When an admission to the intensive care unit (ICU) is elective, such as for a brief postoperative stay, there is time to plan the approach in advance. Several issues should be considered: Is the patient non per os before and after the procedure, and if so, for how long?

Patients should be instructed to take their antiepileptic drugs (AEDs) with sips of water before surgery, even during a presurgical non per os period.

How long is the operation or procedure, and are doses of medication missed during that time?

If the surgery and postoperative period involve a prolonged period when the patient cannot use his or her GI tract, consider changing the regimen to an IV AED. The lack of IV forms for many outpatient oral AEDs complicates ICU care. IV choices in the United States include phenytoin, valproic acid, phenobarbital, or benzodiazepines. If possible, a gradual transition to oral administration of one of these medications before elective surgery is often preferable to an abrupt transition to the IV form postoperatively. If the patient has previously taken the other medication, consider whether he or she experienced any adverse reactions or allergies. If it is not practical to make the transition to a new AED before an elective procedure, it is common to load patients with the IV AED before or during surgery.

The type of IV access available is critical for certain kinds of therapy:

  • IV phenytoin should best be given via high-flow (preferably, central venous) access, owing to the risk of thrombophlebitis with peripheral IV administration.31,32
  • Central access is not required for administration of fosphenytoin, which can be injected more rapidly intravenously or intramuscularly, has a lower reported incidence of thrombophlebitis and other adverse reactions, and is preferable to IV phenytoin in all circumstances.

If the time period off of medications is brief, another option is to place the patient on round-the-clock doses of benzodiazepines as bridge therapy, after which the patient's regular AED can be resumed. This may only involve a few doses of lorazepam, chlorazepate, or clonazepam. A benzodiazepine with a relatively long antiepileptic half-life, such as lorazepam, may be best.

Once the patient has been moved into the recovery room or more central ICU area, the choice of AED remains a function of how long the patient's typical AED regimen is interrupted and which routes of therapy are available for administration.

Will the patient be able to swallow after surgery? Will there be some other access to the gastrointestinal tract? Will enteral nutrition be given?

Using the patient's normally functioning GI tract is almost always preferable to IV administration, but there are several potential pitfalls. Nasogastric tube (NGT) feedings, for example, raise stomach pH and can interfere with absorption of AEDs that require an acidic environment. For example, phenytoin levels can plummet after NGT feedings are initiated.33 One solution for this problem is to bolus-feed the patient when possible, limiting the patient to a number of boluses of nutrition per day, more closely mimicking normal eating. Holding continuous NGT feedings for several hours before and after administration of phenytoin elixir can help, but serum levels must be monitored carefully to assess the effect on absorption.

Administration of AEDs through feeding tubes can pose difficulties. The patient's usual oral medication often can be used, but levels may fluctuate widely if an appropriate preparation is not used correctly. Suspensions (e.g., phenytoin and carbamazepine) must be vigorously shaken before they are administered. AED levels may plummet if they are not. Preparations of AEDs that are designed to be chewed or mixed in with food, such as valproate (Depakote) sprinkles and the coated carbamazepine beads emptied out of Carbatrol capsules, can be given via enteric tubes.34 These preparations of valproic acid and carbamazepine are preferable to crushed enteric-coated oral preparations of these drugs. Whether they are preferable to the liquid form of these medications is not clear.

Table: Antiepileptic Drug Preparations Delivered Via Enteric Tubes lists preparations of standard AEDs that can be administered through gastric or intestinal tubes with reasonable expectations of absorption. In general, sustained-release and enteric-coated preparations should not be used for this purpose. Chewable preparations are acceptable for crushing and administering through enteric tubes, although they may result in more rapid serum peak levels than slower-to-dissolve preparations. Specific medication preparations need to be judged individually. Those marked "Do not crush, chew, or break" on the package insert can be assumed to result in erratic absorption when these directions are ignored. It is also vital to flush feeding tubes to be sure that the full dose of medication is delivered to the patient.

A common problem with assessing the efficacy of AED delivery strategies in the ICU is waiting for changes in the daily level, as it often results in prolonged periods of underdosing while assessing a variety of ineffective strategies.

Does the patient require medications that can alter the seizure threshold or the absorption, metabolism, or excretion of AEDs?

Several medications, most commonly antacids and histamine 2 blockers, can affect absorption of AEDs into the blood stream through the alimentary tract. Through lowering gastric pH, these medications can interfere with absorption of many AEDs, most notably phenytoin. In this case, AED dosing should be timed away from the administration of these agents, when possible. If this does not help to maintain levels, it may be necessary to change to an IV agent. Increasing the dose of medication may be useful in some cases, but the response should be assessed through peak serum levels after a dose of medication.

Adapted from: Kolb SJ and Litt B. Management of epilepsy and comorbid disorders in the emergency room and intensive care unit. In: Ettinger AB and Devinsky O, eds. Managing epilepsy and co-existing disorders. Boston: Butterworth-Heinemann; 2002;515–535. With permission from Elsevier (www.elsevier.com).

Authored By: 
Steven J. Kolb Md PhD
Brian Litt MD
Reviewed By: 
Steven C. Schachter MD
Friday, April 30, 2004