Patients with refractory seizures present both diagnostic and therapeutic challenges. Formulating a treatment plan requires a thorough understanding of the patient's seizure frequency, seizure type(s), serum concentrations of antiepileptic drugs (AEDs) and any underlying systemic conditions that could exacerbate seizures. This section provides a guide to several of these diagnostic evaluations.

Seizure calendar

Encouraging your patients to keep track of their seizures, medication intake, other epilepsy treatments, and side effects can be quite helpful in assessing whether they have refractory seizures, and if so, possible reasons and treatment approaches. Calendars or diaries are the best ways for patients or their families to track seizures and side effects. On, we provide a user-friendly way for patients or families to start and keep a Seizure Diary. Additional sections on will help teach your patients about the importance of tracking seizures, tips for seizure observation, and using seizure diaries.

EEG monitoring

The most direct way to confirm the diagnosis of epilepsy in a patient with refractory seizures is to obtain EEG monitoring. This is feasible if the patient's events are frequent enough to justify either inpatient or outpatient EEG monitoring. The interictal EEG is not useful in confirming the diagnosis of epilepsy.

Two types of monitoring are in general use:

  • an outpatient procedure with a duration of 6 to 8 hours (Daytime Monitoring, or DAYMON)
  • inpatient monitoring, which continues for 24 hours or more

Outpatient studies are less expensive and more convenient than inpatient monitoring. DAYMON is most appropriate for patients with relatively high seizure frequencies-at least three events per week. To increase yield, DAYMON should be carried out when the patient is sleep-deprived.

If the patient's seizure frequency is relatively low, inpatient video-EEG monitoring for 24 hours or more is indicated. This procedure requires hospital admission and a dedicated staff. Although more costly than DAYMON, inpatient monitoring is effective. More than one event may be recorded, increasing diagnostic certainty if the events are stereotyped. Inpatient monitoring also allows recording of a full night's sleep, increasing the possibility of recording sleep-provoked epileptiform activity as well as nocturnal clinical events. Several days of monitoring may be required before the diagnosis is made.

Video-EEG monitoring procedures

During video-EEG monitoring, the patient wears an EEG transmitter connected to a wall outlet by coaxial cable. Wall-mounted video cameras provide continuous behavioral observation. Both EEG and video signals are transmitted to a control room, where the EEG is reformatted and conducted to a video monitor. The EEG signal and video are displayed simultaneously for on-line observation, and both are recorded on videotape. The EEG may be recorded on paper or stored on optical disc.

The patient can move about and carry out normal activities, such as napping, talking, and watching television. Participation by a family member or friend is encouraged, especially someone who has observed the patient's events in the past.

The most important task is to ensure that the recorded event(s) are typical of the patient's spontaneous attacks. This task can be accomplished only by reviewing the recorded attack with a person who has witnessed such events.

Serum AED concentrations

Blood tests in patients taking antiepileptic drugs (AEDs) may have several purposes:

  • monitoring compliance
  • following the results of AED dosage changes
  • establishing a patient's maximum tolerated serum level
  • looking for early signs of adverse effects (e.g., hepatic, renal, hematopoietic)

Within the first 6 months of treatment with a newly prescribed AED, systemic toxicity and neurotoxicity are as likely to contribute to AED failure as lack of efficacy. Allow sufficient time during office visits to determine whether the patient is experiencing any side effects.

The extent of some side effects can be difficult to assess. For instance, cognitive impairment (especially memory loss) is a common complaint. Neuropsychologic testing sometimes is needed to determine the extent of cognitive impairment and whether it is medication-related.

Check serum levels

Attempt to correlate drug serum levels with the patient's side effects before abandoning a medication. This can be done by obtaining levels when a patient is experiencing side effects and comparing them with those obtained when the patient is free from symptoms. Referring to the patient's seizure calendar may be helpful in planning the timing of drug levels, to prove a cause-and-effect relationship between peak levels and side effects. The serum levels associated with toxicity vary from one patient to another and may occur within the usual therapeutic range.

Total serum levels may be misleading. Free unbound serum levels of phenytoin and valproate should be checked in patients with low albumin levels or patients who are taking multiple drugs that are tightly protein-bound. In such patients, free levels should be multiplied by 10 to approximate the desired total serum level.

Other factors may influence serum levels:

  • laboratory error
  • generic substitution for brand-name AEDs
  • variable potency of pills (following improper storage, for example)
  • menstrual cycle (midcycle serum AED levels may be higher than during the premenstrual period or menses)

Adjust medication regimen

For patients who have peak-level side effects from an AED, the usual strategy is to modify the medication regimen or treatment schedule to minimize side effects. For example, suppose that a patient has only nocturnal seizures and takes equal doses of an AED twice a day. If the patient experiences side effects during the afternoon from the morning AED dose, those side effects may be eliminated without compromising seizure control by shifting part of the morning dose to the bedtime dose.

Spreading out the daily dosage over smaller, more frequent doses or using a slow-release form of the same medication is another possible solution to the problem of peak-level side effects.

Adapted from and Schachter SC. Treatment of seizures. In: Schachter SC, Schomer DL, eds. The comprehensive evaluation and treatment of epilepsy. San Diego, CA: Academic Press; 1997. p. 61-74. With permission from Elsevier (

Overnight sleep test

An overnight sleep test is helpful when a sleep disorder is suspected, either in the differential diagnosis of nocturnal events, or as a contributing factor to seizure exacerbation during the day.

Authored By: 
Steven C. Schachter MD
Authored Date: