Anticonvulsants for preventing mortality and morbidity in full term newborns with perinatal asphyxia

Evans DJ, Levene MI, Tsakmakis M

Background - Methods - Results - Characteristics of Included Studies - References - Data Tables and Graphs


Dates

Date edited: 08/06/2007
Date of last substantive update: 08/03/2007
Date of last minor update: / /
Date next stage expected 08/03/2009
Protocol first published: Issue 3, 1998
Review first published: Issue 3, 1998

Contact reviewer

Dr David J Evans
Consultant Neonatologist
Neonatal Intensive Care Unit
Southmead Hospital
Neonatal Intensive Care Unit
Southmead Hospital
Bristol
UK
BS10 5NB
Telephone 1: +44 117 959 5326
Telephone 2: +44 117 959 5325
Facsimile: +44 117 959 5324
E-mail: david.evans@nbt.nhs.uk

Contribution of reviewers

David Evans and Malcolm Levene conducted the original review in 1999 and update in 2001. For the most recent update, Maria Tsakmakis implemented the search and arranged for translation of papers. DE and MT appraised the new studies and DE updated the analyses and text of the review.

Internal sources of support

University of Leeds, UK
North Bristol NHS Trust, UK

External sources of support

None

What's new

This review updates the existing review "Anticonvulsants for preventing mortality and morbidity in full term newborns with perinatal asphyxia", first published in The Cochrane Library, Issue 3, 1998 and updated in Issue 2, 2001 (Evans 1998, Evans 2001).

Our search was updated in January 2007.

Two additional studies have been incorporated in the review since 2001.

Dates

Date review re-formatted: 20/09/1999
Date new studies sought but none found: / /
Date new studies found but not yet included/excluded: / /
Date new studies found and included/excluded: / /
Date reviewers' conclusions section amended: / /
Date comment/criticism added: / /
Date response to comment/criticisms added: / /

Text of review

Synopsis


It is unclear whether giving anticonvulsants to newborn babies soon after possible birth asphyxia at term is safe and effective. More studies are needed.

Seizures (or convulsions) are common following birth asphyxia. These seizures may worsen the brain injury. In theory, anticonvulsant medication given to babies soon after possible birth asphyxia may improve outcome by preventing seizures and protecting the brain. Anticonvulsant drugs are not without side effects and there are concerns that they might impair brain development. The studies included in this review involved relatively small numbers of babies and few studies assessed developmental outcome. At present, there is insufficient information on which to base recommendations about the effectiveness of giving anticonvulsants to newborn babies soon after possible birth asphyxia.

Abstract



Background


Seizures are common following perinatal asphyxia and may exacerbate secondary neuronal injury by increasing cerebral metabolic demand, causing fluctuations in oxygenation and perfusion, and triggering the release of excitatory neurotransmitters. Anticonvulsant therapy has been used in infants with perinatal asphyxia in order to prevent seizures. However, long term anticonvulsant therapy may lead to inhibition of brain development. Therefore, the routine use of anticonvulsant therapy to prevent seizures following perinatal asphyxia needs to be evaluated.

Objectives


To assess the effect of administering anticonvulsants to infants of 37 weeks gestation or more following perinatal asphyxia on death or subsequent severe neurodevelopmental disability and/or the prevention of seizures.

Search strategy


Relevant randomised controlled trials were identified using a combination of electronic database searches, hand searches and a search of the Cochrane Controlled Trials Registry.

Selection criteria


All randomised or quasi-randomised controlled clinical trials that reported data comparing the following outcomes: mortality, neurodevelopmental disability, neonatal seizures and adverse events, following anticonvulsant therapy in term infants (37 weeks or more) compared to controls (with or without placebo) following perinatal asphyxia.

Data collection & analysis


Methodological quality and validity of studies were assessed without consideration of the results. Data relevant to the outcome were extracted and analysed.

Main results


Seven randomised or quasi-randomised controlled trials that met the selection criteria were included. No studies were of sufficient methodological quality and size to demonstrate a valid, clinically significant change in the risk of mortality or severe neurodevelopmental disability. A meta-analysis combining five studies comparing barbiturates with conventional therapy following perinatal asphyxia demonstrated no difference in risks of death, severe neurodevelopmental disability, or the combined outcome of death or severe neurodevelopmental disability.

Reviewers' conclusions


At the present time, anticonvulsant therapy to term infants in the immediate period following perinatal asphyxia cannot be recommended for routine clinical practice, other than in the treatment of prolonged or frequent clinical seizures. Any future studies should be of sufficient size to have the power to detect clinically important reductions in mortality and severe neurodevelopmental disability.

Background


Throughout the world, perinatal asphyxia remains an important cause of perinatally acquired brain injury in full term infants. In technically developed countries, the incidence of death or severe neurological impairment following perinatal asphyxia is 0.5 - 1.0 per 1000 live births (Finer 1981; Levene 1985; Thornberg 1995). In developing countries, perinatal asphyxia appears to be more common (al-Alfry 1990; Airede 1991; Boo 1991; Kinoti 1993; Singh 1991). Although follow up programmes are not well developed in these countries, it is likely that perinatal asphyxia produces a huge burden of world-wide disability.

