytoin first
Diazepam First (%) 30 4.5 0.25 15.0
No. of patients Age (yr) Age range (yr) Convulsion time before anticonvulsant (min) Range (min) Response latency (min) Range (rain) Drug therapy Diazepam only Diazepam plus phenobarbital, phenytoin, or both Ever used phenytoin Ever used phenobarbital Used only one drug Used two drugs Used three drugs Intubated
Phenobarbital or Phenytoin First (%) 24 1.5 0.17 15.3
27.0 1.0 180.0 27.0 1.0 360.0
20.0 1.0 145.0 32.5 1.0 - 255.0
9 (30.0)
0 (0.00)
2! (70.0) 13 (43.3) 12 (40.0) 9 (30.3) 17 (56.7) 4 (13.3) 24 (80)
4 (16.7) 9 (37.5) 20 (83.3) 15 (62.5) 8 (33.3) 1 (4.2) 8 (33.3)
P < .03*
> .62" > .83*
> .80t < .O02t < .05~ < .001t
Results of continuous variables are expressed as median and range. *Mann-Whitney rank sum lest; tFisher's exact test, two tared; cFreeman Halton test.
TABLE 3. Univariate logistic regression analyses - Continuous variables versus intubation
Age Weight Convulsion time before an0convulsant Response latency No. of drugs Total dose of diazepam
Odds Ratio 1.27 1.09
95% Confidence Intervals (1.04, 1.66) (t.00, 1.17)
0.99 1.01 1.89 2.77
(0.98, 1.01) (1.00, 1.02) (0.77, 4.63) (0.37, 20.69)
about the odds ratio did not contain 1.0.
RESULTS Because of the nonparametric nature of the data, results of continuous variables are expressed as median and range. 18 Ninety-eight patients were transported to our hospital by our team during the study period; 54 children met the inclusion criteria. Median age was 2.7 years (range, 0.17 to 15.3 years). Median c o n v u l s i o n time before the first dose of anticonvulsant was 20.0 minutes (range, 1.0 to 180.0 minutes), and median response latency was 30.0 m i n u t e s (range, 1.0 to 360.0 minutes). Intubated children were older (P < 20:9 September 1991
.01) and more likely to have received d i a z e p a m than were those in the group that did not receive diazepam (Table 1). Convulsion time before adm i n i s t r a t i o n of an anticonvulsant, response latency, number of anticonvulsants, and total dose of diazepam for patients receiving diazepam did not significantly differ between the two groups (Table 1 and Figure). The use of diazepam, either alone or in combination with other anticonvulsants, was found to be significantly associated with intubation (P < .001). Eighty-four percent of intubated patients had received diazepam compared with only 31.8% of nonintubated patients {P < .0001). Eighty percent of the patients who received diazepam as the first anticonvulsant required intubation (P < .001 (Table 2). They were also older than those in the group that received phenobarbital or phenytoin first (P < .03). Convulsion time (minutes) before administration of an anticonvulsant and response latency did not differ between these two groups. T h i r t y - n i n e of the 54 c h i l d r e n (72.2%) received an inadequate initial dose of anticonvulsant. Seven of 32 intubated children (21.9%) had an adequate dose for the first anticonvulsant administered compared with eight of 22 n o n i n t u b a t e d children (36.4%) (P > .24, Fisher's exact test). Annals of Emergency Medicine
T w e n t y - f i v e of t h e 39 c h i l d r e n (64.1%) whose initial dose was inadequate were intubated. Of these, 19 of 25 (76.0%) were initially given diazepam. Nine of the 14 nonintubated children (64.3%) whose initial dose of a n t i c o n v u l s a n t was inadequate were never given diazepam. Adequacy of dose did not contribute significantly to the odds of intubation (P > .24). With univariate logistic regression analysis used for continuous variables, statistically significant associations were found between intubation and both age and weight (Table 3). The unadjusted odds ratios relating age and weight to intubation were 1.27 (P < .01) and 1.09 (P < .05), respectively. Because age and weight are related (Spearman's correlation, .92), a logistic regression analysis shows that after adjusting for age, the association between weight and intubarton is no longer statistically significant (P > .7). Statistically significant associations were not found bet w e e n i n t u b a t i o n and c o n v u l s i o n time, response latency, n u m b e r of drugs, or dose of diazepam. For categorical variables, use of diazepam at any time (odds ratio, 11.57~ P < .001), alone (odds ratio, 24.0; P < .001), or in combination with phenobarbital, phenytoin, or both (odds ratio, 9.50; P < .0001) was significantly associated with intubation. Diagnosis was not significantly associated with intubation. In a stepwise logistic regression analysis, administration of diazepam entered the model on the first step (P < .001) as being significantly associated with intubation. After adjusting for age and weight, diazepam use still was found to be significantly associated w i t h i n t u b a t i o n (odds ratio, 8.12; P < .0001). Diagnosis was not significantly associated with intubation (P > .4) after adjustment for diazepam use. When using a multivariate logistic regression analysis, only the use of diazepam was found to be significantly associated with intubation after adjusting for confounding variables (adjusted odds ratio, 49.39; P < .001) (Table 4). DISCUSSION Since Gastaut et al first reported the use of parenteral diazepam in the treatment of status epilepticus, 6 this drug has received wide attention and 1Oll/109
DIAZEPAM & INTUBATION Orr et al
