Accident Analysis and Prevention 41 (2009) 869875

Contents lists available at ScienceDirect

Accident Analysis and Prevention

journa l homepage: www.e lsev ier .co

Convicted of fatigued driving: Who, why and how

Igor RaduDepartment of

a r t i c l

Article history:Received 9 NoReceived in reAccepted 29 A

Keywords:SleepinessFatigueMotor-vehicleSeasonalityAlcoholAmphetamine

c ac. the

termivingdriveatigull driue-requenthat mnishe

f the c35 yrta. Th

and unambiguous criteria for recognizing the contribution of fatigue to accident causation, Finnish policeand the courts punish a signicant number of drivers every year on the basis of fatigue.

2009 Elsevier Ltd. All rights reserved.

1. Introduc

Recognizwhen dealiAn adequatnecessary treduction obasis of tireability of th

The ofcrelated accroutinely rewhen a chereporting fothe choicesand revaluathe numbe1995). For ereported ac

Corresponment of PsychTel.: +358 9 19

E-mail addURL: http:

0001-4575/$ doi:10.1016/j.ation

ing an accident as sleep related is a central problemng with fatigue and sleepiness problems while driving.e identication of the risk factors and risk groups iso be able to conduct safety campaigns aiming at thef sleep-related accidents. Prosecuting drivers on thedness/sleepiness also depends on the validity and reli-e accident investigation process.ial statistics from many countries regarding fatigue-idents are missing because causal factors are notcorded (Horne and Reyner, 1999). Furthermore, evencklist of accident causes is included on police accidentrms, it does not necessarily include fatigue as one of(Knipling and Wang, 1994). Nevertheless, reanalyzingting police reports is a usual practice when estimatingr of fatigue-related accidents (e.g., Horne and Reyner,xample, in the US from 1989 to 1993, 1% of all police-cidents and 3.6% of fatal accidents were attributed to

ding author at: Human Factors and Safety Behavior Group, Depart-ology, PO Box 9, 00014 University of Helsinki, Finland.129508; fax: +358 9 19129481.ress: igor.radun@helsinki. (I. Radun).//www.mv.helsinki./home/radun (I. Radun).

driver drowsiness (Knipling and Wang, 1994). Similarly, in Switzer-land sleepiness was a causative factor in 1% of trafc accidents thatincluded bodily injury or property damage in excess of 500 Swissfrancs (Laube et al., 1998). In Sweden, from 1994 to 2001, 3% ofsingle vehicle accidents reported to the police were fatigue related(Anund et al., 2002). Finally, in New South Wales, Australia about20% of fatal accidents were found to be fatigue related (Roads andTrafc Authority, 2001).

In addition to the blurred concept of fatigue, the inexistence of avalidated and reliable device for detecting the level of sleepiness (cf.the breath analyzer for alcohol levels) aggravates the work of inves-tigating ofcers. Unfortunately, there is little evidence that such adevice will be available in the (near) future. Therefore, investigat-ing ofcers have to rely on other clues coming from the participantsthemselves, eyewitnesses, and the characteristics of the crash itself.However, there are no criteria for the unambiguous detection offatigue/sleepiness as a major or contributing factor in accident cau-sation. Whether the particular accident will be coded into ofcialstatistics as sleep related or not depends solely on police ofcerssubjective opinions.

When judging whether driver fatigue contributed to an accident,investigating ofcers, obviously, do not have direct knowledge of adrivers preaccident condition. In addition, a drivers arousal andemotional state changes after the accident and does not necessaryreveal anything about the preaccident condition. Drivers might alsohave difculties recollecting the period prior to the accident and

see front matter 2009 Elsevier Ltd. All rights reserved.ap.2009.04.024n , Jenni E. RadunPsychology, University of Helsinki, Helsinki, Finland

e i n f o

vember 2008vised form 26 April 2009pril 2009

accidents

s

a b s t r a c t

Fatigue is a major cause of road trafthe lack of reliable testing devices (cfincorporate fatigue in operationalizedRoad Trafc Act explicitly forbids drfactors (sickness, etc.) that impair ato investigate the circumstances of fdriver record database we extracted athese drivers, 90.4% committed a fatigwere the most common (92.5%) consethought to be serious, our data showsAlmost every twentieth driver was puof alcohol or drugs was noted in 13% otired or falling asleep. Young men (seasonal effects were clear in this dam/locate /aap

?

cidents. However, due to the blurred concept of fatigue andbreath analyzer for alcohol levels), it is extremely difcult tos into either trafc or criminal law. Even though the Finnish

while tired, it is done only on a general level among otherrs tness to drive (Article 63). The present study was donee driving offenses. From the Finnish Vehicle Administrationvers (N = 768) punished under Article 63 from 20042005. Oflated trafc offense. Accidents, predominantly single vehicle,ce of fatigued driving. Although fatigue-related accidents are

ost of the accidents (81.6%) did not involve personal injuries.d because his vehicle was drifting on the road. The presenceases. Only 3.1% of the punished drivers ofcially denied beings) represented 50% of all punished drivers. Time of day andis study shows that even without a reliable fatigue detector

870 I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875

even the accident itself due to traumatic effects of the accident. Ina case when a driver falls asleep and causes the accident, the rec-ollection might be even poorer. Such an assumption comes fromexperimental results showing that subjects who fall asleep usuallydeny havingand Moore,cases whenthe accidendrivers arerassment ofand legal co

How poledly inuenpractices. Adescriptionrent interethe secondfavored expcrashes at Tof the accidrole of fatigCortsen, 2tory effectsdoses and acentration)consumptiobination (Ba

Given threal challeneither trafdiscussions

An in-deAustralia (VInstitute, 20step towardfatigue wastion of thecauses an ative hours cand sentenc$100,000.

