Article Text


Phenytoin in traumatic brain injury
  1. Elizabeth A Hunt
  1. PICU Fellow at Johns Hopkins Hospital in Baltimore

Statistics from

A 12 year old boy is admitted to the paediatric intensive care unit after a motor vehicle collision, where he sustained a severe closed head injury. He lost consciousness at the scene and was intubated in the emergency department for a Glasgow Coma Score of 8 and no gag reflex. The boy has no history of seizure activity in the past or at the scene. Your local “Traumatic Brain Injury Protocol” recommends that he receive phenytoin for seizure prophylaxis. You have recently cared for a child who nearly died from phenytoin hypersensitivity syndrome and would like to know if there is a good indication for the drug.

Structured clinical question

In a child with a traumatic brain injury [patient] does phenytoin prophylaxis [intervention] prevent subsequent seizures and/or improve neurological outcome [outcome]?

Search strategy and outcome

Secondary sources—none since 1972.

Systematic reviews—two, performed by same authors: Cochrane Library (Issue 3, 2001) and Evidence-Based Medicine (v3, 1998; initially published in J Neurol Neurosurg Psychiatry, 1998); reviewed use of “anti-epileptic drugs”, not just phenytoin.

Primary research—PubMed query using MeSH subject headings: “Craniocerebral Trauma” AND “Epilepsy, Post-Traumatic” AND “phenytoin” and filter “therapy”.

Search results—52 articles; after “therapy” filter: 4 relevant (1 not placebo controlled, blinded or readily available, therefore not included here, but is included in systematic reviews). See table 3.

Table 3


In addition to determining the potential efficacy of phenytoin as prevention for post-traumatic epilepsy, there are other important outcomes to bear in mind. Although anti-epileptics appear to decrease the rate of early seizures, this has not been shown to result in a lower rate of post-traumatic epilepsy, mortality, or improved cognitive outcome.2,4 In addition, multiple studies have shown that continued use of phenytoin in this population is associated with “negative effects on cognitive performance”.5

One must also consider that phenytoin is not a benign drug. Although studies have not revealed a high rate of side effects in patients treated with phenytoin for one week post-injury, physicians must still be aware of the potential for not only serious but potentially fatal reactions to the drug. A search of the literature for adverse reactions to prophylactic phenytoin in head injured patients yielded case reports of intravenous site reactions,6 exfoliative dermatitis,7 granulocytopenia,8 and transient hemiparesis.9 In other patient populations, phenytoin has been associated with permanent B cell immunodeficiency10 and fatal phenytoin hypersensitivity syndrome.11 There have also been multiple publications which show that total phenytoin concentrations can be normal in a trauma patient, yet be associated with high free, unbound concentrations because of low albumin. This may be associated with a higher rate of side effects at normal “therapeutic” concentrations.12

Recently, a consensus guideline was published stating that “the routine use of seizure prophylaxis later than 1 week following head injury is not recommended”.13 The guideline recommends using a risk–benefit analysis to decide whether phenytoin or carbamazepine should be considered for high risk patients in the first week post-injury.

Clinical bottom line

  • In the first seven days after serious head injury, phenytoin does prevent seizures in some patients (NNT = 10) but has not been shown to decrease mortality or post-traumatic epilepsy. For high risk patients, a physician must weigh the potential risks and benefits before making a decision.

  • If phenytoin is used in trauma patients, measure free, “unbound” phenytoin concentrations to avoid toxicity.

  • There is currently no indication for prophylactic phenytoin beyond the first week post head injury.


View Abstract

Review history and Supplementary material



    Table 3

    CitationStudy groupStudy type (level of evidence)OutcomeKey resultsComments

    Schierhout et al (2001)6 controlled trials, 1218 randomised patients; includes studies with both adults and children, some studies used other antiepileptics than phenytoin Systematic review (level 1a)Early seizures (in 1st 7 days post-injury) Late seizures (after 1st 7 days post-injury) Mortality Neurologic disability, (Glasgow Outcome Score) RR for early seizure prevention 0.34 (CI 0.21–0.54); NNT=10; No? mortality, RR=1.15 (CI 0.89–1.51); No ? in death and neurologic disability RR= 0.96 (CI 0.72–1.26); No ? in late seizures Pooled RR =1.28 (CI 0.90–1.81) Early seizure prevention values represent 4 studies with phenytoin and 2 with other agents Studies very heterogeneous in relation to late seizure outcome
    McQueen et al (1983)164 patients ages 5 to 65 with serious head injuries (pts with seizures 1st wk post injury excluded)Double blind RCT (level 1b)Late seizures, overall mortality, skin rashesNo significant differences Late seizure ARR = -0.77% (CI -9.59 to +8.04) Mortality ARR = -3.45% (CI -9.56 to +2.66) Rash ARR = -4.7% (CI -10.32 to +0.91) Loss to follow up at 2 years 1.2% (5% if include deaths) Authors' note: based on low incidence of events, future trials will have to be ~6x larger than theirs to detect an effect
    Young et al (1983)41 head injured children; (total study 244 pts, adult data published separately with similar results)Double blind RCT (level 2b)(Early seizures; data not published in this study, was collected for total study) Late seizures Overall mortalityNo significant differences Late seizures ARR = -5.75% (CI -23.16 to +11.66) Mortality = 4.55% (CI -14.88 to +23.88)Loss to follow up at 2 years 26% for overall study For children (1 withdrawn, 4 deaths in first wk not included in analysis of final 41 pts), otherwise no loss to follow up
    Temkin et al (1990)404 patients with serious head trauma, 16 years of age and olderDouble blind RCT (level 1b for early seizure outcome, but for late seizure outcome large loss to follow up in both groups at 2 years, thus level 2b) Early seizures (in 1st wk post-injury)3.6% in treatment arm, 14.2% placebo, RR=0.27 (CI 0.12–0.62)24% loss to follow up at 24 months (22.5% in treatment group, 25% in placebo group)
    Late seizures (from day 8 to the end of the year)No significant difference; 21.5% treatment, 15.7% placebo RR=1.2 (CI 0.71–2.02)
    Overall mortality, skin rashesNo significant difference Mortality ARR = -2.64% (CI -10.74 to +5.27) Rash ARR = -3.35% (CI -9.27% to 2.58)

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.