has been used in many institutions as the drug of choice for this purpose in both adults and children 1-9 despite its well-known complication of respiratory depression. 1,6,lO-13 Because seizures may be associated with neuronal cell damage and s e c o n d a r y metabolic consequences, DelgadoEscueta et al suggested that convulsions should be stopped as soon as possible. 1 Thus, one may choose to use a drug that offers more rapid control of seizures than other anticonvulsants despite its effect on the respiratory system, while being prep a r e d to c o n t r o l t h e a i r w a y . Intravenous diazepam has therefore been suggested as the drug of choice because of its presumed rapidity in control of clinical seizures. 1 Approximately 20% of all pediatric interhospital transports performed by our transport team involve patients with seizures, many of whom have respiratory compromise as indicated by the frequency of intubation. Sixtyfive percent of our study population had received diazepam for treatment of seizures; 55% had received it as the initial drug. There has been a lack of carefully controlled clinical studies comparing the various anticonvulsant agents for their effectiveness in treating status epilepticus in both adults and children. 19 In this study, diazepam, as initial or subsequent therapy, offered no better control of seizures (response latency) than phenobarbital, phenytoin, or both. Furthermore, patients who received diazepam were more likely to require intubation, even after adjustment for other potentially confounding variables. In a prospective study in adults, Shaner et al found that a combination of diazepam and phenytoin given as initial anticonvulsant therapy in status epilepticns did not prove to be superior to phenobarbital. 20 In fact, they found that both the cumulative convulsion time and the response latency were shorter for patients given phenobarbital as the initial anticonvulsant. This raises the issue of whether diazepam used as the first drug in treating seizures offers any advantage over phenobarbital. This case-control study involved a select group of patients who were transported to a pediatric center with witnessed seizures requiring therapy. Information on the number of patients treated for seizures who were 110/1012
TABLE 4. Multivariate logistic regression model: Association of diazepam with intubation, adjusting for other covariates Odds Ratio Age Weight Diagnosis (Baseline = Prior seizure disorder) Seizure with fever Metabolic cause CNS infection Idiopathic or tumor
1.62 0.96 1.00 46.50 2.07 3.34 11.78
Convulsion time before anticonvulsants Response latency Therapy (Ever versus never used) Phenytoin Phenobarbital Diazepam Total dose of diazepam Adequacy of dose
1.00 1.01 0.09 2.09 49.39 1.18 0.50
not referred and the timing of administration of diazepam in each case was difficult to obtain. Patients with prolonged seizures may be at greater risk for airway compromise in general, making them more likely to require intubation regardless of the anticonvulsant used. If that were so, the intubation rate should have been similar in patients with and without diazepam. It is also possible that diazepam was more likely to have been used in patients whose seizures were deemed more prolonged, in an attempt to gain rapid control, making these patients likely candidates for airway control. Arterial blood gases reflecting the acid-base status of these patients were rarely available in our population. Nevertheless, convulsion times before the first dose of anticonvulsant were similar for the patients who received diazepam and those who had not. Hypoxia and traumatic injuries to the airway may be more likely when intubation is performed in an uncontrolled setting or by an operator who is not trained in management of the pediatric airway. If an alternative ant i c o n v u l s a n t drug were k n o w n to have equal efficacy and a decreased incidence of respiratory compromise, it might be preferred to diazepam as first-line therapy in the control of seizures. A n o t h e r benzodiazepine, lorazepam, is four times more potent than diazepam in stopping status epilepticus in animals, and its 'efficacy Annals of Emergency Medicine
95% Confidence Intervals
(0.66, 4.01) (0.73, 1.25)
(0.80, (0.02, (0.09, (0.30,
2,709.28) 26922) 123.24) 461.21)
(0.96, 1.03) (0.99, 1.04) (0.01, (0.22, (1.62, (0.05, (0.05,
1.42) 20.21) 1,501.35) 27.19) 5.16)
in the treatment of status epilepticus in neonates and children is now well documented. 21-23 However, this drug shares the same respiratory effects as diazepam. Also, although few of our patients had received an adequate loading dose of the initial anticonvulsant, adequacy of dose did not appear to be associated with intubation. In this population, response lat e n c y was n o t s h o r t e r w h e n diazepam was used as the initial anticonvulsant instead of phenobarbital or phenytoin, and a significant association was found between diazepam use and i n t u b a t i o n . P r o s p e c t i v e studies are needed to d e t e r m i n e whether diazepam, particularly when given as the first anticonvulsant, offers any advantage over alternative drugs in initial treatment of seizures in children or is potentially harmful. CONCLUSION In a case-control study of patients with seizures who were transported to our institution from nonpediatric EDs, patients were more likely to require intubation if they had received diazepam as first or subsequent therapy than were those who had received phenobarbital, phenytoin, or both. Convulsion time (minutes) before the first dose of anticonvulsant and response latency did not differ between intubated and nonintubated patients or between those who received diazepam and those who did not. This study indicates that di20:9 September 1991
DIAZEPAM & INTUBATION Orr et al
azepam is more likely to be associated w i t h e n d o t r a c h e a l i n t u b a t i o n and is n o t s u p e r i o r to a l t e r n a t i v e drug therapy in controlling seizures. The authors thank Ms Lisa Cohn and Patr i c k M Kochanek, MD, for editorial assis-
6. Gastaut H, Naquet R, Poire R, et al: Treatment'of status epilepticus with diazepam (Valium). Epilepsia I965;6:167-182.