During perinatal asphyxia, hypoxia and ischaemia cause primary neuronal injury because of cell necrosis (Hossman 1983). Neonatal resuscitation results in improved oxygenation and reperfusion, which lead to delayed, or secondary, neuronal injury. The mechanisms believed to be important in this secondary phase of neuronal injury include oxygen free radical production (McCord 1985), intracellular calcium entry (Siesjo 1992) and apoptosis (Buttke 1994).

Seizures are a common feature of Hypoxic-Ischaemic Encephalopathy (HIE) (Sarnat 1976). Seizures can substantially increase central nervous system (CNS) metabolic demand (Younkin 1986), cause the release of excitatory neurotransmitters such as glutamate (McDonald 1990), lead to fluctuations in systemic arterial pressure (Clozel 1985) and result in hypoxia and hypercapnea. Therefore, seizures themselves may cause further neuronal injury following asphyxia.

The potential benefits of preventing further neuronal injury associated with seizures following asphyxia has prompted the widespread use of anticonvulsants, and barbiturates in particular, for the prevention of seizures. In addition to their anticonvulsant activity, barbiturates are known to decrease CNS metabolic rate when given in high doses (Nilsson 1971), reduce calcium entry post-ischaemia and scavenge free radicals (Demopoulos 1977). These actions may potentially attenuate the cascade of damaging processes initiated following hypoxic-ischaemic insults, thereby reducing secondary neuronal injury. For these reasons, barbiturates have been used prophylactically (whether or not the infant has seizures) following perinatal asphyxia in neonates, although long-term therapy in animals leads to inhibition of brain growth and development.

Objectives


To determine the effect of administering anticonvulsant therapy on death or subsequent severe neurodevelopmental disability and/or the prevention of seizures in infants 37 weeks gestation or more following perinatal asphyxia.

Criteria for considering studies for this review



Types of studies


All published, unpublished and ongoing randomised or quasi-randomised trials with reported data comparing the specified outcomes in asphyxiated term infants given anticonvulsants with outcomes in controls with or without placebo.

Types of participants


Neonates, of 37 or more completed weeks of gestation, following perinatal asphyxia. Perinatal asphyxia was considered a clinical diagnosis, characterised by signs of fetal distress, depression at birth, neonatal encephalopathy or other signs of multi-organ dysfunction. The presence or absence of seizures before trial entry was not a required inclusion criterion.

Types of interventions


Anticonvulsants administered in the early neonatal period (within the first seven days of life) with the intention of preventing neuronal injury following perinatal asphyxia or the intention of preventing seizures.

Types of outcome measures


Primary outcome: death or severe neurodevelopmental disability assessed at greater than, or equal to, 12 months of age. Severe neurodevelopmental disability was defined as any one or combination of the following: cerebral palsy, developmental delay (DQ < 70), blindness.

Secondary outcomes: seizure control in the neonatal period; childhood epilepsy; hypotensive episodes following administration of anticonvulsant; reported adverse events. Seizures were either apparent clinically or detected by electro-encephalographic recordings.

Search strategy for identification of studies


See Neonatal Review Group search strategy. The specific search strategy for MEDLINE and the Cochrane Controlled Trials Registry as given below. The abstracts of the Society for Pediatric Research and European Society for Pediatric Research, published in the Journal of Pediatric Research were searched from January 1980 to January 2007. Cited references from retrieved articles were searched for additional studies. Abstracts and letters to the editor were reviewed to identify randomised controlled trials which had not been published. Editorials, indicating expert opinion, were reviewed to identify any further studies, not already included in the review.

1. MEDLINE Search

Dates: 1966 - Nov 2006

Strategy

#1 explode 'Asphyxia-Neonatorum' / all subheadings
#2 asphyx$
#3 hypox$ ADJ ischaemi$
#4 hypox$ ADJ ischemi$
#5 encephalopath$
#6 #1 or #2 or #3 or #4 or #5
#7 explode 'Anticonvulsants-' / all subheadings
#8 anticonvuls$
#9 explode 'Seizures-' / all subheadings
#10 convulsi$
#11 seiz$
#12 #7 or #8 or #9 or #10 or #11
#13 #6 and #12

This output was combined with the search filter for randomised controlled trials (see Appendix 5c of the Cochrane Handbook).

2. Cochrane Controlled Trials Registry Search

Date: Issue 3, 2006

Strategy (Update Software)

#1 ASPHYXIA-NEONATORUM*:ME
#2 ASPHYX*
#3 HYPOX*
#4 ENCEPH*
#5 (((#1 or #2) or #3) or #4)
#6 ANTICONVULSANTS*:ME
#7 SEIZURES*:ME
#8 CONVULSI*
#9 SEIZ*
#10 ANTICONVULS*
#11 ((((#6 or #7) or #8) or #9) or #10)
#12 (#5 and #11)

Methods of the review


The primary reviewer screened the title and abstract of studies identified by the above search strategy. Both primary and a secondary reviewer re-screened the full text of the report of each study identified as of potential relevance. Studies meeting any of the pre-specified inclusion criteria were included. Studies were evaluated for inclusion and methodological quality without consideration of results. Reviewers were not blinded to authorship, institution or journal.

