Epilepsy and Driving in Europe Final Report v2 en
Epilepsy and Driving in Europe Final Report v2 en
Epilepsy and Driving in Europe Final Report v2 en
__________
Members:
Mr Joël Valmain European Commission, DG TREN.
Dr Jaume Burcet Darde Spain.
Dr Bernhard Gappmaier Austria.
Dr John Kirker Ireland.
Prof Guenter Kraemer Germany.
Dr Nicole Markschies Germany.
Dr Mikael Ojala Finland.
Dr Eric Schmedding, Belgium. President
Dr Anders Sundqvist Sweden.
Dr Elena Valdès Spain.
Prof Hervé Vespigniani, France.
Dr Graham Wetherall, United Kingdom. Secretary
Dr Jörgen Worm-Petersen Denmark.
Consultants:
Prof D Chadwick UK
Prof Van Donselaar The Netherlands
Prof Ettore Beghi Italy
Dr Marc de Krom The Netherlands
Mrs Maria Teresa Sanz-Villegas European Commission, DG TREN
The SEWGED had meetings on 5th May 2004; 20th October 2004; 19th and 20th January 2005 and 2nd and 3rd
March 2005.
1
Final report; 3 April 2005 2
Contents
1. Introduction 4
Historical background
A practical problem: the question of compliance
3. When does the patient reach this risk-threshold? The influence of the
seizure-free interval in different situations 14
First unprovoked epileptic seizure
What happens in the first year?
What happens after five years?
The first unprovoked cryptogenic seizure
The first provoked epileptic seizure
Epilepsy: multiple seizures
The remaining risk of recurrence after a given seizure-free period
A possible consequence: limited licence
Recommendation 4
5. Legal issues 21
Who should do the assessment?
The question of immunity
The legal status of a statistical decision
Legal obligations
Should taxi drivers etc. be included in group 2, if so, should the risk assessment be equally severe?
Recommendation 6
2
Final report; 3 April 2005 3
9 References 35
10 Annexes 39
Annex 1: The questions put to the working group
Annex 2: Glossary of terms
Annex 3: The calculation of increased risk for group 2
Annex 4: Other data used for calculation
Annex 5: Recommendations for the future
3
Final report; 3 April 2005 4
1. Introduction
The cumulative incidence of epilepsy is at least 4% of the population. The prevalence of active epilepsy in the adult
population is 4 to 10 in 1000 people (Hauser et al. 1996; Goodridge et al. 1983). For the European union we
assumed a value of 6 in 1000. Of these patients, a substantial number hold a drivers licence (Sonnen 1995). In
general, driving is experienced as one of the top concerns of people with epilepsy, as is noticeable in the daily
practice of any neurologist. In surveys, driving is listed as a first or second concern by people with epilepsy, after the
wish to be seizure-free (Gilliam et al. 1997; Taylor et al. 2001; Fisher et al. 2000). On the other hand, driving while
having active epilepsy clearly poses an increased risk, ( Krauss et al. 1999; Berg et al. 2000) while drivers with
epilepsy who are in compliance with driving restrictions and with medication intake pose no excess danger (Krauss
1999). This makes the topic of “epilepsy and driving” of importance to neurologists and the regulators of driver
licensing alike.
HISTORICAL BACKGROUND
In the European Union, the regulations about driver licensing used to differ greatly among member states. (Fisher et
al 1994) With the support of the European government, this led to the formation of European workshops on driving
licence regulations in May 1995 and March 1996 organised by the International League against Epilepsy (ILAE) and
the International Bureau for Epilepsy (IBE)(Sonnen 1995 and 1997). The recommendations of these workshops were
not reflected in an official European guideline or in European law. In these recommendations, as well as in an
American consensus statement, control or remission of seizures, measured as the “seizure-free interval” is the main
determinant in the assessment of the ability to drive (Sonnen 1997, Krumholz 1994) and it will be the subject of a
large part of this report.
Table 1
European Council Directive 91/439/EEC of 29 July 1991 on driving licences
Official Journal L 237 , 24/08/1991
Group I Group II
A licence may be issued or renewed subject to an Driving licences shall not be issued to or
examination by a competent medical authority and to renewed for applicants or drivers
regular medical check-ups. The authority shall decide on suffering or liable to suffer from epileptic
the state of the epilepsy or other disturbances of seizures or other sudden disturbances of
consciousness, its clinical form and progress (no seizure in the state of consciousness.
the last two years, for example), the treatment received and
the results thereof.
European member states have to stay within a Council directive: they can be more restrictive, but not more liberal.
(table 1: see references. Group 1 refers to categories A: motorbike and B: car. Group 2 refers to categories C: lorry
and D: bus. – for a full explanation see European commission transport internet site)
After the 1995 / 1996 workshops, national legislation was adapted to an important degree in several European
countries, but remained unchanged in others. This situation led to a renewed call for harmonisation and the
installation of several medical advisory boards like the working group on epilepsy and driving.
The purpose of this report is to give an overview of current knowledge of the subject of epilepsy and driving and to
give regulations for implementation in European law.
It has to be stressed that rigorous scientific proof is not always sufficiently available for the decisions that have to be
taken with regard to epilepsy and driving. In such cases, the best available evidence and reasonable estimates are
used.
The recommendations of the 1996 European working group stated: rules must be as liberal as possible, simple and
clear (Sonnen 1997).
They should also be based on calculated risk.
More liberal rules may persuade people with seizures to undergo an assessment and stick to the rules for several
reasons:
-they may accept the rules as reasonable
-they have the perspective of getting their licence back
-they feel relieved of the responsibility and the uneasiness of doing something that may endanger other people,
including their relatives.
-they can drive legally and have an insurance
Shorter seizure-free periods will also increase the reporting of seizures to their physician.
One needs to realise that there is a relationship between the social expectation or need to drive and the number of
people with active epilepsy that drive illegally. This has been shown in the study of Berg (Berg et al 2000): seasonal
or irregular employment increases the chance of driving illegally (as does being male, young and having a licence).
There is likely an inverse relationship with the availability of public transport. The number of experienced seizures
is also likely to influence compliance with the rules.
Compliance can only be increased if we can give an explanation of the risk-increase in terms that are understandable
and convincing for the patient.
5
Final report; 3 April 2005 6
2 Statistical studies
Another approach is to try and look at statistics about the possible impact of epilepsy on road safety.
If one looks at the increase of risk for the population at large, epilepsy-related accidents constitute a small minority:
-Only one hospital admission after a traffic accident in 250 has an associated medical factor. Of these, 37% was
caused by epilepsy in the study by Taylor et al. (1995). This amounts to one in 675 hospital admissions, or 0.15% of
serious accidents.
-Epilepsy-related accidents constitute an estimated 0.25% of all traffic accidents according to Parsonage. (Parsonage
1992). This is 1 in 400 accidents.
Only 11% of all accidents among individuals identified with epilepsy are reported as due to seizures. (Krumholz A et
al. 1991)
First seizures (unavoidable) constitute on average 15% (Sonnen 1995) of seizure-related accidents.
In the Australian guidelines it is stated that epilepsy caused only 9 to 19 traffic accidents in 1991,being 0.025-
0.053% of all traffic accidents in that year (Austroads), which is a factor 10 lower. These were medical causes
collected via police reports, probably under-reported (Black 1996).
A population based study in the USA gave a number of 86 deaths per annum (total number per annum is 44,027;
population about 275 million), which is 0.2% of all traffic deaths; 4.2% of deaths associated with medical conditions
(Sheth SG 2004: see table 2)
An European data bank (table 3) gives data about traffic accidents for the year 2001: 1,248,896 road accidents and
38,828 deaths. This is 31.2 fatalities per 1000 road accidents involving personal injury. In addition, 18% of all
accidents are considered serious. (European internet site) These are figures about all traffic accidents. Cars seem to
have a somewhat smaller death ratio: 22 per 1000 car accidents. The contribution of epilepsy to these figures is not
known.