15. American Academy of Pediatrics Committee on Drugs: Emergency drug doses for infants and children. Pediatrics 1988;8i:462-465.
7. Carter S, Gold AP: The critically ill child: Management of status epilepticus. Pediatrics 1969;44:732-733.
16. Conover WJ: Practical Non-parametric Statistics, ed 2. New York, John Wiley & Sons, 1980.
8. Shaywitz BA: Management of acute neurologic synd r o m e s in i n f a n t s and children. Yale J Biol M e d 1984;57:83-95.
17. Hosmer DW Jr, Lemeshow 8: Applied Logistic Re gression. New York, John Wiley & Sons, 1989.
tance. They also thank Steven H Belle, PhD, for his statistical assistance.
9. Packer RJ, Berman PH: Neurologic emergencies, in Fleisher G, Ludwig S (eds): Textbook of Pediatric Emergency Medicine. Baltimore, Williams & Wilkins, 1983, p 333-335.
REFERENCES
10. Litchfield NB: Complications of intravenous diazepam: Respiratory depression (an a s s e s s m e n t of i6,000 cases). Anesth Prog 1981;29:11-17.
1. Delgado-Escueta AV, Wasterlain C, Treiman DM, et aI: Current concepts in neurology: Management of stares epilepticus. N Engl J Med 1982;306:1337-1340. 2. Lombroso CT: Treatment of status epilepticus with diazepam. Neurology (Minneapolis) 1966;16:629-634.
11. Hall 8C, Ovassapian A: Apnea after intravenous di azepam therapy, lAMA 1977;238:1052.
18. Glantz SA: How to summarize data, in Barry BK, White J (eds): Primer of Biostatistics, ed 2. New York, McGraw-Hill Book Co, 1987, p i0-29. 19. Goldberg MA, McIntyre HB: Barbiturates in the treatment of status epilepticus, in Delgado-Escueta AV, Wasterlain CG, Treiman DM, et al (eds): Advances in NeurologG New York, Raven Press, Publishers, 1983, vol 34, p 449-503. 20. Shaner DM, McCurdy 8A, Herring MO, et al: Treatment of status epilepticus: A prospective comparison of diazepam and phenytoin versus phenobarbital and optionai phenytoin. Neurology 1988;38:202-207.
3. McMorris S, McWilliam PKA: Status epilepticus in infants and young children treated with parenteral diazepam. Arch Dis Child 1969;44:604-611.
12. Larson GF, Hurlbert BJ, Wingard DW: Physostigmine reversal of diazepam-induced depression. Anesth Analg 1977;56:348-350.
4. Bailey DW, Fenichel GM: The treatment of prolonged seizure activity with intravenous diazepam. J Pediatr 1968;73:923-927.
13. Brauninger G, Ravin M: Respiratory arrest following intravenous Valium. A n n Ophthalmol 1974;6: 805-806.
22. Lacey D, Singer W, Horwitz S, et al: Lorazepam therapy of status epilepticus in children and adolescents. J Pediatr 1986;108:771-774.
5. Smith BT, Masotti RE: Intravenous diazepam in the treatment of prolonged seizure activity in neonates and infants. Dev Med Child Neural 1971;I3:630-634.
14. Aijian P, Tsai A, Knopp R, et al: Endotracheal intubation of pediatric patients by paramedics. Ann Emerg Med 1989;18:489-494.
23. D e s h m u k h A, W i t t e r t W, 8 c h n i t z l e r , et al: Lorazepam in the treatment of refractory neonatal seizures. A m J Dis Child 1986;140:1042-1044.
