For Patients With Seizures, Is it Safe to Slowly Titrate Antiepileptic Drugs to Clinical Effect Rather Than Blood Levels?

Michael R. Sperling, MD


December 27, 2000


When using carbamazepine, sodium valproate, or phenytoin in general practice for patients with epilepsy, is there any evidence to support slowly titrating the drug dose to clinical effect rather than to adequate blood levels? I have seen this done; the patient is well with good seizure control and no side effects. The drug level is often subtherapeutic, but the treating physician measures the level only to ensure that the patient is not toxic. Part of the motivation to treat in this manner is to avoid oversedation.

Thank you for your opinion.

T. Quin, MD

Response from Michael R. Sperling, MD

The notion of titrating the dose to clinical effect has some practical and theoretical problems. However, there are 2 situations that can be discussed separately.

The first is that of patients whose seizures can be controlled completely with medication: The desired clinical effect is to prevent all seizures. One only knows that seizures have been prevented by observing that nothing has happened. This knowledge is attained only in retrospect, and one must wait a long time -- years -- to be certain that the needed dose has been given. Hence, a slow titration of dose to desired clinical effect means that you must accept that seizures will recur during the course of titration until the necessary serum level is established. This does not seem like a good idea, given the adverse consequences of seizures -- including delays in restoring driving privileges, the psychosocial impact, and potential for injury or even death. Hence, it seems more reasonable to initially target a therapeutic dose or level, perhaps at the low end or middle of the therapeutic range, to minimize the probability of seizure recurrence. The dose could then be titrated higher from this point if seizures recur.

If side effects occur at a low serum level or dose, many epileptologists might try a lower dose after discussing it with the patient. This applies to those with seizures that can lead to impaired consciousness or falling, or have the potential to cause injury; minor simple partial seizures that pose no disability or risk could be treated with a low-dose titration schedule as suggested by the questioner, because there are no adverse consequences to the seizures.

The second case is patients whose seizures cannot be completely controlled (known as refractory) and/or who are already taking 1 or more anticonvulsant drugs. These individuals often can be treated in the suggested way, slowly titrating the dose upward looking for therapeutic effect. This method only works with relatively frequent seizures, because the titration schedule could be too slow if seizures only happen rarely.

In general, the approach to treat should be sensible and goal directed. When the aim is immediate and complete control, stronger measures are needed. If slow, incremental control is desired, then a more gradual approach can be adopted.