The methodological quality of the various components of the study design known to be important in minimising bias were scored separately by reviewers before a consensus was obtained. The components were those relating to the following sources of potential bias, namely: selection, performance, attrition and detection.

Selection bias (Allocation concealment - Section 6 of the Cochrane Handbook):
A = adequate, B = can't tell / unclear, C = not used / inadequate.

Performance bias (blinding of caretakers to the intervention):
A = adequate, B = can't tell / unclear, C = not used / inadequate.

Attrition bias (post-randomisation exclusions and loss to follow-up):
A: < 5% loss, B: 5 - 9.9% loss, C: 10 - 20% loss, D: > 20% loss or unable to calculate because of lack of data.

Detection bias (blinding of outcome assessment):
A: Double blind, neither investigator nor participant (parent or guardian) knew, or was likely to correctly identify, treatment allocation.
B: Single blind, either investigator or participant knew, or was likely to correctly identify, treatment allocation.
C: No blinding. Includes studies where effects, or side effects, of treatment meant that it was likely that the treatment allocation could be correctly identified in a significant proportion (>20%) of participants.
D: Can't tell.

The treatment effects of individual trials were examined by comparing groups allocated to the treatment under study (prophylactic anticonvulsant therapy versus placebo or standard treatment, that is anticonvulsants for the treatment of seizures). Data relating to the primary and secondary outcomes (described above) were compared. Where relevant and if possible, relative risk (RR) and risk difference (RD) were calculated for dichotomous data and weighted mean difference (WMD) for continuous data, with 95% confidence intervals (CI) for all analyses. Analysis of outcome data was by intention to treat. Where relevant, meta-analyses was performed using the fixed effects "assumption free" model. Where relevant and if possible, heterogeneity between trial results was examined using the I-squared test for dichotomous outcomes.

Subgroup analyses were performed according to the types of anticonvulsants administered. Subgroup analyses according to (a) the grade of Hypoxic-Ischaemic Encephalopathy and (b) the presence or absence of seizures were only be performed where the data allowed grouping according to characteristics measured before trial entry (as both grade of Hypoxic-Ischaemic Encephalopathy and seizure rates are likely to be affected by anticonvulsant therapy).

Description of studies


Nine randomised or quasi-randomised controlled trials using neonatal anticonvulsive therapy following perinatal asphyxia were identified.

Two of these studies were excluded. Ruth 1988 randomised very low birthweight infants to receive phenobarbital or placebo following birth; perinatal asphyxia was not an eligibility criterion. Wilkinson 1989 compared four anticonvulsants in a randomised control trial. The anticonvulsants were used for treatment (not prophylaxis) of electro-encephalographically apparent seizures of any aetiology, not exclusively perinatal asphyxia. Infants were a mixture of term and premature neonates. It was not possible to extract the data relating only to term infants following asphyxia.

The characteristics of participants, interventions and outcomes of the remaining seven studies are given in the table: Characteristics of Included Studies. Five studies compared a barbiturate with conventional therapy (Goldberg 1986; Hall 1998; Ruth 1991; Singh 2004; Vargas-Origel 2004), Vela 1987 compared phenobarbital with phenytoin and Kuzemko 1972 compared chloral hydrate with diazepam. The clinical definition of perinatal asphyxia was vague in Kuzemko 1972. Three studies recorded neurodevelopmental outcome beyond one year of age: Goldberg 1986; Hall 1998 and Ruth 1991. The remaining studies only reported short-term neonatal outcomes (deaths, abnormal neurological status and seizures within the neonatal period): Kuzemko 1972; Vela 1987; Singh 2004 and Vargas-Origel 2004.

Methodological quality of included studies



Selection

All studies used formal randomisation, although the method was only stated in Hall 1998. Only Hall 1998 used adequate allocation concealment. No attempt at concealment was made in Kuzemko 1972 and the method of concealment was unclear in Goldberg 1986; Vela 1987; Ruth 1991; Singh 2004 and Vargas-Origel 2004 .

In the studies by Kuzemko 1972 and Singh 2004, randomisation resulted in unequal numbers of infants being allocated to the two groups (17 allocated to chloral hydrate versus 11 allocated to diazepam treatment and 25 allocated to phenobarbital versus 20 allocated to control, respectively.). No adequate explanation for this imbalance is evident.

Performance

No study used a placebo. Only in the studies by Vela 1987 and Kuzemko 1972 were the neonatal clinicians, responsible for the care and assessment of neonatal outcomes, blind to study group.

Potential bias arising from co-intervention remains a possibility in the five studies that did not use caretaker blinding: Goldberg 1986; Hall 1998; Ruth 1991; Singh 2004 and Vargas-Origel 2004. In Goldberg 1986, 14 infants in the group treated with thiopental and 12 infants in the control group received phenobarbital for the treatment of clinical seizures. Hall 1998 compared phenobarbital with conventional treatment for the prevention of seizures following asphyxia. The control group received a mean of 27 mg/kg of phenobarbital as treatment for seizures, compared to a mean of 39 mg/kg in the experimental group. The incidence of seizures was greater in the control group, compared to the phenobarbital group, and this may have led to the control group requiring more anticonvulsants for treatment (a downstream treatment effect). The clinicians caring for the infants were aware of treatment allocation, however, and may have had a lower threshold to diagnose and treat clinical seizures in the control group (co-intervention bias). In Vargas-Origel 2004, 3 out of 36 infants in the control group received phenobarbital for the treatment of seizures: additional phenobarbital was needed in 2 out of 37 infants in the experimental group. In Ruth 1991 and Singh 2004, the numbers of infants treated with additional anticonvulsants were not stated.