20:9 September 1991
Annals of Emergency Medicine
21. Levy R, Kroll RL: Treatment of status epilepticus with Iorazepam. Arch NeuroI 1984;41:605-611.
1013/111
Diazepam and Intubation in Emergency Treatment of Seizures in Children Study objectives: This study was undertaken to determine the incidence of endotracheal intubation after the use of diazepam compared with phenobarbital or phenytoin in emergency treatment of seizures in children. Design: The records of all children (98) were reviewed in a case-control fashion. A logistic regression model was used to determine whether there was an association between diazepam administration and intubation, adjusting for all other covariates (age, weight, convulsion time before first anticonvulsant was given, response latency, diagnosis, and therapy). Setting: All children were treated in an emergency department and then transported to a tertiary pediatric center by a pediatric transport team. Type of participants: All patients were children, w~th a median age of 2.7 years flange, 0.17 to 15.3 years). Interventions: None. Measurements and main results: Only the use of diazepam was found to be significantly associated with intubation after adjusting for all other covariates (adjusted odds ratio, 49.4; P < .001). In the comparison of diazepam versus phenobarbital or phenytoin used as the first anticonvuIsan& median response latency did not differ (27.0 vs 32.5 minutes, P > .83). Conclusion: A significant association was found between diazepam use and intubation. Response latency was not shorter when diazepam was used as the initial anticonvulsant compared with phenobarbital or phenytoin. [Orr RA, Dimand R J, Venkataraman ST, Karr VA, Kennedy KJ: Diazepam and intubation in emergency treatment of seizures in children. Ann Emerg Med September 1991;20:1009-1013.]
INTRODUCTION Children are often seen in the emergency departments of nonpediatric hospitals for the treatment of seizures before they are referred to a pediatric center. Diazepam has been recommended for the initial treatment of seizures 1-9 because of its presumed rapid onset of action compared with other antieonvulsants. The emergency physician, who is responsible for treating the acute onset of seizures, thus may choose diazepam as the initial drug. Apnea and respiratory depression are well-known side effects of diazepaml,6, lo-13 that may necessitate intubation for both airway protection and assisted ventilation. These conditions demand personnel who are able to recognize respiratory compromise and are skilled in the management of the pediatric airway. It has been shown that prehospital personnel have difficulty in airway management in children. 14 However, many emergency medical services crews carry diazepam for the management of seizures in children. The incidence of apnea and respiratory depression requiring intubation after the use of diazepam in this setting is not known. All patients who were transported to our institution by the interhospital transport team from other EDs and had a diagnosis of seizures requiring anticonvulsant therapy were studied in a case-control fashion. The goals of this study were to determine the incidence of endotracheal intubation in these patients, to identify factors associated with intubation, and to test the hypothesis that diazepam results in a greater incidence of intubation while offering no better control in stopping seizures than phenobarbital or phenytoin.
20:9 September 1991
Annals of Emergency Medicine
Richard A Orr, MD* Robert J Dimand, MD* Shekhar T Venkataraman, MD* Vaterie A Karrt Kathleen J KennedylPittsburgh, Pennsylvania From the Departments of Anesthesiology/ Critical Care Medicine and Pediatrics, University of Pittsburgh School of Medicine;* and the Transport Team, Department of Nursing, Children's Hospital ef Pittsburgh,t Pittsburgh, Pennsylvania. Received for publication May 23, 1989. Revisions received September 18, 1990, and February 14, 1991. Accepted for publication March 25, 1991. Presented in part at the University Association for Emergency Medicine Annual Meeting in Cincinnati, Ohio, May 1988. Address for reprints: Richard A Orr, MD, Department of Anesthesiology/Critical Care, University of Pittsburgh School of Medicine, Children's Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, Pennsylvania 15213.
1009/107
DIAZEPAM & INTUBATION Orr et al
lOO 80
0--0 e--I
~
,9
- - / e /
~i
60
°
e~
o,.~/o /
20
I
I
<10
10 29
I
I
30 49 50-69
I
I
70 119
>120
RespDnseLatency(rnin)
F I G U R E . Cumulative frequency of i n t u b a t e d versus n o n i n t u b a t e d patients for response latency.
