0 JAccid Emerg Med 1997;14:204-208

ORIGINAL ARTICLES

Violence: the relation between seriousness ofinjury and outcome in the criminal justice system

Jonathan Shepherd

AbstractObjective-To investigate the relation bet-ween injury severity in assault, offenceseriousness, and outcome in the criminaljustice system.Methods-Prospective random sample of93 assault victims who attended hospitalwere interviewed and examined and fol-lowed through the criminal justice system.Patient and police interviews were carriedout at Bristol Royal Infirmary accidentand emergency and outpatient depart-ments, wards, and in Avon police stationsand criminal courts. Severity of injury(injury severity score [ISS] assessed bythe major trauma outcome study group),offence seriousness (Delphi panel ofcriminologists), and outcome in thecriminal justice system were recorded.Results-Median ISS was 4 (range 1-17).There was no significant correlation bet-ween ISS and outcome in the CJS (r. =

0.07). There was a weak correlation bet-ween offence seriousness and outcome (r.= -0.24, P = 0.019).Conclusions-Outcome in the criminaljustice system was not predictable frominjury severity scores and was only weaklylinked to offence seriousness. Contactswith medical services are opportunitiesfor protecting those at risk of violence.(T Accid Emerg Med 1997;14:204-208)

Keywords: assaults; criminal justice; outcome; violence

Department of OralSurgery, Medicine andPathology, UniversityofWales College ofMedicine, CardiffJ Shepherd

Correspondence to:Professor Jonathan ShepherdPhD, Department of OralSurgery, Medicine andPathology, University ofWales College of Medicine,Heath Park, Cardiff, CF44XY

Accepted for publication28 February 1997

Research on the links between offence serious-ness, criminalisation, and punishment in rela-tion to violence has recently been facilitatedthrough access to injured victims who attendhospital accident and emergency (A&E) de-partments." Importantly, this has allowedstudy of the relation between violent behaviourand subsequent attempts at criminalisation. Ithas also allowed an investigation of the factorswhich-on both sides of the Atlantic-lead tounderreporting and underrecording of of-fences. Past research on the correspondencebetween offence seriousness and criminalprosecution has shown that at every stage ofthe criminalisation process cases are filteredout. This may be because of the victim's deci-

sion not to go to the police, or because of thepolice response, both immediately and in sub-sequent investigations, or because of events incourt.

In Britain, only about 25% of assaults result-ing in the need for treatment in the emergencydepartment are known to other agencies3 andsubstantial numbers of crimes of all types gounrecorded.4 The research reported in thispaper was carried out as part of a project toinvestigate the relation between seriousness ofoffence and outcome in the criminal justicesystem. The main purpose was to quantify therelation between the severity of injury and out-come in terms of reporting, police investiga-tion, conviction, and sentencing of assailants.This research programme has resulted inreports in the criminology literature,' 2 butthese have not addressed the issue of injuryseverity, assessed using modern, objectiveinjury severity scoring. The various drop outpoints identified in the previously reportedresearch are summarised in fig 1.

MethodsNinety three randomly selected assault victimswho attended an inner city A&E departmentwere interviewed and their cases followedthrough the criminal justice system. Seventyeight victims were male (median age 31 years;range 18 to 76 years) and 15 were female(median age 28 years; range 18 to 64 years).This sample was representative of the hospitalcatchment population in ethnic terms. Thesample was necessarily limited in size becauseof the need to follow cases through the report-ing, recording, prosecution, and sentencingprocess. The research required detailed inter-views with all criminal justice personnelinvolved as well as with the victims themselves.Injury severity scoring was carried out by theUnited Kingdom major trauma outcome study(MTOS) group from A&E injury records, andoffence seriousness was ranked by a panel ofcriminologists.Medical injury severity scoring methods

provide a robust and objective means ofcomparing those injured. A variety of scalesand scoring methods is available, which arebased either on anatomical or physiologicalvariables or on a combination of these.5Although there is still undoubtedly room for

204

on February 1, 2022 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.4.204 on 1 July 1997. Dow

nloaded from

Seriousness of injury and outcome after assault

Dafindantimprisoned

Defendantconvicted

Case

prosecuted

Policenvestigated

Assault

reportedto police

Assault notI IIreported to 1 4 7 10 13 16 19 22

policeSeriot

2 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76

usness of offence (1 = most serious, 93= least serious)82

Figure 1 Criminalisation of assault cases matched against seriousness of offence.