Attrition

There was a post-randomisation loss of 23% in the study by Hall 1998, 13% in Kuzemko 1972, 3% in Goldberg 1986 and 0% in Vela 1987; Singh 2004 and Vargas-Origel 2004. There were no data to calculate attrition in Ruth 1991.

Detection

Only three studies had blind assessment of outcome; only one of these (Goldberg 1986) assessed outcomes outside the neonatal period. Vela 1987 utilised a double-blind study design. Kuzemko 1972 used an independent clinician to administer the medications and the neonatal staff responsible for assessment were blind to study group allocation. Goldberg 1986 used a blinded outcome assessment at 1-3 years but the high incidence of hypotension in the experimental group may have led to the shorter term outcomes, such as clinical seizures, being recorded without the clinician being blind.

Results


None of the studies reported data to enable analyses according to subgroups other than type of anticonvulsant.

BARBITURATES VS. CONTROL (Comparison 01):

Death (before neurodevelopmental assessment by 3 years) (Outcome 01.01):

There was no significant difference in mortality rates between experimental and control groups reported in the five studies comparing barbiturates with standard therapy (Goldberg 1986; Hall 1998; Ruth 1991; Singh 2004; Vargas-Origel 2004). The meta-analysis, combining results from these studies, also showed no significant difference in mortality rates (typical RR 1.13, 95% CI 0.59, 2.17).

Severe neurodevelopmental disability in survivors examined (Outcome 01.02):

There was no significant difference in the rates of severe neurodevelopmental disability between experimental and control groups reported in the three studies comparing barbiturates with standard therapy and reporting this outcome (Goldberg 1986; Ruth 1991; Hall 1998). The meta-analysis, combining results from these studies, also showed no significant difference in rates of severe neurodevelopmental disability (typical RR 0.61, 95% CI 0.30, 1.22).

Death or severe neurodevelopmental disability (Outcome 01.03):

Only one study demonstrated a significant reduction in the relative risk of the combined outcome of severe neurodevelopmental disability or death in infants treated with phenobarbital (RR = 0.30, 95% CI 0.10 to 0.93; Hall 1998), although the significant post-randomisation loss, potential for co-intervention bias and unblinded outcome assessment raise concerns about the validity of these findings. The other two studies that reported both death and neurodevelopmental outcomes showed no difference in the risk of the combined outcome of severe neurodevelopmental disability or death, when comparing barbiturates with standard therapy (Goldberg 1986 and Ruth 1991). The meta-analysis, combining results from these three studies, also showed no significant difference in the risk of the combined outcome of death or severe neurodevelopmental disability (typical RR 0.78, 95% CI 0.49, 1.23).

PHENOBARBITAL VS. CONTROL (Comparison 02):

Seizures within neonatal period (Outcome 02.01):

Two studies compared Phenobarbital versus control and reported the rates of seizures within the neonatal period (Hall 1998; Vargas-Origel 2004); neither reported any significant difference in rates. The seizures rates were relatively low in both groups in the study by Vargas-Origel 2004, reflecting the fact that only a small proportion of infants enrolled had moderate or severe HIE.

IQ in survivors at 6 years of age (Outcome 02.02):

There was no significant difference in the IQ of survivors at six years of age, treated with phenobarbital or control as neonates (Ruth 1991).

Hypotension requiring inotropes (Outcome 02.03):

Singh 2004 stated that side effects of phenobarbital (hypotension, respiratory depression or excessive sleepiness) were not seen during the study. Vargas-Origel 2004 reported arterial pressures two hours after trial entry and there was no significant difference in values between those infants given prophylactic phenobarbital and those in the control group.

THIOPENTONE VS. CONTROL (Comparison 03):

Goldberg 1986 compared Thiopentone versus control and reported the proportions of infants suffering from seizures within three days of age and at three days of age.

Seizures within 3 days of age (Outcome 03.01):

No significant difference reported (Goldberg 1986).

Seizures at 3 days of age (Outcome 03.02):

No significant difference reported (Goldberg 1986).

Hypotension requiring inotropes (Outcome 03.03):

Goldberg 1986 reported an increase in the relative risk of neonatal hypotension, requiring inotropic support, associated with thiopental therapy, although this was not significant (RR = 1.76, 95% CI 0.98 to 3.16).

PHENOBARBITAL VS. PHENYTOIN (Comparison 04):

Seizures within the first 7 days of age (Outcome 04.01):

Vela 1987 reported similar rates of seizures occurring within the first week, when comparing Phenobarbital with Phenytoin.