MATERIALS A N D M E T H O D S The records of all children w i t h a h i s t o r y of seizures w h o were transported to our hospital by our transport t e a m between October 1986 and September 1987 were reviewed. Children were i n c l u d e d in the s t u d y if they had seizures w i t n e s s e d by personnel in the referring ED and t h e y received at least one anticonvulsant. Excluded were children who had seizures not w i t n e s s e d by the referring ED or w h o were i n t u b a t e d before a w i t n e s s e d seizure or a n t i c o n v u l s a n t therapy. For each child, the data recorded included age, weight, convulsion t i m e (minutes) before the first dose of anticonvulsant, elapsed t i m e from i n i t i a t i o n of t h e r a p y to cessat i o n of c l i n i c a l c o n v u l s i v e m o v e ments (response latency), treatment, p r e s e n c e or a b s e n c e of i n t u b a t i o n , and diagnosis. Diagnostic categories were defined as prior seizure disorder (included patients with a confirmed diagnosis of e p i l e p s y ) , s e i z u r e w i t h f e v e r (inc l u d e d p a t i e n t s w h o s e s e i z u r e was associated with fever but did not have evidence of central nervous syst e m [CNS] infection or a history of a s e i z u r e disorder), m e t a b o l i c c a u s e s (included patients whose seizure was r e l a t e d to a m e t a b o l i c d e r a n g e m e n t ( s u c h as h y p o g l y c e m i a or h y p o calcemia), CNS infection (included p a t i e n t s w h o s e s e i z u r e w a s associated w i t h a d o c u m e n t e d CNS infection such as m e n i n g i t i s or encephalitis), a n d i d i o p a t h i c or t u m o r (inc l u d e d p a t i e n t s w h o s e s e i z u r e was associated w i t h an intracranial lesion and those in w h o m a cause for the seizure was never determined). Diagnosis was assigned on review of the records. Children were often t r e a t e d w i t h an anticonvulsant in the ED before a 108/1010
TABLE l. Comparison of intubated ~ t h nonintubated patients
--//O
Inlubatedpafierls Nonntubatedpaden!s
No. of patients Age (yr) Age range (yr) Diagnosis Prior seizure disorder Seizure with fever Metabolic CNS infection Tumor Idiopathic
Convulsion time before anticonvulsant (min)
Intubated
Nol Intubated
(%)
(%)
32 4.5
22 1.3
0.5 - 15.3
0.2 - 10.0
5 (15.6) 10 (31.3) 2 (6.3) 5 (15.6) 1 (3.1) 9 (28.1)
6 5 4 3 0 4
< .01"
> .50t
1 - 180
1
30.0 145
> .25*
Range (min)
Response latency (rain)
40.0 1 360
15.0 1 150
> .30*
8 (25.0) 19 (59.4) 5 (15.6) 27 (84.4) 13 (40.6) 5 (15.6) 24 (75.0) 3 (9.4) 5 (15.6) 12 (37.5) 16 (50.0) 4 (12.5)
1 (4.5) 6 (27.3) 15 (68.2) 7 (31.8) 9 (40.9) 14 (63.6) 6 (27.3) 2 (9.1) 14 (63.6) 12 (54.5) 9 (40.9) 1 (4.5)
0.23 0.06 - 1.50
0.140 0.11 - 2.30
Range (min)
16.5
(27.3) (22.7) (18.2) (13.6) (0.0) (18.2)
P
Drug therapy Diazepam only Diazepam plus phenobarbital, phenytoin, or both No diazepam Ever used diazepam Ever used phenytoin Ever used phenobarbital First used diazepam First used phenytoin First used phenobarbital Used only one drug Used two drugs Used three drugs
Total dose of diazepam for patients receiving diazepam (mg/kg) Range (mg/kg)
< .001, < .0001§ > .90§ > .09§ < .002*
> .35t
> .50*
Results of continuous variables are expressed as median and range• *Mann-Whitney rank sum test; tFreeman Halton test; :x 2 test for independent samples; 9Fisher's exact test, two-tailed.
diagnosis was d e t e r m i n e d . N o n e of the " n o n i n t u b a t e d " children required b a g - v a l v e - m a s k v e n t i l a t i o n at a n y time. Adequacy of the initial dose of a n t i c o n v u l s a n t (mg/kg) was determ i n e d for each patient based on the r e c o m m e n d a t i o n of t h e A m e r i c a n A c a d e m y of Pediatrics C o m m i t t e e on Drugs. * 15 Statistical analyses were performed using t h e M a n n - W h i t n e y r a n k s u m test to compare c o n t i n u o u s variables between two groups and the Fisher's e x a c t test, F r e e m a n - H a l t o n test~ 16 and X2 for categorical variables where appropriate. P < .05 was considered significant. A univariate logistic regression *The minimum recommended dose for each anticonvulsant was considered an adequate initial dose: diazepam, 0.3 mg/ kg; phenobarbital, 10 mg/kg; phenytoin, 10 mg/kg. Annals of Emergency Medicine
m o d e l 17 was used to test the association b e t w e e n i n t u b a t i o n and each of the continuous v a r i a b l e s (age, weight, convulsion time, response latency, n u m b e r of drugs, and drug doses) as well as b e t w e e n i n t u b a t i o n and categorical variables (diagnosis and drug therapy). T h e u n a d j u s t e d o d d s r a t i o s a n d 95% c o n f i d e n c e l i m i t s were calculated for each indep e n d e n t variable separately. A stepwise logistic regression analysis t h e n was performed to fit the m o s t parsim o n i o u s model. A m u l t i v a r i a t e logistic regression model was used to determine w h e t h e r there was an association between diazepam administration.and i n t u b a t i o n , a d j u s t i n g for all o t h e r covariates, including adequacy of the initial dose of anticonvulsant. A statistically significant association w i t h i n t u b a t i o n was d e f i n e d as t h a t for w h i c h the 95% confidence i n t e r v a l 20:9 September 1991
DIAZEPAM & INTUBATION Orr et al
TABLE 2. Comparison of patients: Diazepam first versus phenobarbital or phenytoin first
Diazepam First (%) 30 4.5 0.25 15.0
No. of patients Age (yr) Age range (yr) Convulsion time before anticonvulsant (min) Range (min) Response latency (min) Range (rain) Drug therapy Diazepam only Diazepam plus phenobarbital, phenytoin, or both Ever used phenytoin Ever used phenobarbital Used only one drug Used two drugs Used three drugs Intubated
Phenobarbital or Phenytoin First (%) 24 1.5 0.17 15.3
27.0 1.0 180.0 27.0 1.0 360.0
20.0 1.0 145.0 32.5 1.0 - 255.0
9 (30.0)
0 (0.00)
2! (70.0) 13 (43.3) 12 (40.0) 9 (30.3) 17 (56.7) 4 (13.3) 24 (80)
4 (16.7) 9 (37.5) 20 (83.3) 15 (62.5) 8 (33.3) 1 (4.2) 8 (33.3)
P < .03*
> .62" > .83*
> .80t < .O02t < .05~ < .001t
Results of continuous variables are expressed as median and range. *Mann-Whitney rank sum lest; tFisher's exact test, two tared; cFreeman Halton test.