6
Final report; 3 April 2005 7
Table 3 http://europa.eu.int/comm/energy-transport-safety/
Europe 2001 Europe 2000
A Road accidents, involving 1,248,896 1,295,600
personal injury
B Per 1000 population 3.4 (1.4-5.2)
C Victims
D Slightly injured
E Seriously injured 159.9 per 1000 of A
F Road fatalities 31.2 per 1000 of A 40,812 (31.5 per 1000
of A)
G Serious + fatalities in % 201.1 per 10000 of A
H Fatalities in car accidents 22.0 per 1000
NOTE
In recent studies the accident rate for people with epilepsy does not seem exclusively attributable to seizures. In the
RESt-1 Group European cohort study (Van den Broek M and Beghi E 2004) accidents were investigated by type and
circumstances. The risks for street accidents (many of them traffic accidents) were 5% at 12 months and 7% at 24
months (controls 2% and 4%; p<0.001) After exclusion of seizure- related events these figures decreased to 4% and
6% (p<0.05). Part of this increased risk might be due to the effect of medication. This increase in risk is not easily
quantifiable and subject to individual assessment. In the following, we will deal with the risk increase due to seizures
while driving.
SO THE NUMBERS ARE LOW, BUT ARE THEY INCREASED IN PEOPLE WITH EPILEPSY?
Two questions remain.
Is the number of accidents increased in people with epilepsy?
Do seizure-related accidents caused by persons with epilepsy more often lead to serious injury?
CONCLUSION
The highest accident rate ratio found in overviews are 1.4 for serious accidents and 1.84 for all accidents in people
with epilepsy compared to the general population. We will mainly look at serious or fatal accidents in this report and
will use an accident rate ratio of 1.8 (!)for our risk calculations: a maximum estimate.
1 A = Attributable risk.
This is the increase in risk for the population, because there is a group of drivers that has an increased risk of an
accident. The magnitude of the risk depends on the number of people in the group and the risk increase per
individual in this group.
The formula that gives the relationship between the individual risk increase "R" and the risk for the population is:
A = P(R-1) formula (1)
7
Final report; 3 April 2005 8
A quantitative approach to risk assessment as far as a sudden incapacity by a seizure is concerned requires a decision
about the maximum acceptable accident rate for a person with epilepsy, on top of the risk for the population.
If no increase in overall risk is acceptable, any increase in risk because of epilepsy would have to be counterbalanced
by a diminished risk elsewhere, be it a medical or a non- medical risk. In this report we advise to accept a small
increase in risk for the population.
We propose to accept: A = 1% (The same percentage that was proposed by Sonnen in 1996: see ANNEX4)
Then the individual risk ratio of an accident R can be 4.3 (times, compared to someone without epilepsy)
If the percentage of people with epilepsy that drove would rise to a theoretical maximum of 70% (the percentage that
will become seizure-free) (P) will become 0.42 and the corresponding R will be: 3.4.
Conclusion: the relative risk can be between 3 and 4.3 when we accept that A can be 1%
2 R = Relative risk
This is the risk of the group with the characteristic under consideration (epilepsy) compared to the group without this
characteristic. (Ref: Spencer MB 2001)
To make a choice about which relative risk can be acceptable to us we have to look at other risks that occur in the
general population.
Age>75 3.1
Age>70 2
Female<25 3.2
Male<25 7
0 1 2 3 4 5 6 7 8
Data from the website of the Belgian Traffic Bureau (BIVV) and the "IMMORTAL" project, a study funded by the
EU 2004
A number of common or unavoidable situations are mentioned in figure 1. The highest estimated risk for a driver
with epilepsy is lower than most of these: 1.84.
8
Final report; 3 April 2005 9
Figure 1 shows some unavoidable variables that occur in the general population. These refer in part to Belgium, in
part to Europe. The fact that young males or elderly people have got an increased accident rate is accepted as
unavoidable and, in the elderly are due to an altered physical (and/or mental) functioning. It seems only fair to
accept an increase in accident rate ratio for people with epilepsy that is comparable to these ratios, especially in view
of the fact that the social consequences of not being able to drive for these people are often very important (loss of
job, social contacts, disclosure of the epilepsy etc.).
Conclusion:
An acceptable individual accident rate ratio could be 3 to 6
Remark:
It seems more acceptable for the patient to base regulations on the "R", rather than on the "A" type of risk. It would
also be easier to defend, so it might lead to better compliance with the rules.
At the same time, it is more fair: if we define "A" for a certain disease or for all medical causes, a disease that is
prevalent like epilepsy or diabetes would have a resulting low "R"-value. Considering the small contribution of
medical causes to accident statistics, the ultimate effect for the population will be low anyway, but the consequences
for the patients are often important.
THE MEASURING STANDARD: THE "COSY" AND THE CHANCE OF CAUSING AN ACCIDENT
In the quantification of risk several factors have to be taken into account.
The additional risk that a driver with epilepsy runs (in comparison to those without epilepsy) so the risk of having an
accident is mainly determined by four factors.
The likely outcome of an accident. This is generally measured in terms of serious injuries or fatalities. Less data
seem to exist about material damage or danger for the environment. Statistics give the following numbers.
To stay at the safe side (ANNEX 4), we will assume that of all accidents there is:
21% chance of a serious accident
3% chance of a fatality (included in 21%)
0.6% of the population has epilepsy
60% of people with epilepsy have a driving licence.
All these values are "negative" estimates.
9
Final report; 3 April 2005 10
The relation between the individual risk increase "R"and the COSY
If:
r = COSY (in %)/ 100
COSY means: the Chance of an Occurrence of a Seizure in the next Year, so the expected per annum seizure rate.
This is often expressed as a percentage. Here it will be expressed as a ratio:
R = individual risk ratio compared to someone who does not fulfil the criterion (has no epilepsy)
F = the present fatal casualty rate per driver per year (in fact statistics give per car) It is estimated less than 1 in 7000
drivers (Annex 4 )
D = proportion of time spent at the wheel. For group 1 this is 4.2% (one hour) or 0.042
X = probability of a seizure at the wheel leading to a fatality. The probability of a seizure at the wheel leading to an
accident is 0.6 The chance of 3% of having a fatality can be multiplied with this factor: X is 0.6 x 0.03 = 0.018
The relation between R and r is given by formula (2)
If for group 1:
F=0.00014; D=0.042 X= 0.018Then:
DX is 0.00076 F/DX is: 0.183
And:
r=(R-1). 0.183
Conclusion
This implies that even in the worst case scenario if we accept an R = 3, as a value for the individual risk increase, the
COSY can be as high as 37% while the risk for the population remains under 1. If one takes 20% COSY as the limit,
R = 2.1 and A < 0.5%.in the worst case. As argued above, we do not advocate this last choice.
RECOMMENDATION 1
It is recommended that risk assessment shall be based on the risk for the individual ("R": relative risk)
rather than on the risk for the population. For group 1, a relative risk of about 3 is acceptable. This implies
that a COSY between 20% and 40% is acceptable.
The European workshop of 1996 accepted an arbitrary factor of 5 for the severity of an accident if caused by a
heavy-goods vehicle compared to a private car (Sonnen in: Commission on Epilepsy, Risks and Insurance of the IBE
1994). A professional driver typically spends up to 8 hours per working day behind the wheel, which is 20% of his
lifetime - about six times as long as car drivers.
From these approximations, it was taken that the maximum COSY for group 2 would have to be 30 times less. This
resulted in an acceptable chance of a seizure in the next year of 2% (60% as a COSY for group 1 divided by 30 is 2%
COSY for group 2).
This same percentage is used in the American consensus statement and in the official Australian Guidelines (AAN et
al.1994; Austroads Incorporated 2003).
The European workgroup recognised the differences in risk for the respective categories of vehicles for which a
group 2 assessment is required, but “for the sake of simplicity” the same rate for all was put forward.
10
Final report; 3 April 2005 11
Sonnen in his report (Sonnen 1997) calculated the risk for the different vehicles. He attributed factors of severity for
4 items:
1. driving time
2. toll of accidents
3. seizure/accident ratio
4. passenger transport.
We tried to find European statistics to base these factors on. For an extensive account, see ANNEX 3
Conclusion
The overall factor 20 applies to buses and HGV´s. So we have no reason to change the 2% rule. Some other
vehicles that, in some countries, drive under group 2 conditions do not need to be assessed as severely. (e.g. taxis)
In countries where the transport of people organised by the employer in a car is assessed according to the criteria of
group 2 the assessment should take account of the number of people transported and an arbitrary safety factor (3-5?).
Motorbikes should be assessed more severely than cars, because of a greater danger for the driver, or at least should
the patient be informed about this. The danger to the driver is 2 to 3 times increased compared to a car.
RECOMMENDATION 2:
If we accept 40% or even 33% for group 1, the 2% rule (COSY should be <2%) can still apply for group 2.
Car driving used for professional purposes carries in general a lower risk, as does taxi driving and the like.
The COSY can be accordingly higher.
We propose to accept a ratio of 2 to 3 times. The most important groups however will have a risk increase no greater
than R = 2 (see the next chapter). Fatal accidents were 3 % of all accidents in 2001. If this last number was increased
up to 2 times it would become 6% for people with epilepsy. The number of injury accidents per year is about 3.4 per
1000 inhabitants, so 3,400 per 1 million population (this figure pertains to all accidents, not just vehicle or car
accident!). Of these 63% are car accidents this is 2,142 per million. If 0.6% of the population has epilepsy and 70%
of patients qualify for a drivers licence, 0.42% of the driving population could have epilepsy. These people would
produce a proportional number of fatalities: 9. If we accept an accident rate ratio of 2, they will produce 18 fatalities.
11
Final report; 3 April 2005 12
So for car accidents the extra number of fatalities because of epilepsy is about 9 fatalities of a total of 3,400 per
million !
CONCLUSION
The risk for the population is acceptable if people comply to the rules.
What would happen if these people would all continue to drive their car till the second seizure (and then all
stop driving)?
The number that recurs within a year is the average total recurrence (49%) times the percentage that has a second
seizure in the first year (68%) (see below) = 33.32%. In 6 months this would be (53%) 25.97% and in 3 months
(32%) 15.68%.
NOTE: for more explanation about these figures see page the next chapter: first seizure.
So of 700 people per 1,000,000 the number that will recur is 110 in the first 3 months, 182 in 6 months and 233 in
the first year. In the long run 49% of 700 will recur: 343 people.
These people have a chance of an accident of (60% times 1/24) 2.5%. So 6 people will produce an accident in the
first year, 9 in total.
Of the total number of accidents, 21% will produce a severely injured person including about 3% fatalities (so about
1 in 7 serious accidents leads to a fatality).
If we accept that the chance of having a serious accident or a fatality in people with epilepsy is twice that of the
population (in most situations it will be less) these figures would be 42% and 6% respectively, but the increase is
21% and 3%. So 0.53% (21% x 2.5%) of people with a second epileptic seizure will have a serious accident because
of this seizure: 1.8 people per 1 million population, of which 0.26 fatalities. In the first 6 months the number is 1..
In the first year it is 1.2 per million.
If these 700 people would all be allowed to continue driving (without restraint in time period) and if they all stop
driving after the next (second) seizure this would cost 1.8 serious accidents per million population and 1 fatality per 4
million population.
These might be avoidable but many people will not be compliant to the rules, so the effect will be less, but the social
consequence will be more: more people will drive illegally and possibly without insurance.
CONCLUSION
Here especially, the effect on the level of the population is small. The social and economic consequences are bound
to be high in this group that has not (yet) adjusted to their new status. The difference between 3 months or 6 months
driving ban for first-seizure patients is 0.7 serious accidents and 0.1 life per million population.
12
Final report; 3 April 2005 13
Table 5: Data from the Netherlands suggest that driving a car is one of the least dangerous forms of road use.
The Netherlands 2001:victims per 1 billion driving kilometres per vehicle type
Victims Ratio / car
Car (driver) 4.57 1.00
Car (passenger) 3.22 0.70
Moped 86.68 18.97
Light moped ("snorfiets") 116.8 25.55
Bicycle 15.76 3.45
Pedestrian 33.21 7.27
If people cannot drive their car, they very often have to use another means of traffic, like a moped, that, in some
countries, does not require a drivers licence. These are all 3-25 times more dangerous per kilometre driven. What to
advise them?
RECOMMENDATION 3:
The dangers for the population are low if people continue driving in a car after a first seizure and the effect
on the population level of not letting them drive may not outweigh the socio-economic disadvantages. The
increased risk is often acceptable to the patient. The alternatives of car driving: are often more dangerous
for the patient. This is a group that has not accepted (yet) their disease and resulting incapacity.
Compliance is bound to be low. After a neurological workup, restricted driving without passengers might
be a solution for a prognostically favourable group.
13
Final report; 3 April 2005 14
70
60
50
40
37.4
30
20
7.8
3.4
10
2.6 1.3
0
Legend of fig 2
Elwes 1985 0
Stroink 1998 0
FIRST 1993 untreated 0
Shinnar 2000 / 96 14
Hopkins 1998 15
NGPSE 1990 15
Van Donselaar 1991 15
Hirtz 1984 non-provoked 27
Camfield 1989 30
Annegers 1986 61
Camfield 1985 68
Hauser 1990 / 82 80
FIRST 1993 treated 80
Average % recurrence
14
Final report; 3 April 2005 15
From this, a reasonable estimate of the recurrence rate after a first seizure in untreated patients would be 55%, in
treated patients 33%.
In the meta-analysis of Berg et al. (1991), the average percentage recurrence risk in carefully selected studies was
42% in 2 years (treated and untreated patients). In an overview of the literature, Beghi et al. (1997) finds a range of
25-52% with an average of 38% in 2 years. Assuming a 87% recurrence in the first 2 years these averages imply
about 49% and 44% of total recurrence risk respectively
The second question is the behaviour of the a priori recurrence rate over time in the group where there were
recurrences. We found eight studies with a follow-up of longer than 2 years (in: Berg et al 1991; and: Hart et al.
1990; Sander et al. 1990 ) and two studies providing data about recurrences in the first 3 months and 6 months after a
first seizure (leaving the Hirtz study out, because it provides data about very young children)( FIRST 1993; Annegers
et al. 1986). Berg et al. (1991) in their meta-analysis calculated the percentage recurrence risk as a percentage of the
total risk after 4 years in the studies that provided data over a longer period and found an average of 87%. We
recalculated these data with the addition of later studies (Hart et al. 1990; Shinnar et al. 2000; Hui et al. 2000) and
found a very similar number: 86,5% after 2 years. A recurrence risk in the 3rd, 4th and 5th year after a first seizure
can be estimated from their data: respectively 8%, 5% and 4%.
80
70
60
50 32
Percentage
40
30
21
20
10 15 19
8 5
0
1y 2y 3y 4y
15
Final report; 3 April 2005 16
in which these recurrences occurred. Hauser (Hauser A in Jallon 2003) states: “Relapse was about 1% per year for
the first 10 year after fulfilling remission criteria…” (5 years remission with or without medication)
Table 7
Recurrence risk as a percentage of the total
recurrence over the observation period
After a number of seizures
1 2 3 or more
After the first 3 32 44% 41%
months
At 6 months 56 56% 63%
At 12 months 68 78% 80%
At 24 months 87 83,5% 88%
16
Final report; 3 April 2005 17
An interesting study is the one of Krauss (Krauss et al 1999) They compared 50 patients with a seizure-related
accident with 50 patients without an accident and found that the period of seizure freedom has a strong influence on
the chance of an accident. If everybody had kept a seizure-free period of 12 months, 93% of the accidents would
have been avoided. This percentage was 85% for a seizure-free period of 6 months. For 3 months it was not
significant.
Figure 4: Remaining percentage recurrence risk (RRR) for six different situations at 4 seizure-free intervals (S =
seizures)
60,0
50,0
40,0
Remaining % risk
R = acceptable
30,0
20,0
10,0
0,0
3 mo 6 mo 1y 2y
Many S 95% 3S Hauser 76%
1S Untreated 55% 1S Beghi 44%
1S Treated 33% 1S Idiopathic 25%
RECOMMENDATION 4
It is recommended that for every driving assessment an estimate of the COSY of the relevant situation is
determined according to the described principles. This result can be compared with the acceptable R as
described in chapter 2.
17
Final report; 3 April 2005 18
UNFAVOURABLE FACTORS
For the assessment of drivers in group 2, extra care is needed in drivers of buses and drivers that transport dangerous
goods.
Motor cycle drivers are likely to have an accident with most seizures while driving (not just a 50% or 60% risk).
Moreover there are 2 to 3 times as many seriously injured + fatalities per vehicle in motorcycles compared to cars.
(Belgian data 2001: cars 12.7; motorbikes 33.0 per 10,000 vehicles). The ratio serious accidents/all accidents, is 15
for cars and 27 for motorbikes.(Belgium 2001: BIVV website) Here the risk increases accordingly.
The RESt-1 group found an increased relative risk for non-seizure related events of 1.4 in patients without seizures.
(Beghi 2002) It was argued that this could be an effect of medication.
FAVOURABLE FACTORS
Favourable mitigating factors will decrease the chance of recurrence or the chance of an accident.
- Provoked seizures
- Seizures during medically directed medication changes
- Seizures that do not interfere with consciousness or motor control
- Established pattern of pure nocturnal seizures.
Could be added:
- Treatment of the first seizure
Treatment of the first seizure does on average improve the recurrence risk (FIRST study 1993), but only for
the first 2 years.
For most of these favourable modifiers, specific articles are part of the medical criteria in some countries. They are
discussed in the above chapter. Others that are mentioned in the American consensus statement (AAN et al. 1994) or
18
Final report; 3 April 2005 19
by Beaussart (1994) are not. Seizures after sleep deprivation are not considered avoidable in most instances. An
exception could be seizures after excessive sleep deprivation. Prolonged and consistent auras are proven not to be
safe enough (Krauss et al. 1999).
Drachman (Drachman 1999) gives a list of questions to which the physician should provide an answer:
1. The cause of the episode (type of disorder) suffered
2. The means by which the condition is controlled (including medications and dosages)
3. The degree of impairment or disability suffered during an episode (extent of the episode)
4. The probability of recurrence of the episode (including frequency of recurrence, degree of assurance that the
event will not reoccur, and the basis of estimate of this probability)
5. The date of the most recent episode
6. Certification, to a reasonable degree of medical certainty, that the individual's medical condition and medication
will not interfere with the safe operation of a motor vehicle
EEG
First seizure. The EEG is part of the diagnosis of epilepsy and therefore a necessary investigation after the
occurrence of a first seizure. It also has a prognostic significance, as will be discussed below, in the case of a first
seizure without detectable cause and normal EEGs.
Epilepsy and the prediction of seizure-free episodes. Limited prognostic value in this situation.
Exclusively seizures in sleep. The EEG can be useful in the diagnosis of generalised versus partial epilepsy, which is
a prognostic factor in the study of Park et al.(see below).
Provoked seizures. The occurrence of epileptic activity after the acute period is a logical argument against the
diagnosis of "provoked seizure"
Stopping medication: here Berg (BERG A et al 1997) found 15 studies. The range in relative risk for an abnormal
EEG just before discontinuing medication was 0.82 to 6.47 with a typical estimate of 1.45.
Suspicion of "subclinical" seizures - mainly in some primary generalised epilepsy syndromes. Testing with the aid
of video-EEG monitoring is advisable in this situation to determine the influence of epileptiform discharges on
consciousness and cognitive function.
Medication
In the study of Taylor no increased risk was found in people taking AED (Odds ratio: 0.97 CI: 0.87-1.07). (TAYLOR
1996)
In the RESt-1 Group European cohort study (Van den Broek M and Beghi E 2004) accidents were investigated by
type and circumstances. After exclusion of seizure- related events the risks decreased to 4% and 6% (p<0.05) but
remained higher than the controls (2% and 4%; p<0.001). Part of this increased risk might be due to the effect of
medication. A recent Belgian review classified drugs in groups according to their possible influence on driving.
More detailed data were not found. Phenobarbitone, Primidone and Hydantoine were classified in group 3 ("probably
a serious negative influence is possible"); All the other antiepileptic drugs in class 2 (probably a light (Valproate) or
moderate (Carbamazepine; benzodiazepines) negative influence is possible). For the newer antiepileptic drugs
influence was uncertain because of lack of data. (THE TOXICOLOGICAL SOCIETY OF BELGIUM AND LUXEMBOURG
"Invloed van geneesmiddelen op de rijvaardigheid" Literature-study for the Belgian Institute of Traffic Safety
(BIVV) 1999. Ed: BIVV, Haachtsesteenweg 1405 Brussels 1130 Belgium. The study is available in French). In
clinical practice the main concern is during the period of drug initiation. Cognitive side effects vary, but are
generally less in the newer AEDs.
A different question is the use of AEDs as a treatment of the first seizure or in the treatment of epilepsy. See below.
Driving time
If we accept the calculations based on driving time, a licence that restricts the time behind the wheel seems a
alternative that is worth contemplating in certain situations. See "Limited licence"
RECOMMENDATION 5
There are many factors at play in the ability to drive that cannot easily be quantified. An individual
assessment by a neurologist is recommended for every patient that has had one or more seizures.
19
Final report; 3 April 2005 20
20
Final report; 3 April 2005 21
5. Legal issues
It is preferred that the criteria should appear in guidelines rather than in the law. The law cannot provide more than a
framework of minimum criteria. Each person has to be assessed individually.
LEGAL OBLIGATIONS
Should the licence holder or applicant inform the authorities of any relevant disease/change in his/her health?
This should be imposed by law, both while applying for a licence and in the case of any relevant change in health
status of a holder of a licence.
SHOULD TAXI DRIVERS ETC. BE INCLUDED IN GROUP 2, AND IF SO, SHOULD THE RISK ASSESSMENT BE EQUALLY
SEVERE?
The medical assessment for group 2 in some countries is also applicable to the transport of people in a broader sense,
notably taxi drivers; rental services with driver; public transport; drivers for school transport; transport of people if
this is organised and run by the employer. The decision to do so is a national competence. The reason to do so
might be to give an additional assurance that the driver is fit to drive in the case of the transport of people (required
risk), or might be because of a perceived increase in risk because of an increase in driving distance etc. (calculated
risk). The difference between required risk and calculated risk for these groups reflects the additional "safety
margin". Some quantification of risk for taxi drivers is attempted in this report.
It is clear, however, that the risk for these group 1 vehicles is by far not as high as for heavy goods vehicles or buses.
These drivers should be assessed accordingly.
21
Final report; 3 April 2005 22
RECOMMENDATION 6
It is the responsibility, not of the doctor, but of the patient whether he drives or not and it is his duty to
report to the authorities. This is inherent part of the proposals in this report and should be part of European
and national law. Mandatory reporting by the physician is recognised as working against road safety
because it discourages the declaration of symptoms by the patient.
The assessing physician should be legally protected as regards his advice about driving ability and if he
reports or does not report the non-compliant patient to the authorities.
22
Final report; 3 April 2005 23
GENERAL REMARK
The recommendations here proposed are meant to serve as a base for implementation in European law. The Working
Group has taken the position that, in view of the different levels of scientific evidence, the recommendations are
minimal medical criteria for safe driving. Sometimes it might be advisable to be more severe than the
recommendations. In that case the opinion of the Working Group is expressed elsewhere in the report.
DEFINITION OF EPILEPSY
For the purpose of these regulations epilepsy is defined as having had 2 or more epileptic seizures, less than 5 years
apart.
GENERAL RULES
Drivers, assessed under group 1 with epilepsy should be under review till they have been seizure- free for at least 5
years.
For group 2 this period will be dependent on the clinical situation and will be longest for the person with a diagnosis
of epilepsy, i.e. until 10 years freedom from seizures has been demonstrated without the aids of anti-epileptic drugs
A person who has an initial or isolated seizure or loss of consciousness should be advised not to drive
A specialist report is required, stating the period of driving prohibition and the requested follow-up
If the person has epilepsy, the criteria for an unconditional licence are not met. The patient should notify the
Licensing Authority
It is extremely important that the patient's specific epilepsy syndrome and seizure type are identified so that an
adequate evaluation of the person's driving safety can be undertaken (including the risk of further seizures) and the
appropriate therapy instituted. This should be done by a neurologist.
On extensive neurological investigation no relevant cerebral pathology has been established and there is no
epileptiform activity on the EEG.
The subject can only be declared able to drive subject to neurological opinion.
23
Final report; 3 April 2005 24
A note of caution
Three questions are of major importance
1. Is the event epileptic vs. non-epileptic?
2. Is the event a first epileptic seizure or a recurrence: in using statistics for recurrence after a first seizure it is
presumed that it really was the first one. If more than one seizure has occurred other statistics apply that are
less favourable. More than 50% of patients that first come to a neurologist will have had more than one
seizure! One should specifically ask for the occurrence of isolated auras, absences and myoclonic jerks
(Wolf P 1997).
3. Is the seizure provoked or unprovoked?
For each of these questions a number of anamnestic and hetero-anamnestic data should be collected.
The applicant who has had a first unprovoked epileptic seizure can be declared able to drive after a period without
seizures of six months, if there has been an appropriate medical assessment.
National authorities may allow drivers with recognised good prognostic indicators to drive sooner.
National authorities may allow drivers with recognised good prognostic indicators to drive sooner.
For group 2, some situations that are usually considered provoked seem to have a recurrence risk that is too high
Post-traumatic seizures
The recurrence risk after early post-traumatic seizures is 25-60% according to Jennett (Jennett 1975). Late seizures
occur 50% - 60% in the first year; 85% in the first 2 years (Caveness 1979). After the first late seizure the recurrence
risk is 86% in 2 years (Haltiner et al. 1997).
Cranial trauma is considered serious if one of the following characteristics is present.
In a population-based study (Annegers 1998) risk factors for late post traumatic seizures were: brain contusion,
subdural haematoma; skull fracture; loss of consciousness or amnesia of more than one day and age above 65. The
standardised incidence ratio was 1.5 after mild injury, 2.9 after moderate injury and 17.0 (CI12.3-23.6) after severe
injury. According to the criteria of Jennett (Jennett 1995 and 1997) head injury is serious if there is an acute
intracerebral haematoma, required surgery, compound depressed fracture or dural tear with more than 24 hours post-
traumatic amnesia. The presence of one of these factors is a reason for assessment.
24
Final report; 3 April 2005 25
The driving authority should consider whether these patients require a practical driving assessment. Very often they
have psychological or cognitive alterations, which are misjudged by themselves and their physician (Hawley 2001).
These two causes (serious trauma and encephalitis) were excluded as acceptable provoking factors in the case of
early seizures for drivers of group II in a Belgian consensus. Driving is consider under group 2 criteria after serious
traumatic cerebral injury only if the following prerequisites have been fulfilled:
- no early seizures
- after individual assessment of the seriousness of the trauma
- after a seizure-free period of 2 years
For some other diseases high risk have been stated: arteriovenous malformation 30-45%; intracerebral haemorrhage
25% with a follow-up of 4.6 years (half of these occurred in the first 24 h); in another series 32%; subarachnoid
haemorrhage 27.8% (Fisher 2001). Most of these figures are not usable as such to determine a COSY.
Other causes
Causes that are not considered sufficiently avoidable in general are: sleep deprivation unless exceptional and
excessive; stress.
For alcohol and alcohol withdrawal seizures and seizures in association with the use of illicit drugs a controlled
period of abstinence is needed. A seizure after a night of binge-drinking might be more avoidable than a withdrawal
seizure, because the latter occurs in general in chronic alcoholics.
An EEG and an appropriate neurological assessment should be performed after the acute episode.
Someone with a structural intracerebral lesion who has increased risk of seizures should not be able to drive vehicles
of group 2 until the epilepsy risk has fallen to 2% per annum or less.
PROPHYLACTIC BAN
As discussed above, certain disorders have a high risk of epileptic seizures, even if early seizures have not occurred.
As to whether this situation exists for group 1 is debatable, but for group 2 some recurrence risks are clearly too high,
even if these are given as total recurrence risk over a longer time period and the COSY is not exactly known.
25
Final report; 3 April 2005 26
Certain disorders have an increased risk of seizures, even if seizures have not yet occurred. In such a situation an
assessment should be done: the risk of having a seizure should be 2% per annum or less
SEIZURES WITHOUT INFLUENCE ON CONSCIOUSNESS OR ABILITY TO ACT AND WITHOUT (EVER) HAVING (HAD)
ANY OTHER KIND OF SEIZURE
Some seizures are not considered to be of influence on driving ability, mainly some myoclonias and simple partial
seizures. Evidence about the (non-) harmfulness of these is lacking. In the first European Committee, there was no
consensus about this. In the experience of the authors this situation is rare.
26
Final report; 3 April 2005 27
CHILDHOOD SEIZURES
BREAK-THROUGH SEIZURES
27
Final report; 3 April 2005 28
EPILEPSY
EXCEPTIONS.
If the period between the last seizure and penultimate seizure is more than 5 years, the last seizure may be considered
in a similar fashion to a first unprovoked seizure for licensing purposes, subject to neurological opinion.
The applicant or driver who has had seizures exclusively during sleep or seizures which have been demonstrated
exclusively to affect neither consciousness nor cause any functional impairment can be declared fit to drive so long
as this pattern has been established for a period which must not be less than the seizure free period required for
epilepsy.
National authorities may allow drivers with recognised good prognostic indicators to drive sooner.
Group 1
It was decided that after the period of driving prohibition ("safety period"), the licence should be reviewed after one
year, then 5 years. For seizures that do not influence driving ability or seizures that occur exclusively in sleep, the
review will be annually for the first 5 years. After these periods no specific assessment for epilepsy was considered
necessary.
28
Final report; 3 April 2005 29
Safety periods and periods of driving validity for both groups: the first 10
years
First S
Epilepsy
Epilepsy- surgery
No influence
Sleep SS
Therapy change S
Provoked S
Sporadic SS
Provoked S (Gr 2)
First S (Gr 2)
Epilepsy (Gr 2)
20 years
Group 2
The applicant for group 2 will be regularly checked during the required periods of freedom. It is not recommended
that there should be a regular check-up once he has received his drivers licence.
29
Final report; 3 April 2005 30
30
Final report; 3 April 2005 31
A person who has an initial or isolated seizure or loss of consciousness should be advised not to drive
A specialist report is required, stating the period of driving prohibition and the requested follow-up
Drivers, assessed under group 1 with epilepsy should be under licence review till they have been seizure- free for at least 5
years.
If the person has epilepsy, the criteria for an unconditional licence are not met. The patient should notify the Licensing Authority
It is extremely important that the patient's specific epilepsy syndrome and seizure type are identified so that an adequate
evaluation of the person's driving safety can be undertaken (including the risk of further seizures) and the appropriate therapy
instituted. This should be done by a neurologist.
Drivers or applicants can be declared fit to drive after a 1-year period free of further
- Epilepsy
seizures.
Special situations
If the period between the last seizure and penultimate seizure is more than 5 years,
- Sporadic seizures the last seizure may be considered in a similar fashion to a first unprovoked seizure
for licensing purposes, subject to neurological opinion.
31
Final report; 3 April 2005 32
General conditions for all Group 2 drivers On extensive neurological investigation no relevant cerebral pathology has
been established and there is no epileptiform activity on the EEG.
The applicant who has had a first unprovoked epileptic seizure can be
declared able to drive once 5 years freedom of further seizures has been
achieved without the aid of anti-epileptic drugs, if there has been an
First unprovoked seizure appropriate neurological assessment
National authorities may allow drivers with recognised good prognostic
indicators to drive sooner.
10 years freedom of further seizures has been achieved without the aid of
anti-epileptic drugs.
Epilepsy
National authorities may allow drivers with recognised good prognostic
indicators to drive sooner.
32
Final report; 3 April 2005 33
GROUP 1
Does the syndrome or seizure type have influence on the policy after a first seizure?
There are no data that would indicate a different approach for different seizure types. The same holds true for
epilepsy syndromes, at least as far as the short term prognosis is concerned.
LIMITED LICENCE?
The existing list (for situations different from epilepsy):
-not during the night
-never alone
-not with passengers
-without trailer
-limited radius
-limited speed
-not on motorways
-without any alcohol
A limitation in time was considered, but it was thought impossible to control and therefore not practical to
implement.
33
Final report; 3 April 2005 34
GROUP 2
Prophylactic ban
Do we need a list of situations where, apart from the functional consequences of the underlying disease, the risk of
the de novo seizures is increased to a level that group 2 driving should be advised against?
Sonnen 1997 page 117
Only for trauma there are sufficient data. For the other situations only overall risks are given. According to Doelman
(in Sonnen 1997) brain abscess: best permanent ban.
Surgically treated AVM might also be a reason for an exclusion.
For the recommendations mentioning of the problem seems best. For guidelines a thorough study is needed.
Ideally, the collected data could have the following format.
Percentages of recurrence risk are given for a certain disorder for four different situations.
34
Final report; 3 April 2005 35
9 References
AAN, AES AND EFA Consensus statements, sample statutory provisions and model regulations regarding driver
licensing and epilepsy. Epilepsia 1994;35:696-705
ANNEGERS JF, HAUSER WA et.al. The risk of unprovoked seizures after encephalitis and meningitis Neurology
1988;38(9):1407-10
ANNEGERS JF HAUSER WA COAN SP AND ROCCA WA A population-based study of seizures after traumatic brain
injuries NEJM 1998;338(1):20-24
ANNEGERS JF, SHIRTS SB, HAUSER WA, KURLAND LT Risk of recurrence after an initial unprovoked seizure.
Epilepsia 1986;27(1):43-50
ASSURALIA- INTERNET SITE: "Opsplitsing slachtoffers in 2001 tussen bestuurders en passagiers". http//:
www.assuralia.be/
AUSTROADS INCORPORATED 2003: Australian medical criteria. In: "Assessing fitness to drive 3rd ed." ISBN 0 85588
507 6
BEAUSSART M In: Epilepsy and Risk, a first-step evaluation. IBE 1994
BEGHI E, BERG A AND HAUSER A Treatment of single seizures. In: Engel J Jr et al :Epilepsy a comprehensive
textbook Lippincott-Raven 1997
BEGHI E CORNAGGIA C Morbidity and accidents in patients with epilepsy: results of a European cohort study
Epilepsia 2002;43(9):1076
BELGIAN TRAFFIC DATA 2001: http://www.statbel.fgov.be/figures/d37_nl.asp
BERG A & SHINNAR S The risk of seizure recurrence following a first unprovoked seizure: a quantitative review.
Neurology 1991;41:965-972
BERG AT, SHINNAR S Relapse following discontinuation of antiepileptic drugs: A meta-analysis. Neurology
1994;44:601
BERG A, SHINNAR S , CHADWICK D Discontinuing antiepileptic drugs in: Engel J Jr et al :Epilepsy a comprehensive
textbook Lippincott-Raven 1997
BERG A AND ENGEL J: Restricted driving for people with epilepsy Neurology 1999;52(7):1306-1307
BERG AT, VICKREY BG, SPERLING MR et al. Driving in adults with refractory localisation-related epilepsy: multi-
centre study of epilepsy surgery. Neurology 2000;54:625-630
BIVV INTERNET SITE: www.bivv.be/main/PublicatieMateriaal/Statistieken.shtml
BLACK AB, LAI NY Epilepsy and driving in South Australia - an assessment of compulsory notification Medicine
and Law 1997; 16:253-267
BOON P, DE DEYN PP, HAUMAN H, MOL L, SCHMEDDING E, VLIETINCK R, WILLAERT B.: Epidemiologie van
epileptische toevallen in Vlaanderen. Tijdschrift voor Geneeskunde 1996;52:47
BOULLOCHE I. et al. Risk of recurrence after a single unprovoked generalized tonic-clonic seizure. Dev Med Child
Neurol 1989;31:626-632
CAMFIELD PR et al Epilepsy after a first unprovoked seizure in childhood. Neurology 1985;35:1657-1660
CAMFIELD PA et al. A randomized study of carbamazepine versus no medication after a first unprovoked seizure in
childhood. Neurology 1989;39:851-852
CANADIAN CARDIOVASCULAR SOCIETY CONSENSUS CONFERENCE 2003 Assessment of the cardiac patient for fitness
to drive or fly: internet site:http://www.ccs.ca
COCKERELL O et al. Prognosis of Epilepsy: A Review and further analysis of the first nine years of the British
National General Practice Study of Epilepsy, a prospective population-based study. Epilepsia 1997, 38 (1): 31-46
COMMISSION ON EPIDEMIOLOGY AND PROGNOSIS, ILAE Guidelines for Epidemiologic Studies on Epilepsy Epilepsia
1993;34(4):592-596
COMMISSION ON EPILEPSY, RISKS AND INSURANCE OF THE IBE Epilepsy and Risks. A first-step evaluation. 1994:
IBE, PO box 21, 2100 AA Heemstede The Netherlands
COUNCIL DIRECTIVE 91/439/EEC of 29 July 1991 on driving licences. Official Journal L 237 , 24/08/1991 P. 0001 –
0024
D´ALESSANDRO RD et al.. Pure sleep epilepsies: prognostic features. In: Epilepsy, an update on research and therapy
Alan R Liss Inc, New York, 1983; 235-239
DRACHMAN D Risk factors of seizure-related motor vehicle crashes in patient with epilepsy Neurology
1999;53(9):2214-2215
DRAKOWSKI JF, FISHER RS, SIRVEN JI, ET AL. Seizure-related motor vehicle crashes in Arizona before and after
reducing the driving restrictions from 12 to 3 months. Mayo Clin Proc 2003;78:819-825
EGLI M, HARTMANN H, HESS R Driving licences in epileptic patients. Schweiz Med Wochenschr 1977;10/12:389-
397
ELWES R et al. Prognosis after first tonic-clonic seizure. Lancet 1985;2:752-753
ENGEL J A proposed diagnostic scheme for people with epileptic seizures and with epilepsy: report of the ILAE task
force on classification and terminology Epilepsia 2001;42(6):1-8
35
Final report; 3 April 2005 36
36
Final report; 3 April 2005 37
MA C & CHAN K Benign childhood epilepsy with centro-temporal spikes: a study of 50 Chinese children. Brain &
Development 25 (2003) 390-395
MARSHALLSC et.al. Restricted driver licensing for medical impairments: does it work? CMAJ 2002;167(7):747-51
MCINTOSH A, KALNIS R, et al Temporal lobectomy: long-term seizure outcome, late recurrence and risk for seizure
recurrence Brain 2004; 127: 2018-2030
MEDICAL RESEARCH COUNCIL ANTIEPILEPTIC DRUG WITHDRAWAL STUDY GROUP Randomised study of antiepileptic
drug withdrawal in patients in remission. Lancet 1991;337:1175-80
MUSICCO M, BEGHI E et al Treatment of first tonic-clonic seizure does not improve the prognosis of epilepsy
Neurology 1997;49:991-998
OJALA M AND OJALA I: translation in English of an article "Epilepsy and assessment of fitness to drive: Finnish
guidelines" in Suomen Lääkärilethi 2001; 49-50: 5121-5123
PARK Clinical courses of pure sleep epilepsy. Seizure 1998;7(5):369
PARK K, BUCHHALTER J et.al. Frequency and significance of acute postoperative seizures following epilepsy surgery
in children and adolescents Epilepsia 2002;43(8):874-81PARSONAGE M Epilepsy and driving licence regulations.
Report by the ILAE / IBE commission on drivers´licensing IBE/ ILAE 1992 PO Box 21,210 AA Heemstede, The
Netherlands
PORTER JR, MATTSON RH, CRAMER JA, DIAMOND I Alcohol and Seizures. Basic Mechanisms and Clinical Concepts
F.A. Davis Co Philadelphia 1990 ISBN 0-8036-7008-7
REPORT OF THE QUALITY STANDARDS SUBCOMMITTE OF THE AMERICAN ACADEMY OF NEUROLOGY Practice
Parameter: A guideline for discontinuing antiepileptic drugs in seizure-free patients - Summary Statement.