CHLORAL HYDRATE VS. DIAZEPAM (Comparison 05):

Seizures beyond 3 days of age (Outcome 05.01):

Kuzemko 1972 did not find any significant difference in the proportions of infants with seizures occurring beyond three days of age, in infants treated with either Chloral Hydrate or Diazepam after birth.

Discussion


Perinatal asphyxia remains an important cause of death and neurodevelopmental disability. In this review, seven randomised controlled trials of anticonvulsants for preventing mortality and morbidity in full term infants with perinatal asphyxia were identified. It is disappointing that all the studies reporting mortality and neurodevelopmental disability are small in number (n < 40) and only one study assessed neurodevelopment blind to allocated intervention. Currently, there is little evidence from these studies to suggest that treatment with anticonvulsants following perinatal asphyxia in term infants leads to improvement in these outcomes.

Asphyxia also leads to seizure activity, which, in turn, may further compromise neuronal recovery. Seizures within the neonatal period is an outcome reported by most studies in this review. There is little evidence that treatment with anticonvulsants soon after an episode of perinatal asphyxia significantly reduces the subsequent seizure burden. It also remains to be established whether prevention of such seizure burden would represent a significant clinical achievement, i.e. whether reducing seizure activity will be translated into improvements in rates of death or severe neurodevelopmental disability.

No uniform definition of perinatal asphyxia was used between studies. This is because, in the absence of a valid and practical single marker of an hypoxic-ischaemic insult (with the potential to cause significant cerebral injury), the definition of perinatal asphyxia will remain clinical. Studies should strive towards a common clinical definition and endeavour to collect data relating to the markers in the neonatal period known at present to be associated with a poor neurodevelopmental outcome (e.g. magnetic resonance imaging, EEG data, etc.). It is also important for future research to identify other early and reliable markers associated with greatest risk of asphyxial cerebral injury, in order that potential neuroprotective strategies, commenced soon after birth, can be assessed in future studies.

Reviewers' conclusions



Implications for practice


At the present time, anticonvulsant therapy to term infants in the immediate period following perinatal asphyxia cannot be recommended for routine clinical practice, other than in the treatment of prolonged or frequent clinical seizures.

Implications for research


Any future studies should be of high quality: randomised control trials with allocation, performance and outcome assessment blinding. Such studies should be of sufficient size, with minimal attrition, to have the power to detect clinically important reductions in mortality and severe neurodevelopmental disability, as the primary outcome measures. Potential adverse effects, such as hypotension following barbiturate therapy, should be prospectively defined and reported. The relatively low incidence of perinatal asphyxia is such that there is a need for collaborative effort in order to mount such future studies.

Acknowledgements


The authors wish to acknowledge the help given by Dr Eduardo Moya (for translation of Vela 1987) and Dr Maria Juarez (for translation of Vargas-Origel 2004).