The Finntired in Artness to drivdriving becwhose healfor grantingtranslation)fatigued driare of suchdescribe hocumstancesof risk facto

2. Method

2.1. Data

From thewe extracteRTA from 20directly frofrom the cople were reqstationed a

to the public, and obtaining them for research purposes is free ofcharge. We obtained altogether 768 out of 776 cases (99%); dueto unknown reasons only eight cases could not have been traced.More specically, we obtained 628 out of 632 (99.4%) prosecutor

ns, 129 out of 131 (98.5%) district court decisions, and 11 out84.6%) courts of appeal decisions.

osecutor and court decisions

rosecutor decision is based on police investigation form thates a l

sho), andor thdoine post, w

gatinationecutgniatemn, tharyone

ssiblationy, deearc

e law

re arand,ant.

ies foonmlight) arey-ner (Jond a

offenerytigu

thaepencanCh.ty (Rcrim

ppennlex

ata a

anaand

sion o

e trafer limmittined D 1ews.bbeen asleep if woken up within two minutes (Bonnet1982; Horne and Reyner, 1999). On the other hand, indrivers are able to recall information about the causes oft, they do not necessarily report it. It might be that somenot willing to admit to falling asleep due to the embar-being labeled a bad driver or concerns about insurancensequences (Cortsen, 1999; Reyner and Horne, 1998).ice ofcers attribute the cause of an accident is undoubt-ced by current developments in safety research ands Ogden and Moskowitz (2004) pointed out, Police

s of crashes are typically assigned to the cause of cur-st. These authors offer an example of how throughhalf of the twentieth century inattention became alanation instead of loss of control for the same type of-intersections. Furthermore, focusing on other aspectsent, typically on alcohol involvement, might leave theue/sleepiness in the accident causation unnoticed (e.g.,003). However, it is well known that besides stimula-, alcohol has sedative effects that are present at highert the descending phase of the BAC (blood alcohol con-curve (Roehrs and Roth, 2001). Even minimal alcoholn and increased sleepiness represent a hazardous com-nks et al., 2004).ese difculties in dening and detecting fatigue, it is age to incorporate fatigue in operationalized terms intoc or criminal law. Worldwide, there are many currenton how this challenge can be met.pth debate on this topic was recently conducted inictorian Government, 2004; Tasmania Law Reform07) and the UK (House of Commons, 2006). The biggestaddressing the legal responsibilities concerning drivertaken in the state of New Jersey, US, with the introduc-

so-called Maggies Law. Under this law, a driver whoccident after being awake for more than 24 consecu-an be convicted of second-degree vehicular homicideed to up to 10 years in prison and ned a maximum of

ish Road Trafc Act (RTA) explicitly forbids driving whileicle 63 (3.8.1990/676), which addresses the drivers t-e: a person that does not meet the requirements for

ause of illness or tiredness or another similar reason orth condition no longer fullls the requirements needed

a drivers license must not drive a vehicle (unofcial. It is unknown how many drivers are charged because ofving under Article 63 and what the actual consequencesfatigued driving. Therefore, the aim of the study was tow this law is applied in practice and to determine the cir-of fatigue driving offenses, including the identication

rs and risk groups.

Finnish Vehicle Administration driver record databased all drivers (N = 776) punished under Article 63 of the04 to 2005. Of these drivers, 632 received a punishment

m the prosecutors, 131 from the district courts and 13urts of appeal. The decisions covering the selected sam-uested from the prosecutor ofces and courts in charge

round Finland. Court decisions in Finland are available

decisioof 13 (

2.2. Pr

A pincludetc.), awordsbasis fand byhas thincideninvestinoticto proshave siness stdecisioa summwe hadwas poinformseveritthe res

2.3. Th

Thein Finlimportpenaltimpris(e.g., sto 120of a daoffendin Finlatrafc

A vbids faoffenseever, ddriver98; PCc Safebrings(see Awww.

2.4. D

Ourecutorprofes

1 Sincany uppfor comwas n(http://nist of predened information to ll in (date, time, place,rt description of the offense (in police ofcers own

in most cases an explanation of the offense and thee charge. The accused has to sign the completed formg so accepts what is written there; however, the accusedsibility to give a statement providing his view of the

hich can be contrary to the information provided by theg police ofcer. If the defendant ofcially contests the, a charge might be brought in a district court. Comparedor decisions, district court and court of appeal decisionscantly more text and information available (e.g., eyewit-ents). They are written by the court and include the nale reasoning and the grounds on which it was based, andof the court process. Although for each and every casetype of court decision, the analysis of the whole samplee because all three types of courts decisions contained

(age and sex of defendant, vehicles involved, injuryfendants testimony, etc.) necessary for the purpose ofh.

e a number of acts and decrees covering road safetyof which the Road Trafc Act (267/1981) is the most