Neurology 1996;47:600-602
REST-1 GROUP Social aspects of epilepsy in the adult in seven European countries. Epilepsia 2000; 41(8):998-1004
SANDER JW, HART YM, JOHNSON AL, SHORVON SD National General Practice Study of Epilepsy: newly diagnosed
epileptic seizures in a general population. Lancet 1990;336:1267-71
SASIC et al. 2002 Epilepsia:43(S8);111
EUROPEAN COMMISSION Saving 20,000 lives on our roads 2003 ISBN 92-894-5893-3
SCHMEDDING E Personal inquiry in a group of 70 Belgian neurologists. Chantilly 1996
SCHMEDDING E on behalf of the "Belgian Working Group on Epilepsy and Driving" Epilepsy and Driving in
Belgium: proposals and justification Acta Neurologica Belgica 2004;104:68-79
SHETH SG, KRAUSS G, KRUMHOLZ A AND LI G Mortality in epilepsy: driving fatalities vs. other causes of death in
patients with epilepsy Neurology 2004;63:1002-1007
SHINNAR S, BERG AT et al. The risk of recurrence after a first unprovoked afebrile seizure in childhood: an extended
follow-up. Pediatrics 1996;98:216-225
SHINNAR S, BERG AT et al. Predictors of Multiple Seizures in a Cohort of Children Prospectively Followed from the
Time of Their First Unprovoked Seizure. Ann Neurol 2000;48:140-147
SONNEN AE AND THE EUROPEAN WORKING GROUP: Epilepsy and Driving Proceedings First European Workshop
epilepsy and Driving Licences Group 1. IBE May 1995
SONNEN AE Epilepsy and driving: A European View. Driving Commission, IBE 1997: 11-32
SPENCER MB The role of risk analysis in the evaluation of fitness to drive QinetiQ 2001
SPENCER SS Long-term outcome after epilepsy surgery. Epilepsia 1996; 37(9):807-813
STROINK H, BROUWER OF, ARTS WF, GEERTS AT, PETERS AC, VAN DONSELAAR CA The first unprovoked,
untreated seizure in childhood: a hospital based study of the accuracy of the diagnosis, rate of recurrence, and long
term outcome after recurrence. Dutch study of epilepsy in childhood. J Neurol Neurosurg Psychiatry 1998;64:595-
600
TAYLOR JF Medical aspects of Fitness to Drive. The Medical Commission on Accident Prevention 1995 35-43
Lincoln Inn Fields London WC2A 3PN
TAYLOR JF, CHADWICK D, JOHNSON “Risk of accidents in drivers with epilepsy”. J Neurol Neurosurg Psychiatry
1996; 60:621-27
TAYLOR DC, MCMACKIN D, STAUNTON H et al. Patients´aim for epilepsy surgery: desires beyond seizure freedom.
Epilepsia 2001;42(5) 629
THE TOXICOLOGICAL SOCIETY OF BELGIUM AND LUXEMBOURG "Invloed van geneesmiddelen op de rijvaardigheid"
Literature-study for the Belgian Institute of Traffic Safety (BIVV) 1999. Ed: BIVV, Haachtsesteenweg 1405 Brussel
1130 Belgium. The study is available in French
VAA T. Impairments, diseases, age and their relative risk of accident involvement: Results from meta-analysis 2003
Report to the Institute of Transport Economics, PO Box 6110 Etterstad, N- 0602 Oslo, Norway
VAN DEN BROEK M, BEGHI E FOT THE REST-1 GROUP Accidents in patients with epilepsy: types, circumstances and
complications: a European cohort study Epilepsia 2004;45(6):667-
VAN DONSELAAR Reliability of the diagnosis of a first seizure Neurology 1989;39:267
VAN DONSELAAR C, HABBEMA J Recurrence after first seizure (letter) Lancet 1991;337(jan5):46
VAN DONSELAAR C et al. Idiopathic first seizure in adult life: who should be treated? BMJ 1991;302:620-623
VAN DONSELAAR C, SCHIMSHEIMER RJ, GEERTS A, DECLERCK A Value of the Electroencephalogram in Adult
Patients With Untreated Idiopathic First Seizures. Arch. Neurol. 1992;49:231-237
37
Final report; 3 April 2005 38
VICTOR M in: PORTER JR, MATTSON RH, CRAMER JA, DIAMOND I Alcohol and Seizures. Basic Mechanisms and
Clinical Concepts F.A. Davis Co Philadelphia 1990 ISBN 0-8036-7008-7 Page 331
WOLF P Isolated seizures in: Engel J Jr et al: Epilepsy a comprehensive textbook Lippincott-Raven 1997
YANG SY, ZHAO CS Review of 140 patients with brain abscess Surg Neurol 1993; 39(4):290-6
38
Final report; 3 April 2005 39
10 ANNEXES
39
Final report; 3 April 2005 40
Sonnen in his report (Sonnen 1997) calculated the risk for the different vehicles. He attributed factors of severity for
4 items and by multiplication of these factors he arrived at a total factor of increased risk of a certain vehicle
compared to cars. There were factors attributed for:
1. driving time
2. toll of accidents
3. seizure/accident ratio
4. passenger transport.
Sonnen (Sonnen 1996) assumed a ratio of 30 between group 1 and group 2 in the severity of assessment. This did
not always seem justified. For taxis and ambulances and lorries the risk seems considerably lower than for buses.
Minibuses have an intermediate risk.
40
Final report; 3 April 2005 41
In some countries, like Belgium, transport of people organised by a company falls under group 2 assessment. This
often involves a driver and one or two other people. This situation carries a calculated risk like a car, not like a bus!
The only reason to be more severe is a (arbitrary) factor for extra-security because it is a professional risk to be a
passenger of such a driver. Five times more severe than a car driver seems reasonable. Other factors do not apply in
this situation.
We tried to find European (or national) statistics to base these factors on.
Driving time
For Belgium data are available.
The above mentioned risks are not the same for different vehicles. For some vehicles of interest data are shown in
the table (Belgian data 2001: http://www.statbel.fgov.be/figures/d37_nl.asp).
These Belgian data show an average driving time for lorries that is 1.5 times that of cars; for buses about 3 times. To
stay at the safe side we could use 2 times for lorries and 4 times for buses. This is almost certainly an overestimation
in view of the fact that the average speed goes up and thereby the average time at the wheel goes down by an
increase of the distance driven, because it is likely that with an increase in distance, more kilometres will have been
driven on motorways.
For the sake of uniformity we will use a factor 5 overall, a very safe value.
We did not find data for taxis. We asked taxi drivers and came to an average of 65,000 km/year. It seems that a
factor 5, accounting for 80,000 km/year is safe. Taxi drivers often make long hours, but much of it is waiting for
customers. In some countries or cities taxi drivers drive around all the time. Here the number of kilometres
increases, but not necessarily the number of kilometres driven with customers!
The theoretical limit of driving time in the European Union is 9 hours per day for 6 days a week, which is 54 hours
per week. Assuming 4 weeks of holiday, this is 48 x 54 = 2592 hours. One year is 8760 hours, so this is over 29% of
the time, compared to 4.2% as an average for cars: 7 times as much. These values can only be reached for long
distance (international?) transport, but even there it seems highly unlikely that they would be reached as an average.
Remember: drivers of cars can drive for 12 or more hours a day too!
The higher ratio for motorcycles and mopeds (group 1) is due to the more exposed position of those drivers, but
could be due to the effect of having less driving experience as well as to an influence of the young age of the drivers.
This group consists of 3 subgroups: motorbikes >400cc; motorbikes <400cc and mopeds that can drive >40km/h.
These subgroups contributed in 2001 to fatalities with ratios compared to cars of respectively 7.9; 1.4 and 4.0
(Belgian data: not corrected for distance driven).
Are there an increased number of serious accidents in group 2 vehicles? The factors " toll of accidents"
If anything, the opposite can be derived from the figures in the table about Belgian data above. As far as the ratio of
serious accidents is concerned, the lesser ratio of accidents per vehicle for buses and lorries will in part be due to the
effect that more experienced drivers produce less accidents. Since the total number of accidents is not given, we
cannot calculate a ratio serious accidents per total number of accidents. Why would this be important?