Potential conflict of interest


None


Characteristics of included studies

StudyMethodsParticipantsInterventionsOutcomesNotesAllocation concealment
Goldberg 1986RCT: assignment "using list of random numbers" Controlled trial. No placebo.
Sources of potential bias:
Selection: B;
Performance: C;
Attrition: B, 2 patients (1 patient each group) lost to neurodevelopmental follow up (6%);
Detection: C (neonatal outcomes), B (assessment at 1-3y).
Term infants with severe perinatal asphyxia. Criteria: signs of hypoxic-ischaemic encephalopathy and requiring mechanical ventilation within first hour of life; plus two of the following: perinatal distress (abnormal fetal heart pattern or requiring prolonged neonatal resuscitation); Apgar score below 5 at 5 minutes; base deficit greater than 16 mEq/l within first hour of life.
At trial entry: n=17 (experimental); n=15 (control).
Experimental: Thiopental 15 mg/kg IV over 30 minutes, then 10 mg/kg/h for 90 minutes, 5 mg/kg/h for 60 minutes, 3 mg/kg/h for 8 hours, 1.5 mg/kg/h for 6 hours and 0.75 mg/kg/h for 6 hours.
Control: conventional therapy (phenobarbitone or phenytoin for clinically evident seizures).
Hypotension requiring inotropes. Hypotension defined by attending clinician, based upon normative data.
Clinically apparent seizure activity over the first three days of life.
Death before neurodevelopmental assessment.
Neurodevelopmental assessment at 1-3 years of age (Bayley). Severe neurodevelopmental defined as Bayley <68.
Co-intervention: 14 experimental and 12 control infants received phenobarbitone for seizures.
B
Hall 1998RCT. Sealed envelope randomisation. Controlled trial. No blinding. No placebo.
Sources of potential bias:
Selection: A;
Performance: C;
Attrition: D, 9 patients (23% post randomisation loss): 5 experimental, 4 control;
Detection: D.
Term infants with severe perinatal asphyxia. Asphyxia defined by one of following criteria: initial arterial pH below 7.0 with base deficit above 15 mEq/l; Apgar score less than 4 at 5 minutes; no spontaneous respiratory effort before 10 minutes of life.
At trial entry: n=20 (experimental); n=20 (control).
Experimental: Phenobarbitone 40 mg/kg, IV over 60 minutes, administered after trial entry.
Control: Standard therapy (phenobarbitone given in order to treat clinical seizures).
Number of infants with clinically evident seizures.
Death before neurodevelopmental assessment.
Neurodevelopmental assessment at 3 years of age (Gessell, Bayley, or Stanford-Binet). Appropriate definitions of severe neurodevelopmental disability.
Co-intervention: control group received mean of 27 mg/kg of phenobarbitone in first 24 hours, compared to a mean of 39 mg/kg in the experimental group.
A
Kuzemko 1972RCT: Randomised. Prescription and administration of medication independent of assessment.
Sources of potential bias:
Selection: C;
Performance: A;
Attrition: C, 4 infants (13% post randomisation loss), 2 patients each group.
Detection: A.
Term infants with cerebral irritability. Vague case definition.
At trial entry: n=17 (Chloral Hydrate); n=11 (Diazepam).
Chloral Hydrate 80 mg 6 hourly for 24-72 hours.
Diazepam 1 mg 6 hourly for 24-72 hours.
Clinical seizures persisting beyond 3 days of life.Short-term (neonatal) outcomes.
C
Ruth 1991RCT: unclear methods of randomisation, allocation concealment, blinding and use of placebo.
Sources of potential bias:
Selection: B;
Performance: B;
Attrition: D (no data);
Detection: D.
Term infants with severe asphyxia (Apgar score less than 4 at 5 minutes or requiring mechanical ventilation beyond 30 minutes of life).
At trial entry: n=21 (experimental); n=17 (control).
Experimental: Phenobarbitone 30 mg/kg IV before 4 hours of age, a further 15 mg/kg 4 hours following first dose, followed by 5 mg/kg/day for 5 days. Control: conventional therapy.Death before neurodevelopmental assessment.
Neurodevelopmental assessment: Disability stated as cerebral palsy.
Cognitive assessment at 6 years of age (WISC-r), expressed as IQ.
Co-intervention: not stated.
B
Singh 2004RCT: assigned by computer-generated random numbers. No blinding. No placebo.
Sources of potential bias:
Selection: B;
Performance: C;
Attrition: A (nil);
Detection: C.
Infants >33 weeks with Apgar <6 at 1 min plus fetal distress (fetal bradycardia or meconium-stained liquor or cord arterial pH<7.16) plus abnormality of tone and conscious level within 6 hours.
At trial entry: n=25 (experimental); n=20 (control).
Experimental: Phenobarbital 20 mg/kg IV within 6 hours of life. Control: standard therapy (no placebo).Death, neurologically abnormal at discharge, CSF levels of malondialdehyde, superoxide dismutase, glutathione peroxidase and plasma levels of vitamins A and E.Although the participants included infants below 37 weeks gestation, the study was included in this review as the majority of infants were near-term (>35 weeks). Co-intervention: not stated. The study reported mainly short term (neonatal) outcomes; only death was considered as an outcome for this review.B
Vargas-Origel 2004RCT. Unstated method of randomisation. Controlled trial. No blinding. No placebo.
Sources of potential bias:
Selection: B;
Performance: C;
Attrition: A (nil);
Detection: C.
Infants >=37 weeks with one of the following: (a) umbilical arterial pH<=7.00 within 15min of life, (b) Base Deficit >=15mmol/l, (c) Apgar <3 at 5min, (d) no spontaneous respiration within first 10min. At trial entry: n=37 (Phenobarbital); n=36 (control).Experimental: Phenobarbital 40 mg/kg, IV over 60 minutes, administered within first hour of age.
Control: Standard therapy (phenobarbital given in order to treat clinical seizures).
Death, Sarnat grade of HIE, number of infants with clinical seizures.Short-term (neonatal) outcomes. Co-intervention: 5/36 infants in control group given Phenobarbital (mean dose 26mg/kg, mean time 9.2 hours. Experimental group given Phenobarbital 40mg/kg (mean time 0.88 hours). Infants in the trial had a relatively low rate of HIE grade II or III (12% overall: 11% experimental, 14% control).B
Vela 1987RCT: randomisation procedure not clear, double-blind, controlled study.
Sources of potential bias:
Selection: B;
Performance: A;
Attrition: A (nil);
Detection: A.
Term infants with perinatal asphyxia, defined as three of the following criteria: multiple late fetal heart rate decelerations, fetal bradycardia for >20 minutes, meconium-stained liquor, Apgar score of less than 8 at 5 minutes, requiring mechanical ventilation, requiring IV bicarbonate therapy.
At trial entry: n=9 (Phenobarbitone); n=8 (Phenytoin).
Phenobarbitone 12mg/kg IM, followed by 6 mg/kg/day for 7 days.
Phenytoin 12mg/kg IM, followed by 6 mg/kg/day for 7 days.
Clinical seizures within the first 7 days.Short-term (neonatal) outcomes.B

Characteristics of excluded studies

StudyReason for exclusion
Ruth 1988Eligibility criteria did not include evidence of perinatal asphyxia
Infants < 1500g eligible and randomised to phenobarbitol or placebo for prevention of potential complications of 'unrecognised' asphyxia.
Wilkinson 1989Anticonvulsants were used for treatment (not prevention) of electro-encephalographically apparent seizures of any aetiology, not exclusively perinatal asphyxia. Infants were a mixture of term and premature neonates. It was not possible to extract the data relating only to term infants following asphyxia.