The Road Trafc Act (RTA) and the Penal Code (PC) dener trafc crimes, but only the PC denes crimes leading toent. Fixed nes are used only for minor trafc offensesly exceeding the speed limit); while day-nes (from 1the predominant form of punishment. The actual sume depends on the monthly income and the assets of theutsen et al., 2001).1 This day-ne system has been useds a form of punishment for all kinds of offenses (not onlyses) since its introduction in 1921.important issue regarding the Article 63 that for-

ed driving is that breaking it, by default, is a trafct brings a day-ne punishment (RTA Art.103). How-ding on the seriousness of the offense, a fatigued

be charged with Endangering Trafc Safety (RTA Art.23, Sec. 1) and possibly with Gross Endangering Traf-TA Art. 99; PC Ch. 23, Sec. 2). Breaking these articlesinal responsibility and might lead to imprisonment

dix). (An unofcial translation of PC can be found at:./en/laki/kaannokset/1889/en18890039.pdf).

nalysis

lysis was based on information extracted from the pros-court decisions. A list of variables included sex, age andf the driver, date, time and place of the incident, vehicles

c offenders pay nes in proportion to their incomes, withoutit, it is possible for someone to pay an extremely high ne

g a trafc offense: for example, in 2004 a Finnish millionaire70,000 or $216,000 for speeding (26 days income of D 6,538)bc.co.uk/1/hi/business/3477285.stm).

I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875 871

involved, possible injuries and intoxication, the presence of othercrimes and charges in addition to the charge with Article 63, anddrivers testimony. Using these variables we present the character-istics of the fatigue-related incidents mostly in terms of descriptivestatistics. Chi-square, t-test and logistic regression were used forthe comparison of drivers who were judged to have fallen asleep orbeen only fatigued. A hierarchical log-linear analysis was used toexamine the seasonality and time-of-day effects in relation to theage and sex of the drivers. Both backward elimination and forwardselection procedures were used to obtain the best tting model. Alog-likelihood ratio goodness-of-t statistic (G2) was used to eval-uate the adequacy of the model. In order to avoid empty cell valuesin cross tabulations in logistic regression and log-linear analyses,the following variables were categorized: age (25, 2635, 3645,5665, 66), time of day (00:0106:00, 06:0112:00, etc.), andseason (January-March, April-June, etc).

3. Results

3.1. Basis fo

Althoughsickness anafter exammost (N = 6committedmostly inclubetes, epileinformation63. In furth(N = 694).

Whethesibly to a colevel depenity, and therather thanthe driversfalling asleemajor contr

3.2. Trafc

In 92.5%kind of an adents) (Tablbecause hissafety evencles. Three dwhile waiticommitteda red light, din two caseincident.

Table 1Type of fatigue

Single accidenHead-on accidRear-end accidOther accidenDriftingFell asleep in aOther drivingUnknown

Total

Naturally, personal injuries were present only in accidents(N = 642), of which 81.6% did not involve injuries. In 17.1% of the acci-dents, someone was slightly or seriously injured, and eight (1.2%)accidents were fatal. There was no connection between accidentseverity (injury vs. no injury) and whether the driver was judged tohave fallen asleep or been only fatigued (Chi-square test, 2 = 0.33,d.f. = 1, p > 0.05).

In addition to the charge of driving while fatigued, a large major-ity of drivers (96.1%) were convicted of endangering trafc safety(RTA Art. 98), 11 drivers (1.6%) of grossly endangering trafc safety(RTA Art. 99), and 16 (2.3%) of a less serious trafc offense (RTAArt.103). A day-ne was the most predominant (90.3%) type of pun-ishment with an average number of 16 day-nes. The rest of thedrivers (9.7%) received a prison sentence: of these 32.8% actuallywent to prison, for 52.2% the prison sentence was suspended, and14.9% did community work instead of going to prison. Similarlyto the severity of the charge, the severity of the punishment was

relace ofsev

coho

presiversthe led asncesrivinof tadiff

udget = 0

ivers

ry si, or sriverthe

our das ther tht. Macced dr

e, se

driv5 ye(Fig

testim

the chhe chrt of ao havo beinallingd thenmatior a charge

Article 63 of the RTA lists not only fatigue, but alsod other factors concerning a drivers tness to drive,ining all decisions (N = 768) in detail, we found that94, 90.4%) of the drivers punished under this articlea fatigue-related trafc offense. Other reasons (6.9%)ded driving impairment due to different diseases (dia-

psy, etc). In 21 cases (2.7%) we were unable to nd anyabout why these drivers were punished under Article

er analysis, only fatigue-related offenses were included

r a particular case went to a district court and later pos-urt of appeal or whether it stayed only at a prosecutorded mostly on drug/alcohol involvement, injury sever-presence of other crimes (drug possession, robbery, etc.)on the severity of fatigue inducing behavior. Amongpunished of fatigued driving, 71.6% were accused ofp behind the wheel; for the rest, fatigue was noted as aibutor to their driving incidents.

incidents and consequences

of the cases, a fatigued driver was involved in someccident, mostly a single vehicle accident (81% of all inci-e 1). Almost every twentieth (4.9%) driver was punishedvehicle was drifting on the road and endangering roadthough there was no physical contact with other vehi-rivers (0.4%) were found asleep in a car in the crossing

ng to turn or join the main road. Some drivers (1.9%)other driving offenses (crossing an intersection againstriving in the wrong direction, etc.) due to fatigue. Only

s (0.3%) was no information available about the driving

-related offense.