TABLE 3. Univariate logistic regression analyses - Continuous variables versus intubation
Age Weight Convulsion time before an0convulsant Response latency No. of drugs Total dose of diazepam
Odds Ratio 1.27 1.09
95% Confidence Intervals (1.04, 1.66) (t.00, 1.17)
0.99 1.01 1.89 2.77
(0.98, 1.01) (1.00, 1.02) (0.77, 4.63) (0.37, 20.69)
about the odds ratio did not contain 1.0.
RESULTS Because of the nonparametric nature of the data, results of continuous variables are expressed as median and range. 18 Ninety-eight patients were transported to our hospital by our team during the study period; 54 children met the inclusion criteria. Median age was 2.7 years (range, 0.17 to 15.3 years). Median c o n v u l s i o n time before the first dose of anticonvulsant was 20.0 minutes (range, 1.0 to 180.0 minutes), and median response latency was 30.0 m i n u t e s (range, 1.0 to 360.0 minutes). Intubated children were older (P < 20:9 September 1991
.01) and more likely to have received d i a z e p a m than were those in the group that did not receive diazepam (Table 1). Convulsion time before adm i n i s t r a t i o n of an anticonvulsant, response latency, number of anticonvulsants, and total dose of diazepam for patients receiving diazepam did not significantly differ between the two groups (Table 1 and Figure). The use of diazepam, either alone or in combination with other anticonvulsants, was found to be significantly associated with intubation (P < .001). Eighty-four percent of intubated patients had received diazepam compared with only 31.8% of nonintubated patients {P < .0001). Eighty percent of the patients who received diazepam as the first anticonvulsant required intubation (P < .001 (Table 2). They were also older than those in the group that received phenobarbital or phenytoin first (P < .03). Convulsion time (minutes) before administration of an anticonvulsant and response latency did not differ between these two groups. T h i r t y - n i n e of the 54 c h i l d r e n (72.2%) received an inadequate initial dose of anticonvulsant. Seven of 32 intubated children (21.9%) had an adequate dose for the first anticonvulsant administered compared with eight of 22 n o n i n t u b a t e d children (36.4%) (P > .24, Fisher's exact test). Annals of Emergency Medicine
T w e n t y - f i v e of t h e 39 c h i l d r e n (64.1%) whose initial dose was inadequate were intubated. Of these, 19 of 25 (76.0%) were initially given diazepam. Nine of the 14 nonintubated children (64.3%) whose initial dose of a n t i c o n v u l s a n t was inadequate were never given diazepam. Adequacy of dose did not contribute significantly to the odds of intubation (P > .24). With univariate logistic regression analysis used for continuous variables, statistically significant associations were found between intubation and both age and weight (Table 3). The unadjusted odds ratios relating age and weight to intubation were 1.27 (P < .01) and 1.09 (P < .05), respectively. Because age and weight are related (Spearman's correlation, .92), a logistic regression analysis shows that after adjusting for age, the association between weight and intubarton is no longer statistically significant (P > .7). Statistically significant associations were not found bet w e e n i n t u b a t i o n and c o n v u l s i o n time, response latency, n u m b e r of drugs, or dose of diazepam. For categorical variables, use of diazepam at any time (odds ratio, 11.57~ P < .001), alone (odds ratio, 24.0; P < .001), or in combination with phenobarbital, phenytoin, or both (odds ratio, 9.50; P < .0001) was significantly associated with intubation. Diagnosis was not significantly associated with intubation. In a stepwise logistic regression analysis, administration of diazepam entered the model on the first step (P < .001) as being significantly associated with intubation. After adjusting for age and weight, diazepam use still was found to be significantly associated w i t h i n t u b a t i o n (odds ratio, 8.12; P < .0001). Diagnosis was not significantly associated with intubation (P > .4) after adjustment for diazepam use. When using a multivariate logistic regression analysis, only the use of diazepam was found to be significantly associated with intubation after adjusting for confounding variables (adjusted odds ratio, 49.39; P < .001) (Table 4). DISCUSSION Since Gastaut et al first reported the use of parenteral diazepam in the treatment of status epilepticus, 6 this drug has received wide attention and 1Oll/109
DIAZEPAM & INTUBATION Orr et al
has been used in many institutions as the drug of choice for this purpose in both adults and children 1-9 despite its well-known complication of respiratory depression. 1,6,lO-13 Because seizures may be associated with neuronal cell damage and s e c o n d a r y metabolic consequences, DelgadoEscueta et al suggested that convulsions should be stopped as soon as possible. 1 Thus, one may choose to use a drug that offers more rapid control of seizures than other anticonvulsants despite its effect on the respiratory system, while being prep a r e d to c o n t r o l t h e a i r w a y . Intravenous diazepam has therefore been suggested as the drug of choice because of its presumed rapidity in control of clinical seizures. 1 Approximately 20% of all pediatric interhospital transports performed by our transport team involve patients with seizures, many of whom have respiratory compromise as indicated by the frequency of intubation. Sixtyfive percent of our study population had received diazepam for treatment of seizures; 55% had received it as the initial drug. There has been a lack of carefully controlled clinical studies comparing the various anticonvulsant agents for their effectiveness in treating status epilepticus in both adults and children. 19 In this study, diazepam, as initial or subsequent therapy, offered no better control of seizures (response latency) than phenobarbital, phenytoin, or both. Furthermore, patients who received diazepam were more likely to require intubation, even after adjustment for other potentially confounding variables. In a prospective study in adults, Shaner et al found that a combination of diazepam and phenytoin given as initial anticonvulsant therapy in status epilepticns did not prove to be superior to phenobarbital. 20 In fact, they found that both the cumulative convulsion time and the response latency were shorter for patients given phenobarbital as the initial anticonvulsant. This raises the issue of whether diazepam used as the first drug in treating seizures offers any advantage over phenobarbital. This case-control study involved a select group of patients who were transported to a pediatric center with witnessed seizures requiring therapy. Information on the number of patients treated for seizures who were 110/1012
TABLE 4. Multivariate logistic regression model: Association of diazepam with intubation, adjusting for other covariates Odds Ratio Age Weight Diagnosis (Baseline = Prior seizure disorder) Seizure with fever Metabolic cause CNS infection Idiopathic or tumor
1.62 0.96 1.00 46.50 2.07 3.34 11.78
Convulsion time before anticonvulsants Response latency Therapy (Ever versus never used) Phenytoin Phenobarbital Diazepam Total dose of diazepam Adequacy of dose
1.00 1.01 0.09 2.09 49.39 1.18 0.50
not referred and the timing of administration of diazepam in each case was difficult to obtain. Patients with prolonged seizures may be at greater risk for airway compromise in general, making them more likely to require intubation regardless of the anticonvulsant used. If that were so, the intubation rate should have been similar in patients with and without diazepam. It is also possible that diazepam was more likely to have been used in patients whose seizures were deemed more prolonged, in an attempt to gain rapid control, making these patients likely candidates for airway control. Arterial blood gases reflecting the acid-base status of these patients were rarely available in our population. Nevertheless, convulsion times before the first dose of anticonvulsant were similar for the patients who received diazepam and those who had not. Hypoxia and traumatic injuries to the airway may be more likely when intubation is performed in an uncontrolled setting or by an operator who is not trained in management of the pediatric airway. If an alternative ant i c o n v u l s a n t drug were k n o w n to have equal efficacy and a decreased incidence of respiratory compromise, it might be preferred to diazepam as first-line therapy in the control of seizures. A n o t h e r benzodiazepine, lorazepam, is four times more potent than diazepam in stopping status epilepticus in animals, and its 'efficacy Annals of Emergency Medicine
95% Confidence Intervals
(0.66, 4.01) (0.73, 1.25)
(0.80, (0.02, (0.09, (0.30,
2,709.28) 26922) 123.24) 461.21)
(0.96, 1.03) (0.99, 1.04) (0.01, (0.22, (1.62, (0.05, (0.05,
1.42) 20.21) 1,501.35) 27.19) 5.16)
in the treatment of status epilepticus in neonates and children is now well documented. 21-23 However, this drug shares the same respiratory effects as diazepam. Also, although few of our patients had received an adequate loading dose of the initial anticonvulsant, adequacy of dose did not appear to be associated with intubation. In this population, response lat e n c y was n o t s h o r t e r w h e n diazepam was used as the initial anticonvulsant instead of phenobarbital or phenytoin, and a significant association was found between diazepam use and i n t u b a t i o n . P r o s p e c t i v e studies are needed to d e t e r m i n e whether diazepam, particularly when given as the first anticonvulsant, offers any advantage over alternative drugs in initial treatment of seizures in children or is potentially harmful. CONCLUSION In a case-control study of patients with seizures who were transported to our institution from nonpediatric EDs, patients were more likely to require intubation if they had received diazepam as first or subsequent therapy than were those who had received phenobarbital, phenytoin, or both. Convulsion time (minutes) before the first dose of anticonvulsant and response latency did not differ between intubated and nonintubated patients or between those who received diazepam and those who did not. This study indicates that di20:9 September 1991
DIAZEPAM & INTUBATION Orr et al
azepam is more likely to be associated w i t h e n d o t r a c h e a l i n t u b a t i o n and is n o t s u p e r i o r to a l t e r n a t i v e drug therapy in controlling seizures. The authors thank Ms Lisa Cohn and Patr i c k M Kochanek, MD, for editorial assis-
6. Gastaut H, Naquet R, Poire R, et al: Treatment'of status epilepticus with diazepam (Valium). Epilepsia I965;6:167-182.