References to studies

References to included studies

Goldberg 1986 {published data only}

Goldberg RN, Moscoso P, Bauer CR, Bloom FL, Curless RG, Burke B, et al. Use of barbiturate therapy in severe perinatal asphyxia: a randomized controlled trial. Journal of Pediatrics 1986;109:851-6.

Hall 1998 {published data only}

Hall RT, Hall FK, Daily DK. High-dose phenobarbital therapy in term newborn infants with severe perinatal asphyxia: A randomized, prospective study with three-year follow-up. Journal of Pediatrics 1998;132:345-8.

Kuzemko 1972 {published data only}

Kuzemko JA, Hartley S. Treatment of cerebral irritation in the newborn: double-blind trial with chloral hydrate and diazepam. Developmental Medicine and Child Neurology 1972;14:740-6.

Ruth 1991 {published data only}

Ruth V, Korkman M, Liikanen A, Paetau R. High-dose phenobarbitol treatment to prevent postasphyxial brain damage: A 6-year follow up [abstract]. Pediatric Research 1991;30:638 (abstract).

Singh 2004 {published data only}

Singh D, Kumar P, Majumdar S, Narang A. Effect of phenobarbital on free radicals in neonates with hypoxic ischemic encephalopathy - a randomized controlled trial. Journal of Perinatal Medicine 2004;32:278-81.

Vargas-Origel 2004 {published data only}

Vargas-Origel A, Espinosa-Garcia JOG, Muniz-Quezada E, Vargas-Nieto MA, Aguilar-Garcia G. Prevencion de la encefalopatia hipoxico-isquemica mediante Fenobarbital en forma temprana y a dosis alta. [Prevention of hypoxic-ischemic encephalopathy with high-dose, early phenobarbital therapy.]. Gaceta Medica de Mexico 2004;140:147-53.

Vela 1987 {published data only}

Vela F, Duran JA, Chunga F, Serrano JS, Valls A. Tratamiento profilactico de convulsiones en asfixia perinatal [Spanish] [Preventive treatment of convulsions in perinatal asphyxia]. Anales Espanoles de Pediatria 1987;27:95-9.

References to excluded studies

Ruth 1988 {published data only}

Ruth V, Virkola K, Paetau R, Raivio KO. Early high-dose phenobarbital treatment for prevention of hypoxic-ischemic brain damage in very low birth weight infants. Journal of Pediatrics 1988;112:81-6.

Wilkinson 1989 {published data only}

Rochefort MJ, Wilkinson AR. The safety and efficacy of alternative anticonvulsant regimes to control newborns seizures [abstract]. Early Human Development 1989;19:218.

Wilkinson AR, Rochefort MJ. Phenytoin reduces frequency and duration of neonatal seizures in the newborn: A randomised trial of four anticonvulsants [abstract]. Pediatric Research 1989;26:522.

* indicates the primary reference for the study

Other references

Additional references

Airede 1991

Airede AI. Birth asphyxia and hypoxic-ischaemic encephalopathy: incidence and severity. Annals Tropical Paediatrics 1991;11:331-5.

al-Alfry 1990

al-Alfry A, Carroll JE, Devarajan LV, Moussa MA. Term infant asphyxia in Kuwait. Annals of Tropical Paediatrics 1990;10:355-61.

Boo 1991

Boo NY, Lye MS. Factors associated with clinically significant perinatal asphyxia in the Malaysian neonates: a case-control study. Journal of Tropical Pediatrics 1991;38:284-9.

Buttke 1994

Buttke TM, Sandstrom PA. Oxidative stress as a mediator of apoptosis. Immunology Today 1994;15:7-10.

Clozel 1985

Clozel M, Daval JL, Monin P, Debruc C, Morselli PL, Vert P. Regional cerebral blood flow during bicuculline-induced seizures in the newborn piglet: effect of phenobarbital. Developmental Pharmacology and Therapeutics 1985;8:189-99.

Demopoulos 1977

Demopoulos HG, Flamm ES, Seligman MC. Antioxidant effects of barbiturates in model membranes undergoing free radical damage. Acta Neurologica Scandinavica 1977;56 Suppl 64:152-3.

Finer 1981

Finer NN, Robertson CM, Richards RT, Pinnell LE, Peters KL. Hypoxic-ischemic encephalopathy in term neonates: perinatal factors and outcome. Journal of Pediatrics 1981;98:112-17.

Hossman 1983

Hossman KA. Neuronal survival and revival during and after cerebral ischaemia. American Journal of Emergency Medicine 1983;1:191-7.

Kinoti 1993

Kinoti SN. Asphyxia of the newborn in east, central and southern Africa. East African Medical Journal 1993;70:422-33.

Levene 1985

Levene MI, Kornberg J, Williams THC. The incidence and severity of post-asphyxial encephalopathy in full-term infants. Early Human Development 1985;11:21-6.

McCord 1985

McCord JM. Oxygen-derived free radicals in postischemic tissue injury. New England Journal of Medicine 1985;312:159-63.