N %

t 562 81.0%ent 41 5.9%ent 14 2.0%

ts 25 3.6%34 4.9%

stopped car in the crossing 3 0.4%incident 13 1.9%

2 0.3%

694 100.0%

mostlypresenon the

3.3. Al

TheFive draboveregardsubstawhile d5 caseswas nowere j(t-test,

3.4. Dr

Evedrowsyished dfour ofthese froad wjust aftin couring byfatigue

3.5. Ag

Themen 3drivers

Table 2Drivers

AcceptsAdmits t

or couAdmits tAdmits tDenies fAdmitteNo infor

Totalted to drug/alcohol involvement, injury severity, and theother crimes (drug possession, robbery, etc.) rather thanerity of fatigue inducing behavior.

l and drugs

ence of alcohol or drugs was noted in 92 (13.3%) cases.had a BAC below the legal limit of 0.05%, 60 drivers were

imit (40 had their BAC above 0.12%, which is in Finlandgross drunkenness). An additional 27 drivers consumedwhich were generally illegal or whose usage is forbiddeng (15 cases of amphetamines, 4 cases of drug mixtures,king sleeping pills, 1 of cannabis, 2 of medicine). There

erence in BAC levels between intoxicated drivers whod to have fallen asleep or who had only been fatigued.84, d.f. = 58, p > 0.05).

testimony

xth driver admitted to falling asleep or being tired,leepy before the incident (Table 2). Only 3.1% of the pun-s ofcially denied being tired or falling asleep; however,

eight drivers that caused a fatal accident did so. One ofrivers suggested that the appearance of a moose on thee cause of the accident. Another told the police ofcere accident that he had fallen asleep but denied this laterost of the drivers admitted the charge of fatigued driv-pting the charge as a whole or a more specic charge ofiving.

x, and profession

ers were predominantly men (80.7%) and of young age;ars old and younger represented 50% of all punished. 1). Regarding the profession of the drivers, there was

ony regarding the charge of fatigued driving.

N %

arge/decision by signing (prosecutor) 469 67.6%arge generally or specically (district courtppeal)

85 12.2%

ing fallen asleep 68 9.8%g tired/drowsy/having a lack of sleep 44 6.3%asleep/being tired 15 2.2%denied 6 0.9%

n (district court) 7 1.0%

694 100.0%

872 I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875

Fig. 1. Number of fatigue-related offenses by sex and age of drivers (N = 694).

a large variability in reporting details in the available documenta-tion. For 5% of the drivers, the profession was unknown. Of the rest,we managed to identify the following large distinct groups: 17.7% ofthe driversretired, andclassify, buting the 58 pthem wereabout duty

3.6. Time o

A clear odistinct timbe noted hecases; theretting modp = 0.81) inctime-of-dayrelated offedo not diffetion betwee

Fig. 2.

Fig. 3. Time of day distribution of fatigue-related offenses by age of drivers (N = 658).

(25 yrs) had their fatigue-related offenses predominantly at night,while the proportion of daytime offenses increases with drivers age(Fig. 3). Almost two-thirds (62.5%) of the offenses for drivers up to

rs old

ason

datapril(Fig.bet

e ma, p =offe

lling

rderor b

pliedincluthe l, andwere students, 8.8% were professional drivers, 7.1% were2.1% were military conscripts. The rest were difcult toroughly 40% of the drivers had low education. Regard-rofessional drivers, we found that at least 41 (70.7%) of

driving on duty; nine truck drivers had broken the rulesand rest hours.

f day

verall time-of-day effect can be seen in Fig. 2, while ae-of-day and age pattern are depicted in Fig. 3. It shouldre that time of the offense was unknown in 5.2% of thefore number of cases in Figs. 2 and 3 is 658. The best

el in hierarchical log-linear analysis (G2 = 16.99, d.f. = 23,luded the main effect of sex and an interaction betweenand age. This indicates that men commit more fatigue-

nces (as also Fig. 1 shows), although men and womenr across age groups in terms of time of day. The interac-n age and time-of-day is such that the youngest drivers

25 yea

3.7. Se

Thefrom Ain Julyactionsonly thd.f. = 38related

3.8. Fa

In oasleepwe apTheseaboveof dayTime of day distribution of fatigue-related offenses (N = 658). Fig. 4occurred between Friday and Sunday.

also shows a clear seasonal effect. After a sharp riseto May, the peak of fatigue-related offenses is reached4). A hierarchical log-linear analysis showed no inter-

ween variables season, and age and sex of the drivers;in effects of all three factors were signicant (G2 = 40.84,0.347). This indicates that the seasonality of fatigue-nses is present across all age and sex groups.

asleep vs. fatigue

to predict whether a driver was judged to have falleneing only fatigued when causing an accident (N = 642),a logistic regression with several predictor variables.ded age and sex of the driver, intoxication (no, BACimit, or under the inuence of drug/medication), time

season. The only two variables with some predic-. Monthly distribution of fatigue-related offenses (N = 694).

