Introduction

The diagnosis of epilepsy can have a devastating impact on the quality of a person's life, not least because of the risks associated with driving. But how real is the problem? Seizures that occur while driving frequently cause serious crashes. In the USA, only patients with "controlled" seizures are allowed to drive - although the definition of "controlled" varies from state to state.¹ Doctors at Johns Hopkins Hospital have evaluated the actual risk factors in such situations, to see whether the present rules can be made more rational, i.e. based on different patients' likelihood of an accident.

What was done

The doctors examined 50 patients who had driving crashes during a seizure (the 'cases'), and compared them with 50 patients with epilepsy who drove but didn't have crashes - these were 'controls'. Case and control patients were matched up by gender, age, and attending the same epilepsy clinic.

The information collected was for a 12-month period leading up to the crash for the cases, and the 12-month period beginning mid-1996 for the controls. It included:

• Age & gender
• Type of seizures
• Number of seizures per month
• Medication
• Driving hours per week
• Number of years driving
• Purpose for driving (e.g. job, shopping, recreation)
• Type of road usually driven
• Purpose for driving at time of crash
• Type of road at time of crash

What was found

Initially, 61 cases were identified, but 11 of these crashed during their very first seizure, so they were not included for the comparison part of the study, leaving 50 cases and 50 controls. There were 41 men and 9 women in the 50 cases (and also, therefore, in the matched controls).

At the time of the crash, 60% of the patients were driving to or from their work. Many of the crashes were severe, with injuries, and in two cases, fatalities. Thirteen of the controls had seizures while driving, but didn't have a crash.

Out of all the possible factors analyzed, only the following were associated with a significantly reduced likelihood of having a crash when driving:

• Long seizure-free interval
• Reliable 'aura' (an awareness that a seizure is about to occur)
• Few prior non-seizure-related accidents
• Recent change epileptic in medication

The single most significant factor - a 12-month or longer seizure-free interval - was associated with a 93% lower chance of having a crash, compared with patients with shorter seizure-free intervals. A seizure-free period of 6 months or long reduced the likelihood of a crash by 85%, compared with shorter seizure-free periods. Seizure-free periods less than 6 months failed to show a significant benefit on the risk of having a crash.

Having a reliable aura (awareness that a seizure was coming) significantly reduced patients' odds of crashing. However, 13 patients still crashed in spite of having an aura - either the aura was too brief, the patient was blocked in traffic and couldn't pull off the road, or the patient tried to get home before the seizure occurred.

Although the number of years driving did not influence the likelihood of having a seizure-related crash, fewer non-seizure related crashes was significantly beneficial - it suggests that good drivers have a better chance of avoiding any sort of crashes.

About 1 in 3 of all the patients (cases and controls) admitted they didn't always take their medications regularly as prescribed. However, among the 50 cases, 10 crashes occurred in patients immediately after they had missed one or more of their medication doses. On the other hand, reducing or switching medication seems to have had a 'protective' effect. (Changing meds may mean the patients were under closer medical supervision, and therefore more likely to take their meds and/or drive more carefully).

Comment

What conclusions can we draw from this analysis? The authors are cautious in making recommendations, as they point out that their study is, in fact, quite small. Obviously, a longer seizure-free period is preferable to a short one - but setting a legal long period has to be weighed against the difficulties imposed on people getting to work, and an increased risk of non-adherence to the laws. A 12-month mandatory period would avert 74% of crashes, whereas it would restrict driving for 52% of patients without crashes. Clearly, trade-offs have to be made. Effects of a shorter mandatory period might be offset by improved observance with the law.

Another consequence: physicians should work with their patients to optimize anti-epileptic medication. Good medication means effectiveness without severe side effects, and therefore a greater likelihood of compliance. Also, each patient should discuss any pre-seizure symptoms (if there are any) with his or her doctor, to try to determine if they do, in fact, represent a reliable aura. And bad drivers with previous non-seizure crashes should be identified and appropriately cautioned.

Studies such as this one need to be duplicated with greater numbers of patients, to provide hard numbers for necessary legal restrictions on driving. In time, as better control of epilepsy is achieved, it may be possible to relax existing no-driving strictures further.