15. American Academy of Pediatrics Committee on Drugs: Emergency drug doses for infants and children. Pediatrics 1988;8i:462-465.
7. Carter S, Gold AP: The critically ill child: Management of status epilepticus. Pediatrics 1969;44:732-733.
16. Conover WJ: Practical Non-parametric Statistics, ed 2. New York, John Wiley & Sons, 1980.
8. Shaywitz BA: Management of acute neurologic synd r o m e s in i n f a n t s and children. Yale J Biol M e d 1984;57:83-95.
17. Hosmer DW Jr, Lemeshow 8: Applied Logistic Re gression. New York, John Wiley & Sons, 1989.
tance. They also thank Steven H Belle, PhD, for his statistical assistance.
9. Packer RJ, Berman PH: Neurologic emergencies, in Fleisher G, Ludwig S (eds): Textbook of Pediatric Emergency Medicine. Baltimore, Williams & Wilkins, 1983, p 333-335.
REFERENCES
10. Litchfield NB: Complications of intravenous diazepam: Respiratory depression (an a s s e s s m e n t of i6,000 cases). Anesth Prog 1981;29:11-17.
1. Delgado-Escueta AV, Wasterlain C, Treiman DM, et aI: Current concepts in neurology: Management of stares epilepticus. N Engl J Med 1982;306:1337-1340. 2. Lombroso CT: Treatment of status epilepticus with diazepam. Neurology (Minneapolis) 1966;16:629-634.
11. Hall 8C, Ovassapian A: Apnea after intravenous di azepam therapy, lAMA 1977;238:1052.
18. Glantz SA: How to summarize data, in Barry BK, White J (eds): Primer of Biostatistics, ed 2. New York, McGraw-Hill Book Co, 1987, p i0-29. 19. Goldberg MA, McIntyre HB: Barbiturates in the treatment of status epilepticus, in Delgado-Escueta AV, Wasterlain CG, Treiman DM, et al (eds): Advances in NeurologG New York, Raven Press, Publishers, 1983, vol 34, p 449-503. 20. Shaner DM, McCurdy 8A, Herring MO, et al: Treatment of status epilepticus: A prospective comparison of diazepam and phenytoin versus phenobarbital and optionai phenytoin. Neurology 1988;38:202-207.
3. McMorris S, McWilliam PKA: Status epilepticus in infants and young children treated with parenteral diazepam. Arch Dis Child 1969;44:604-611.
12. Larson GF, Hurlbert BJ, Wingard DW: Physostigmine reversal of diazepam-induced depression. Anesth Analg 1977;56:348-350.
4. Bailey DW, Fenichel GM: The treatment of prolonged seizure activity with intravenous diazepam. J Pediatr 1968;73:923-927.
13. Brauninger G, Ravin M: Respiratory arrest following intravenous Valium. A n n Ophthalmol 1974;6: 805-806.
22. Lacey D, Singer W, Horwitz S, et al: Lorazepam therapy of status epilepticus in children and adolescents. J Pediatr 1986;108:771-774.
5. Smith BT, Masotti RE: Intravenous diazepam in the treatment of prolonged seizure activity in neonates and infants. Dev Med Child Neural 1971;I3:630-634.
14. Aijian P, Tsai A, Knopp R, et al: Endotracheal intubation of pediatric patients by paramedics. Ann Emerg Med 1989;18:489-494.
23. D e s h m u k h A, W i t t e r t W, 8 c h n i t z l e r , et al: Lorazepam in the treatment of refractory neonatal seizures. A m J Dis Child 1986;140:1042-1044.
20:9 September 1991
Annals of Emergency Medicine
21. Levy R, Kroll RL: Treatment of status epilepticus with Iorazepam. Arch NeuroI 1984;41:605-611.
1013/111