I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875 873

tive power were time of the accident and intoxication. Comparedwith nonintoxicated drivers, those who were under the inuenceof drug/medication were less likely found to have fallen asleep(OR = 0.29, CI 0.100.79), while there was no difference for thosewho were inCI 0.401.624:00, drivethe night ho12:0118:0for morning

4. Discussi

To our kcumstancesdata. Previoand Jones, 2study (Raduon crashesconclusionstion teams afatigue to a

Given thconvictingpresent dathas been ction, court rdrivers in oOf course, iunmentionassume thasigned a stathe charge (given the ge(Joutsen etbring harshnecessary toor being imlaw had beeexample, mfour of thetired prior tare reluctanresponsibilipossibly fac2009), we rdants changthe other bl

This studas well as ththat sleep-r(Horne andmore than 8injuries. Th(1995) in winvolve perdifferencesthe US, espe

It is alsocontributioCortsen, 2convicted oication. In sthat policeasleep behi

their investigation went beyond simply attributing the accidentto high levels of alcohol. However, given that a drunk driver wasinvolved in 18% of 6767 accidents involving personal injuries in2004, while in 2005 the proportion was slightly smaller (16.1% of

ccideen iverylood

ter aan ex

(Mur whverstami

theroundg onan accomlanatis nthercreatand

lem tlice

specry tiseas

his isof

ertimsurvng adge,

on ofrepo

nts inest

asonin res, inc

studng thare

oundandand

revioall

Sumatigur dr

nds).oo liothe

, andeightonscr

longy con

ally wet aed atoxicated above the legal BAC limit of 0.05% (OR = 0.80,1). Compared with accidents occurring from 18:01 tors were more likely found to have fallen asleep duringurs 00:0106:00 (OR = 2.21, CI 1.194.11), and afternoon0 (OR = 2.20, CI 1.184.13), while there was no difference

hours 06:0112:00 (OR = 1.61, CI 0.873.01).

on

nowledge, this is the rst study that analyzes the cir-of fatigue-related trafc offenses from a large set of

us studies reported only single accidents (Rajaratnam004) or a few cases (Desai et al., 2003). In our recentn et al., 2009), we also analyzed only nine fatal head-and reported differences in the discussions held andreached between multidisciplinary accident investiga-nd Finnish district courts regarding the contribution of

ccident causation.e well-known difculties in recognizing fatigue and

a driver on the basis of fatigue, the strength of thea is that the contribution of fatigue to accident causationonrmed on several levels: police ofcer investiga-ulings, and driver testimony. Only 3.1% of the punishedur study ofcially denied being tired or falling asleep.t is possible that some expressed objections remaineded in the available documentation. Nevertheless, we cant such a possibility is unlikely since most of the accusedtement (for a prosecutor decision) or verbally admittedin a district court or court of appeal). On the other hand,nerally low penalties typical of the Finnish legal systemal., 2001) and the fact that breaking Article 63 does noter penalties, it might be that some drivers did not nd it

object to the charge of falling asleep behind the wheelpaired due to fatigue. We may only speculate that if then signicantly different (stricter), as is Maggies Law, forore people probably would have objected. The fact thateight drivers who caused a fatal accident denied beingo the accident supports the hypothesis that defendantst to admit to being tired or falling asleep and take thety in a case of a more serious trafc offense and thuse a harsher penalty. In our previous study (Radun et al.,eported a similar situation in which one of nine defen-ed his statement during the investigation process andamed the cause of a fatal accident on a moose.y brings new ndings concerning fatigue-related incidentse recognition of fatigue in trafc. It is generally believedelated accidents are likely to have serious consequencesReyner, 1999); however, the present data shows that0% of fatigue-related accidents did not involve personal

is is in contrast with, for example, a report by Pack et al.hich only 40.6% of the falling-asleep accidents did not

sonal injuries. However, there is a possibility that somein reporting minor accidents exist between Finland andcially when dealing with single vehicle accidents.thought that police ofcers are likely to neglect the

n of fatigue in the case of an intoxicated driver (e.g.,003). Our data shows that 13% of the drivers were

n the basis of fatigue in addition to the presence of intox-ome cases, it was evident from the available materialofcers pointed to the high BAC as the cause of fallingnd the wheel. This shows that these police ofcers in

7002 anot oftingly, ein his bthat afdays),to 18 huncleathe driamphe2009).

Anowere fdriftincase ofthat acan expfore, ittired. Otheir inundersa probfor a poand suonly ve

Thedata. Tportionsummrecentof falliknowlevariati(1985)accidethe lowthis senationseason2006).

Ourregardieffectsthose f(Horne(Radunwith p50% ofdents (their fyoungeweekesleep tOn thescriptsof theitary cafter amilitarespeciRadunidentints) (Statistics Finland, 2007), it seems that fatigue isnvestigated in connection with intoxication. Interest-

sixth intoxicated driver had traces of amphetaminessample. Amphetamines are stimulants, but it is typical

long acute use and prolonged wakefulness (e.g., severalhausted user will fall asleep and remain asleep for 12rray, 1998). From the available documentation it was

at role, if any, amphetamines had in the development offatigue. Driving under the inuence of drugs, especiallynes, is an increasing problem in Finland (Ojaniemi et al.,

important nding is that 4.9% of all the punished driversguilty of fatigued driving because their vehicle was

the road, endangering road safety. However, as in thecident, drifting is a sufciently signicant trafc offense