McDonald 1990

McDonald JW, Johnston MV. Physiological and pathophysiological roles of excitatory amino acids during central nervous system development. Brain Research. Brain Research Reviews 1990;15:41-70.

Nilsson 1971

Nilsson L. The influence of barbiturate anaethesia upon the energy state and upon acid-base parameters of the brain in arterial hypotension and in asphyxia. Acta Neurologica Scandinavica 1971;47:233-53.

Sarnat 1976

Sarnat HB, Sarnat MS. Neonatal encephalopathy following fetal distress. Archives of Neurology 1976;33:696-705.

Siesjo 1992

Siesjo BK. Pathophysiology and treatment of focal cerebral ischaemia. Part II: Mechanisms of damage and treatment. Journal of Neurosurgery 1992;77:337-54.

Singh 1991

Singh M, Deorari AK, Khajuria RC, Paul VK. A four year study on neonatal morbidity in a New Delhi hospital. Indian Journal of Medical Research 1991;94:186-92.

Thornberg 1995

Thornberg E, Thiringer K, Odeback A, Milsom I. Birth asphyxia: incidence, clinical course and outcome in a Swedish population. Acta Paediatrica 1995;84:927-32.

Younkin 1986

Younkin DP, Delivoria-Papadopoulos M, Maris J, Donlon E, Clancy R, Chance B. Cerebral metabolic effects of neonatal seizures measured with in vivo 31P NMR spectroscopy. Annals of Neurology 1986;20:513-9.

Other published versions of this review

Evans 1998

Evans DJ, Levene MI. Anticonvulsants for preventing mortality and morbidity in full term newborns with perinatal asphyxia. In: Cochrane Database of Systematic Reviews, Issue 3, 1998.

Evans 2001

Evans DJ, Levene MI. Anticonvulsants for preventing mortality and morbidity in full term newborns with perinatal asphyxia. In: Cochrane Database of Systematic Reviews, Issue 2, 2001.

Comparisons and data

Comparison or outcome
Studies
Participants
Statistical method
Effect size
01 Barbiturates vs. control
01 Death (before neurodevelopmental assessment by 3 years)
5
228
RR (fixed), 95% CI
1.13 [0.59, 2.17]
02 Severe neurodevelopmental disability in survivors examined
3
77
RR (fixed), 95% CI
0.61 [0.30, 1.22]
03 Death or severe neurodevelopmental disability
3
110
RR (fixed), 95% CI
0.78 [0.49, 1.23]
02 Phenobarbital vs. control
01 Seizures within neonatal period
2
113
RR (fixed), 95% CI
0.72 [0.42, 1.23]
02 IQ in survivors at 6 years of age
1
30
WMD (fixed), 95% CI
-3.00 [-21.57, 15.57]
03 Hypotension requiring inotropes
2
118
RR (fixed), 95% CI
Not estimable
03 Thiopentone vs. control
01 Seizures within 3 days of age
1
32
RR (fixed), 95% CI
1.03 [0.74, 1.44]
02 Seizures at 3 days of age
1
32
RR (fixed), 95% CI
1.06 [0.40, 2.77]
03 Hypotension requiring inotropes
1
32
RR (fixed), 95% CI
1.76 [0.98, 3.16]
04 Phenobarbital vs. phenytoin
01 Seizures within the first 7 days of age
1
17
RR (fixed), 95% CI
0.89 [0.07, 12.00]
05 Chloral hydrate vs. diazepam
01 Neonatal seizures beyond 3 days of age
1
28
RR (fixed), 95% CI
0.32 [0.03, 3.16]

 

01 Barbiturates vs. control

01.01 Death (before neurodevelopmental assessment by 3 years)

01.02 Severe neurodevelopmental disability in survivors examined

01.03 Death or severe neurodevelopmental disability

02 Phenobarbital vs. control

02.01 Seizures within neonatal period

02.02 IQ in survivors at 6 years of age

02.03 Hypotension requiring inotropes

03 Thiopentone vs. control

03.01 Seizures within 3 days of age

03.02 Seizures at 3 days of age

03.03 Hypotension requiring inotropes

04 Phenobarbital vs. phenytoin

04.01 Seizures within the first 7 days of age

05 Chloral hydrate vs. diazepam

05.01 Neonatal seizures beyond 3 days of age


Contact details for co-reviewers

Prof Malcolm Levene
Department of Pediatrics
General Infirmary at Leeds
D Floor, Clarendon Wing
Leeds
UK
LS2 9NS
Telephone 1: +44 113 292 3905
Telephone 2: +44 113 292 3902
E-mail: medmil@leeds.ac.uk

Dr Maria Tsakmakis
Specialist Registrar
Department of Neonatal Medicine
North Bristol NHS Trust
Neonatal Intensive Care Unit
Southmead Hospital
Bristol
UK
BS10 5NB
E-mail: maria.tsakmakis@btopenworld.com

 
This review is published as a Cochrane review in The Cochrane Library, Issue 3, 2007 (see http://www.thecochranelibrary.com for information). Cochrane reviews are regularly updated as new evidence emerges and in response to feedback. The Cochrane Library should be consulted for the most recent version of the review.