Phenytoin is used to treat acute tonic–clonic seizures in children who have not responded to a benzodiazepine. In the UK, the loading dose of phenytoin is 18 mg/kg. There is limited evidence on whether this loading dose will achieve the desired levels in paediatric patients. Intravenous loading doses of phenytoin were retrospectively and prospectively audited over 19 months. Doses were normalised for weight and compared with the serum phenytoin concentrations. Errors in dose calculations and adverse events were recorded. Serum phenytoin concentrations were measured on 31 occasions in 27 children (24 retrospective and 10 prospective) between 60 and 180 (median, 153) min after completion of the loading dose. Serum phenytoin concentrations were within the therapeutic range (10–20 μg/ml) on 24 occasions. No errors in dose calculations or adverse effects were identified. A phenytoin loading dose of 18 mg/kg gave serum concentrations within the recommended therapeutic range in most children.
Statistics from Altmetric.com
Intravenous phenytoin has been used as a treatment for convulsive status epilepticus in children since 1959.1
There are limited phenytoin pharmacokinetic (PK) data in children. A study of 16 Kenyan children with cerebral malaria treated with a loading dose of phenytoin (18 mg/kg) found a high phenytoin Cmax of 23.3 µg/ml (above the therapeutic range), with a time to Cmax of 20 min.5 Given the different ethnic background and cerebral malaria, it was considered appropriate to audit the use of intravenous phenytoin in Caucasian children to obtain more contemporary PK data using a loading dose of 18 mg/kg.
An audit of all patients who received an intravenous loading dose of phenytoin to treat an acute generalised tonic–clonic seizure was undertaken over 19 months at Alder Hey Children's Hospital (Audit number 3272). Data from 1 January 2009 to 31 May 2010 (17 months) were collected retrospectively, and from 1 June 2010 to 31 July 2010, prospectively.
For retrospective data, serum phenytoin concentrations were retrieved from the Trust computer system (Meditech).
For prospective data, patients were identified through the neurology, neurosurgery, emergency department and intensive care. Hospital guidelines state that all patients who receive a loading dose of phenytoin should have a blood concentration measured 1–2 h following completion of the infusion by finger prick or fresh venepuncture. All wards and units were notified of the audit.
Clinical information was obtained from the case notes and drug prescription charts regarding the indication for phenytoin and possible side effects. Efficacy data were collected for the patients treated prospectively but not for patients treated retrospectively.
Case notes and prescription charts were reviewed to identify any incorrect dose calculations or prescriptions of phenytoin.
Complete data were collected on 34 intravenous loading doses (24 retrospective and 10 prospective) in 27 children (20 retrospective and 7 prospective). None had received intravenous or oral phenytoin in the preceding month. The presenting seizure was tonic–clonic (32 occasions) and clonic (2 occasions).
The serum phenytoin concentration was measured after 31 of the 34 loading doses (91.2%) between 60 and 180 (median 135) min following completion of the infusion. Levels were not measured in two patients and one had a level measured after 495 min. These three patients were included in the dose analysis but excluded from the analysis of serum concentrations.
On 24 occasions, serum phenytoin concentrations were within the therapeutic range (figure 1). The median concentration (n=31) was 15.3 μg/ml. One child had a subtherapeutic phenytoin concentration 110 min after the loading dose (2.7 μg/ml); six children had levels above the therapeutic range (20.2, 20.3, 21.5, 21.6, 22.3 and 23 μg/ml). Mean time for sampling in the children with supra-therapeutic concentrations was 147 min (range 60–175 min) following completion of the loading dose.
On 33 of the 34 occasions (97%) the dose ranged from 16.6 to 19.7 (median 18.0) mg/kg. This variation reflected ‘rounding up or down’ to aid dispensing. The remaining infant was oedematous and received a lower dose of 14.9 mg/kg based on an estimated ‘dry weight’ for his age. Of the 34 loading doses, 24 (71%) were calculated on the patient's actual and not estimated weight. The median dose was 18.0 mg/kg (range 17.0–19.7) and the median serum phenytoin concentration was 16.2 μg/ml in this group.
Four of the six (67%) serum concentrations above and the single one below the therapeutic range were in patients in whom the loading dose was based on actual weight. The serum concentration in the oedematous patient was 13.3 μg/ml, within the therapeutic range. For the nine patients with doses based on estimated weights, the median dose was 18.0 mg/kg (range 16.7–18.2) and the median serum concentration was 14.5 μg/ml. Two patients had a level above and none below the therapeutic range.
Phenytoin terminated the seizure in 8 of the 10 (80%) of the prospectively treated patients with serum concentrations of 11.2–20 μg/ml. The serum concentrations were 15.4 and 23 μg/ml respectively in the two patients whose presenting seizure persisted.
There was no documented error in the calculation or administration of intravenous phenytoin, or reported toxicity.
We have demonstrated that the recommended loading dose of intravenous phenytoin (18 mg/kg) results in a serum concentration within therapeutic range in the majority of children.
Serum phenytoin concentrations are usually measured approximately 1–3 h after completion of the loading dose.6 This is to determine whether the patient may require (and can safely be given), a further part-load of phenytoin and as a guide to when to administer the first maintenance dose of the drug.
Over 90% of serum phenytoin concentrations in the current study were measured within 3 h following completion of the infusion. Peak serum concentrations are likely to be higher than those measured because they will be achieved soon after completion of the infusion. This would also suggest that the use of a loading dose greater than 18 mg/kg may be associated with increased levels. This concern is supported by the only previously reported study.4 Serum concentrations were measured in 20 children receiving 9.4–21.3 mg/kg, and of the five children who received ≥20 mg/kg, all had supra-therapeutic serum concentrations with a median of 32 μg/ml.4
Most (71%) of the loading doses in the current study were calculated and prescribed using actual and not estimated weights; although the numbers are small, supra-therapeutic levels were seen in patients in both groups. It is reassuring that these data suggest that estimated weight is accurate enough to achieve levels of phenytoin within the target therapeutic range.
Limited data from the 10 patients in the prospective group indicated that the loading dose was effective, in that it terminated the presenting seizure in 8 of the 10 patients.
Concern over the use of intravenous phenytoin relates to the rate of infusion and the risk of inducing cardiac arrhythmias.7 Concern has also been raised over errors that might arise from a miscalculation of the dose based on 18 mg/kg. In the current study, no prescribing errors were identified in the 34 patients who received intravenous phenytoin in a dose of 18 mg/kg.
This audit clearly has limitations. The population is small and uses retrospective and prospective data. The relatively low ascertainment rate in the retrospective cohort compared with the prospective highlights one of the problems with retrospective studies. There is also a paucity of data on the correlation of serum concentrations and termination of the seizure because this information was not available for most doses (n=24). Finally, our assessment of phenytoin toxicity is limited. Five of the six patients with supra-therapeutic serum phenytoin concentrations were in the retrospective group, and because no symptoms of toxicity were documented in the case notes, we assume that this was the case but clearly, this may be a false assumption.
A prospective audit is currently in preparation to assess every child treated for an acute tonic–clonic seizure using this hospital's (and national, APLS) guideline, including correlation of serum blood concentrations of phenytoin with prescribing, efficacy, safety and toxicity data.
The authors thank all the staff in Alder Hey (especially those in the Emergency Department and Intensive Care Unit) who helped to identify the eligible patients.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
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.