panying charge of fatigued driving can be rather seen asion for the action than as a more serious charge. There-ot surprising that none of these drivers denied beingwise, it would have been indeed difcult to prove thatsed fatigue had resulted in the vehicles drifting. As wefrom our contacts with Finnish police, this seems to be

hat police ofcers face on a daily basis. It is not unusualofcer to stop a driver because of weird driving (drifting)ted intoxication, only to discover that the driver seemsred.onality of fatigue-related trafc offenses is clear in thissimilar to our two studies where the number and pro-

sleep-related fatal accidents were much higher in thee (Radun and Radun, 2006; Summala et al., 2003). Our

ey data indicated a similar trend in self-reported casessleep while driving (Radun and Radun, 2008). To ouronly a few reports have discussed the possible seasonalsleep-related road accidents. For example, Langlois et al.rted a seasonal pattern in fatigue-related single-vehicle

Texas, with the highest number in May and July andnumber in January and February. The reasons behindality are still to be discovered; we proposed an expla-lation to different driving and lifestyle habits betweenluding sleep quality and age effects (Radun and Radun,

y also conrms previous ndings on a new type of datae characteristics of fatigue-related crashes. Time-of-dayclearly visible in this data and are consistent withon fatigue-related accidents investigated by the policeReyner, 1995) or multidisciplinary investigation teamsSummala, 2004). Age and sex effects are also in line

us results. Men 35 years old and younger representedpunished drivers. Similarly to the data on fatal acci-mala and Mikkola, 1994), older (retired) drivers commite-related offenses mostly during the afternoon, whileivers are overrepresented at night (especially duringMany of these young drivers were students who often

ttle as a part of their lifestyle habits (Carskadon, 1990).r hand, some of these young drivers were military con-although they represented 2.1% of all the drivers, onefatal accidents in this data was caused by a young mil-ipt. This conscript was driving home from the garrisonnight march. There is growing evidence that Finnishscripts are at high risk of falling asleep while driving,hen driving home from the garrison (Partinen, 1982;

l., 2007; Sipinen, 1990). Professional drivers were alsos a risk group, with some of them breaking the rules

874 I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875

about duty and rest hours. Self-reported data shows that 40% ofFinnish long-haul drivers drive more than 10 h within a 24-h period,breaking EC Regulation No. 3820/85 (Hkknen and Summala,2000).

The presaverage one20042005sonal injuriFinland, 20accidents awas altogetmation is tfatigue. Howthis is undeFinland. Throad crashegram. Multiin depth shthe contribumethod shorelated acciteams showdrivers, 10.1while fatigucases (Radusame data gdrivers inclon a represe19.5% of thepoint durinto falling aing duringthese studiegroup.

4.1. Limitat

Estimation the basunderestimtypical for dstill a respeally punishsimilar to ochecklists oing forms.estimate thdents; we rfatigued driing offensereviewers,bias givensample. Itreceive a tic1983). Howare at highies from di(police accidecisions, dsurprising trelated offeof the presefatigue to aof a new tyand conrmdriving.

4.2. Conclusion

In conclusion, the present data shows that even without a reli-able fatigue detector and unambiguous criteria for recognizing the

utiopuniue. S

g onwhdriv

te ca(e.guedonl

s andsentons fafcd dr

wled

authfro

al Trry offor to an

was sation

dix A

angeCh.brea90/2

f, in atencet forssly EA Arver ontly

to ovor we du

r comis co

ty, thsafetars.

nces

., Kecping., Catchol imin parM.H.,tion oon, M.hol, D, M.T

taking, M.T

dent Aent data shows that Finnish police and courts punish ondriver per day on the basis of fatigued driving (years

). Given that the number of accidents involving per-es in 2004 was 6767 while in 2005 was 7002 (Statistics07) and that the number of drivers involved in injurynd punished of fatigued driving during these two yearsher 118, as identied in the present study, a rough esti-hat less than 1% of injury accidents were related to

ever, data available from other sources suggests thatrestimating the role of fatigue in accident causation inere is a long tradition of in-depth analysis of all fatals in Finland that is a part of the national safety pro-disciplinary investigation teams investigating accidentsould by default provide more reliable identication oftion of fatigue in accident causation, and therefore thisuld result in a more accurate estimation of fatigue-

dents. Our analysis based on the data collected by theses that in a sample of nonprofessional nonintoxicated% of the drivers fell asleep and caused a fatal accident,e was a contributing factor in an additional 5.2% of then and Summala, 2004). Another estimation based on theives an upper estimate of 30% with alcohol-intoxicateduded (Partinen, 2004). Furthermore, our recent surveyntative sample of Finnish driving population shows thatsurveyed drivers had fallen asleep while driving at someg their driving career, while 15.9% reported being closesleep or having difculties staying awake while driv-the last 12 months (Radun and Radun, 2006). In all ofs young male drivers were identied as especially risky

ions

ng the proportion of injury accidents as fatigue relatedis of fatigue-related offences obviously produces anation. As previously noted such underestimations areata based on police ofcer judgments. However, this isctable proportion given that these drivers were actu-

ed on the basis of fatigued driving. This proportion isfcial statistics from other countries that are based onf accident causes included on police accident report-