If accidents that occur because of a seizure at the wheel are more serious than accidents with other causes and if we
want to attribute a factor of severity, ideally we would want to know the severity of accidents that were caused by a
seizure at the wheel for these different vehicle groups. Such data are not available. Neither are data available about a
comparable situation: accidents caused by loss of consciousness at the wheel. The next best approximation is to
know the "normal" severity per accident, for each vehicle group. This factor would be applicable over the time that
41
Final report; 3 April 2005 42
someone drives because the occurrence of a seizure is unpredictable. This is different from the normal situation,
where the relative number of accidents tend to decrease with the time driven, because of the effect that more
experienced drivers produce less accidents. So while in normal drivers the risk decreases with distance or time
driven, in the case of possible seizures the chance increases, because it is a function of the time spent at the wheel.
In a Canadian consensus conference (Canadian cardiovascular society 2003) truckers were found to be involved in
2% of accidents, but 6.2% of fatalities. From this, a factor 3.6 for severity could be derived. It seems more realistic
to use all serious accidents as a measure, not just fatalities. The European data on serious accidents have been based
on different criteria in some of the member states. For that reason, we did not use them. For Europe, the same
numbers can be found in the table.
So the same factor that has been found in Canada (3.6) in European date is 2.93. This can be taken as an overall
factor of severity of accidents produced by group 2 vehicles compared to cars.
A Dutch website (www.SWOV.nl slachtoffers) for traffic information states that per one billion vehicle kilometres a
car produces 64 deaths outside the vehicle, a truck 114. This is a factor 1.8.
It seems an established fact that heavier vehicles protect their drivers (and passengers?) from the impact of an
accident. This means that the relative risk shifts in the direction of the other road users. That should be taken into
account with the assessment. It is shown in the data below about passenger transport. The factor "toll of accidents
will not been taken into account separately.
Comparable and other European data for 2002 (others = passengers + outside the vehicle):
Passeng Pass / Ratio Outside Others/ Ratio to
Killed Driver er outside driver to car /driver Ratio to car driver car
car or taxi 12.086 5806 7440 0,48 1,00 1,10 1,00 0,62 1,00
bus or coach 26 78 655 3,00 6,24 28,19 25,72 25,19 40,92
heavy goods
vehicle 430 65 3192 0,15 0,31 7,57 6,91 7,42 12,06
lorry, under 3.5
tonnes 652 244 1713 0,37 0,78 3,00 2,74 2,63 4,27
unknown lorry 209 40 666 0,19 0,40 3,38 3,08 3,19 5,18
These last data seem to suggest that for every bus driver killed 6 times as many passengers are killed, and another 25
as many persons outside the bus! The ratios for HGVs are favourable for passengers, but 7 times as many others are
killed per accident than in a car crash! These data seem to confirm the severity of bus accidents, mainly for people
42
Final report; 3 April 2005 43
on the road outside the bus. It is however not clear how to interpret these data, when the same risks do not show up
in overall statistics that compare car an group 2 accidents!
A conclusion
Motorbikes should be assessed more severely than a car, because of a greater danger for the driver, or at least should
the patient be informed about this. The danger to the driver is 2 to 3 times increased compared to a car.
The 2% rule can still be applied for buses and HGVs. Some other vehicles that, in some countries, drive under group
2 conditions do not need to be assessed as severely. (e.g. taxis)
In countries where the transport of people organised by the employer is assessed according to the criteria of group 2
the assessment should take account of the number of people transported and an arbitrary safety factor (3-5?).
43
Final report; 3 April 2005 44
THE PERCENTAGE OF THE POPULATION THAT HAS A FIRST SEIZURE PER YEAR
Jallon (Jallon 2001) lists studies in western countries and concludes that there is a mean incidence per year of 70 /
100.000 Hauser finds 61 per 100,000 person-years for first unprovoked seizures (and 44 / 100,000 for epilepsy:
Hauser 1996) A Swedish study finds an incidence of first seizures of 56/100,000/year between ages 17 to 60. Above
age 65 the incidence was 139/100,000/year (men 166, women 116) (Forsgren 1996)
We will take 70 / 100,000 for our calculations
THE PERCENTAGE OF PEOPLE WITH EPILEPSY THAT HOLDS A DRIVING LICENCE IN THE EUROPEAN UNION
In Finland the percentage of young people with epilepsy that drive is 69% versus 89% in the same age group of the
general population (Ojala M 2001). In a USA study, the percentage of drivers with epilepsy that hold a licence was
57% (Krauss G 2001). Sonnen (Sonnen 1997) finds an average of 50%. The RESt-1 Group study (RESt-1 Group
2000) recently found an average of 50.3% in a group of 570 adult patients from six European countries compared
with 75% in the control group (we excluded the Russian patients and controls).
The theoretical maximum of people with epilepsy that holds a drivers licence is the percentage of people that become
seizure free: 70%.
The percentage of the general population that holds a drivers licence is 53% (200/375 million).
This makes the maximum percentage of people with epilepsy that has a licence at the time of writing: 53% * 70% =
37.1%
The theoretical maximum is: 70% times the theoretical maximum of the population. In most cases we have used
70%.
DO SEIZURE-RELATED ACCIDENTS CAUSED BY PERSONS WITH EPILEPSY MORE OFTEN LEAD TO SERIOUS INJURY?
There is uncertainty about this.
In the UK study the number of accidents was not increased in people with epilepsy with a valid driving licence, but
there were twice the expected number of fatalities. These data were somewhat uncertain because of the methodology
used (they could not calculate a relative risk, because there were no fatalities in the control group) (Taylor et al.
1996). The same UK study found an increase in the chance of a serious accident of 40%. (OR: 1.37 CI: 1.02-1.84)
44
Final report; 3 April 2005 45
Other studies only look at the overall chance of an accident. If this is increased, it seems likely that the number of
serious accidents is increased proportionally.
Sonnen (1997) summarises four older studies and concludes that the question of increased severity is unsolved.
Data from The Netherlands (Sonnen 1997) show a 10% average risk of an accident per year per person with a vehicle
insurance. (Some minor accidents might not come to the attention of the assurance, but they are likely to be of no
importance for safety) In the large UK study of Taylor et al. (1996) it was about 7%. Data from the literature vary
between 5.7% and 12% with an average of 10% (Sonnen 1995).
Sonnen looked at the factors that contribute to this percentage, there are important variations, even for variables that
can not be influenced at all. We already mentioned these factors above. (see: Figure 1)
The suggestion of the European workgroup was to take a 10% increase on top of the average of 10% accidents per
year as an acceptable risk, which is an absolute increase of 1%. Clearly this is a very small increase in risk compared
to the above-mentioned variables.
For group 1, a 10% increase of the average risk of an accident for the population was proposed. This population
average is 10% per year, so the resulting increase in risk is 1% (10% x 10%).
45
Final report; 3 April 2005 46
STUDIES
Prospective study of the recurrence risk and the timing of recurrences in the first months to 2 years
Prospective study about the long-term recurrence-risk in seizure-free patients: 5 to 10 (20?) years of follow-up.
Meta-analysis of studies about both these situations.
European statistics about epilepsy (and other medical causes) and traffic accidents to underpin our decisions by better
data about the dangers.
European data about the Percentage of people with epilepsy that hold a drivers licence.
An overview of the current knowledge on the recurrence risk of the different types of provoked seizures as a base for
guidelines concerning driving ability.
An overview of disorders that might qualify for the installation of a prophylactic ban and the COSY thereof as a base
for guidelines concerning driving ability.
Creation of an European registry of difficult problems in the assessment of people with epilepsy or seizures and
discussion in this context.
Prospective study about the compliance with the rules once they are in place as well as for the occurrence of
accidents and the severity thereof, for the respective clinical situations mentioned in the overview of the regulations
and for both groups.
IMPLEMENTATION
Support consistency of the approach with other areas of medical risk assessment.
Sending of the Power Point slides to the members of the Working Group for presentations on the national level April
2005.
Presentation of the development of the arguments as described in this report on the International Congress on
Epileptology in Paris, August/ Sept 2005.
Presentation of the document on the 7th European congress on Epileptology in Helsinki, Finland, 2-8 July 2006.
Submission of the document to Epilepsia, if possible to be published as an annex, as soon as the renewed Annex III
is officially in place (2006).
Posting of the final version of the PP slide show and pdf document in downloadable version on the ILAE website as
soon as the renewed Annex III is officially in place (2006).
Inform the National Leagues of the Member States as soon as the renewed Annex III is officially in place, so they
can contact their governments for further advice on the national level (2006).
46