Furthermore, the purpose of this research was not toe prevalence of fatigue-related injury accidents or inci-ather described the application of the law that forbidsving, and determined the circumstances of fatigue driv-s. Another limitation, as pointed out by one of therelates to the possibility of prosecution or convictionthe overrepresentation of young male drivers in ouris well known that (young) men are more likely toket after being stopped by a police ofcer (e.g., Homel,ever, it is also well documented that young male driversrisk of having fatigue-related trafc incidents. Stud-

fferent parts of the world using various types of datadent databases, multidisciplinary investigation teamsrivers self-reports, etc.) conrm that. Therefore, it is nothat young male drivers committed the most of fatigue-nses in our data. Despite these limitations, the strengthnt data regarding the evaluation of the contribution ofccident causation remains. This relatively large samplepe of data allowed us to present several new ndings

some of the well-known characteristics of fatigued

contribcourtsof fatigdriftindrivingness toconcreoffenseof fatigusuallychargethe preobjectiinto trfatigue

Ackno

TheEkebomNationMinistCentreand twstudyFound

Appen

End98, PCgentlyAct (10thereobe senonmen

Groard (RTthe drinegligestartstakingheed thanothethe actor safetrafctwo ye

Refere

Anund, ALink

Banks, Salcorisk

Bonnet,func

CarskadAlco

Cortsenrisk-

CortsenAccin of fatigue to accident causation, Finnish police and thesh a signicant number of drivers every year on the basisome drivers are even punished due to only the act ofa road. Although Finnish trafc law explicitly forbids

ile tired only on a general level regarding a drivers t-e, it allows a wide interpretation of the evidence in a

se. However, it seems that in the case of a serious trafc., an accident or drifting into an opposite lane), a chargedriving does not bring a signicantly higher penalty,

y in terms of day-nes. Therefore, it is clear why theconvictions of fatigued driving in the cases covered instudy have gone smoothly through without too many

rom the defendants. The struggle to incorporate fatigueor criminal law and adequately recognize and punishivers continues.

gments

ors are thankful to Anu Yl-Pietil and Johanna Rokka-m AKE, Pertti Mkelinen and Jussi Pohjonen from theafc Police, Harri Kukkola and Heikki Pietil from theInterior, and Marjatta Syvter from the Legal Registerheir help in data collection. We also thank Jussi Ohisaloonymous reviewers for their valuable comments. Thisupported in part by a grant from the Finnish Cultural.

.

ring Trafc Safetycausing a trafc hazard (RTA Art.23, Sec. 1): A road user who deliberately or negli-ches the Road Trafc Act (267/1981) or the Vehicle002) or the regulations or orders issued on the basismanner conducive to causing a hazard to others, shall

d for endangering the trafc safety to a ne or to impris-at most six months.ndangering Trafc Safetycausing a serious trafc haz-

t. 99, PC Ch. 23, Sec. 2): If in the causing of a trafc hazard,f a motor-driven vehicle or tram deliberately or grossly(1) signicantly exceeds the maximum speed limit; (2)ertake while the visibility is insufcient for safe over-hile overtaking is otherwise not allowed; (3) fails toty to stop or give way required by trafc safety; or (4) inparable manner breaches the trafc regulations, so thatnducive to causing serious danger to anothers health

e offender shall be sentenced for gross endangering they to at least 30 day-nes or to imprisonment for at most

klund, G., Larsson, J., 2002. Trtthet I focus (No. VTI rapport 933-2002).: Statens Vg och Transportforskningsinstitut.heside, P., Lack, L., Grunstein, R.R., McEvoy, R.D., 2004. Low levels ofpair driving simulator performance and reduce perception of crashtially sleep deprived subjects. Sleep 27, 10631067.Moore, S.E., 1982. The threshold of sleepperception of sleep as af time asleep and auditory threshold. Sleep 5, 267276., 1990. Adolescent Sleepiness: increased risk in a high-risk population.rugs and Driving 56, 317328.., 1999. Fatigue among young male night-time car drivers: is there a

group? Safety Science 33 (12), 4757.., 2003. Tiredness! a natural explanation to The Grand Rapid DIP.nalysis and Prevention 35, 401406.

I. Radun, J.E. Radun / Accident Analysis and Prevention 41 (2009) 869875 875

Desai, A.V., Ellis, E., Wheatley, J.R., Grunstein, R.R., 2003. Fatal distraction: a caseseries of fatal fall-asleep road accidents and their medicolegal outcomes. MedicalJournal of Australia 178, 396399.

Hkknen, H., Summala, H., 2000. Driver sleepiness-related problems, health status,and prolonged driving among professional heavy-vehicle drivers. TransportationHuman Factors 2, 151171.

Homel, R., 1983. Young men in the arms of the lawan Australian perspective onpolicing and punishing the drinking driver. Accident Analysis and Prevention15, 499512.

Horne, J., Reyner, L., 1999. Vehicle accidents related to sleep: a review. Occupationaland Environmental Medicine 56, 289294.

Horne, J., Reyner, L., 1995. Sleep-related vehicle accidents. British Medical Journal310, 565567.

House of Commons, Transport Committee, 2006. Roads policing and technology:getting the right balance. Tenth Report of Session 200506. The Stationery OfceLimited, London.

Joutsen, M., Lahti, R., Plnen, P., 2001. Criminal justice systems in Europe and NorthAmerica. European Institute for Crime Prevention and Controlafliated with theUnited Nations, Helsinki, Finland.

Knipling, R. R., Wang, J. S., 1994. Research note: Crashes and Fatalities Relatedto Driver Drowsiness/Fatigue. National Highway Trafc Safety Administration,Washington, D.C.

Langlois, P.H., Smolensky, M.H., Hsi, B.P., Weir, F.W., 1985. Temporal patterns ofreported single-vehicle car and truck accidents in Texas, U.S.A. during 1980-1983.Chronobiology International 2, 131140.

Laube, I., Seeger, R., Russi, E.W., Bloch, K.E., 1998. Accidents related to sleepiness:review of medical causes and prevention with special reference to SwitzerlandSchweizerische Medizinische Wochenschrift. Journal Suisse de Medecine 128,14871499.

Murray, J.B., 1998. Psychophysiological aspects of amphetamine-methamphetamineabuse. Journal of Psychology 132, 227237.

Ogden, E.J., Moskowitz, H., 2004. Effects of alcohol and other drugs on driver perfor-mance. Trafc Injury Prevention 5, 185198.

Ojaniemi, K.K., Lintonen, T.P., Impinen, A.O., Lillsunde, P.M., Ostamo, A.I., 2009. Trendsin driving under the inuence of drugs: a register-based study of DUID suspectsduring 19772007. Accident Analysis and Prevention 41, 191196.

Pack, A.I., Pack, A.M., Rodgman, E., Cucchiara, A., Dinges, D.F., Schwab, C.W., 1995.Characteristics of crashes attributed to the driver having fallen asleep. AccidentAnalysis and Prevention 27, 769775.

Partinen, M., 1982. Suomalaisten miesten nukkumistottumukset ja unihirit ennenvarusmiespalvelusta, varusmiespalveluksen aikana ja sen jlkeen (Sleeping

habits and sleep disorders in Finnish conscripts before, during and after militaryservice). Sotilaslket Aikakauslehti 57 (Suppl 1), 196.

Partinen, M., 2004. Vsymys ja nukahtaminen kuolemaan johtaneissa liikenneon-nettomuuksissa (Fatigue and falling-asleep in fatal trafc accidents). VALT,Helsinki.

Radun, I., Ohisalo, J., Radun, J.E., Summala, H., Tolvanen, M., 2009. Fell asleep andcaused a fatal head-on crash? A case study of multidisciplinary in-depth analysisvs. the court. Trafc Injury Prevention 10, 7683.

Radun, I., Summala, H., 2004. Sleep-related fatal vehicle accidents: characteristics ofdecisions made by multidisciplinary investigation teams. Sleep 27, 224227.

Radun, I., Radun, J.E., 2006. Seasonal variation of falling asleep while driving: anexamination of fatal road accidents. Chronobiology International 23, 10531064.

Radun, I., Radun, J.E., 2008. Vsyneena ajaminen. Mik on tilanne Suomessa? (Fatigueand driving: the situation in Finland). Liikenneturva, Helsinki.

Radun, I., Radun, J.E., Summala, H., Sallinen, M., 2007. Fatal road accidents amongFinnish military conscripts: fatigue-impaired driving. Military Medicine 172,12041210.

Rajaratnam, S.M.W., Jones, C.B., 2004. Lessons about sleepiness and driving fromthe Selby rail disaster case: R v Gary Neil Hart. Chronobiology International 21,10731077.

Reyner, L.A., Horne, J.A., 1998. Falling asleep whilst driving: are drivers aware of priorsleepiness? International Journal of Legal Medicine 111, 120123.

Roads and Trafc Authority, NSW, 2001. Road Trafc Accidents in New SouthWales 2000, Available from http://www.rta.nsw.gov.au/roadsafety/downloads/accident statistics dl4.html.

Roehrs, T., Roth, T., 2001. Sleep, sleepiness, sleep disorders and alcohol use and abuse.Sleep Medicine Reviews 5, 287297.

Sipinen, L., 1990. Varusmiehet kuolemankolareissa (Conscripts in fatal accidents).Liikenneturva, Helsinki.

Statistics Finland, 2007. Statistical Yearbook of Finland 2007. Statistics Finland,Helsinki.

Summala, H., Karola, J., Radun, I., Couyoumdjian, A., 2003. Kohtaamisonnettomuudetptieverkolla kehitys ja syyt (Head-on crashes on main road networktrendsand causation). Finnish Road Administration, Helsinki.

Summala, H., Mikkola, T., 1994. Fatal accidents among car and truck driverseffectsof fatigue, age, and alcohol-consumption. Human Factors 36, 315326.

Tasmania Law Reform Institute, 2007. Criminal liability of drivers who fall asleepcausing motor vehicle crashes resulting in death or other serious injury: IssuesPaper No.12.

Victorian Government, Department of Justice, 2004. Culpable and Dangerous DrivingLaws: Discussion Paper.

Convicted of fatigued driving: Who, why and how?IntroductionMethodDataProsecutor and court decisionsThe lawData analysis

ResultsBasis for a chargeTraffic incidents and consequencesAlcohol and drugsDrivers' testimonyAge, sex, and professionTime of daySeasonFalling asleep vs. fatigue

DiscussionLimitationsConclusion

AcknowledgmentsAppendix AReferences