improvement, and some scoring methodsdemand medical tests and information notlikely to be available to the courts, injury sever-ity scores have become an established and reli-able means of assessing trauma patients in hos-pital and other medical settings, particularly inrelation to survival. Recent applications of thisscoring system have included the investigationof relative merits of large dedicated traumacentres and smaller casualty departmentsoutside major centres6 and calculation of com-pensation awards.78The injury severity score was developed in

an attempt to quantify severity, taking intoaccount the number of body areas involved inindividuals with multiple injuries.9 Examples ofrecent applications to forensic medicine andresearch are the assessment of the relationbetween blood alcohol concentration andinjury severity in victims of violence andaccidents, and investigation of the associationbetween conscious level and injury severity.'0The ranking of offences in order of serious-

ness took account of physical injury (assessedby lawyers); potential for more serious injury;degree of threat; use of a weapon; likelihood ofemotional or psychological harm; duration ofassault; and provocation by the victim. In a

preliminary validation exercise, 20 lawyerswere asked to rank the assaults, taking intoaccount the variables listed above.' There wasconsensus and high interjudge reliability inrespect of assaults in the upper and lower quar-tiles, but less consensus elsewhere. This find-ing influenced subsequent grouping of data(see fig 4).This research depended on developing a

scale of offence seriousness: a difficult andcontroversial exercise in view of the unquanti-fiable nature of many of the factors listedabove." For the purpose of this research, thecriminal law hierarchy of seriousness, (grievousbodily harm with intent; malicious wounding;actual body harm; battery; and assault) was toocrude to be useful. Some past criminologicalresearch was helpful, however. Surveys havebeen conducted on the general public'2 andcrime victims" in which respondents were

asked to rank offences by seriousness. Recent

research has investigated both the immediateimpact of crime, physical injury, andemotional/psychological harm,'4 and its longerterm effects such as depression, distrust, anddisruption of lifestyle.'5 In the United States,victim assessments of harm are used in manyStates, where "victim impact statements" areintroduced before sentencing." " In variousjurisdictions over the past decade, sentencingcommissions charged with developing sentenc-ing guidelines have attempted to rank offencesin order of seriousness (for example, the Min-nesota Sentencing Commission).For the purpose of this research, the ranking

system sought to reflect harm caused and atthe same time to reflect the assault as an eventin context. Therefore, while the location of theassault and the relation between victim andassailant were taken into account, extraneousfactors such as the prevalence of a particularcategory of assault or the offenders' previousconvictions were not taken into account.

Statistical analysis was carried out using theSpearman correlation coefficient. Because theresearch reported here was principally con-cerned with establishing the strength of anyrelation between injury severity, offence seri-ousness, and outcome, and because compo-nents reported elsewhere'2 depended on quali-tative research methods, a bivariate analysiswas appropriate. Clearly, other variables couldbe taken into account in any subsequent studyto control, for example, for the victim/offenderrelationship.

Table 1 Site of injury by type (number of victims/patients= 93)

Number of injuries by type

Site of injury Fractures Lacerations Bruises Total

Head 3 8 34 45Face 23 24 30 77Upper limb 8 2 13 23Lowerlimb 2 0 8 10Chest 0 2 13 15Abdomen 3 3Total 36 39 98 173

*Includes 19 brain injuries not associated with scalp lacerationsor skull fractures.

if II

205

i

on February 1, 2022 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.4.204 on 1 July 1997. Dow

nloaded from

Shepherd

Injury severity score(ISS) groupings- ISS 1-4

ISS = 5-8

12 = ISS=917m

3

Reported/ Investigated/ Prosecuted/not Convicted Imprisonednot investigated not convicted (Non-custodial

prosecuted sentence)

(r, = 0.5177; P << 0.0001; fig 4). Sites of injur-ies are presented in table 1. Thirty sevenpatients sustained multiple injuries. Injurieswere predominantly facial or of the head. Thenext most frequent site of injury was the upperlimb. Not surprisingly in a British context,there were no firearm injuries and none of theinjured subsequently died of their injuries,though 24 required surgical operations torepair injuries. Only seven of the 10 offencesgiving rise to the highest injury severity scorewere investigated by the police, but eight out ofthe 10 producing the lowest scores were notinvestigated.

In 19 cases (22%) there was a failure toinform the police in the first instance; in a fur-ther 20 cases (23%) no formal complaint wasmade. In 11 cases (17%) the assault wasbrought to the attention of the police but thepolice were unable to identify the assailant oramass sufficient evidence against him. Fortyfive per cent of those investigated were notprosecuted, 18% of those prosecuted were not

Figure 2 Relation between injury severity score and outcome.

The patients were interviewed by one of ateam of three research workers during theirinitial attendance in the A&E department, intheir hospital ward, when they returned forsubsequent outpatient treatment, and at home.Where the assault had been reported to thepolice, the relevant police officers were inter-viewed, and where prosecution followed, allcourt hearings were attended. Of the 29 casesprosecuted, 16 (55%) were dealt with in theMagistrates Court and 13 (45%) in the CrownCourt.

ResultsMedian injury severity score was 4 (range 1 to17). There was no evidence of correlationbetween injury severity score and outcome (r,= 0.07; 95% confidence interval -0.19 to+0.27; P = 0.52; fig 2), but a weak correlationbetween offence seriousness and outcome (r, =-0.24; 95% CI -0.042 to -0.426; P = 0.019;fig 3). Offence seriousness and injury severityscore were highly significantly related, however

Injury severity score(ISS) groupings

- ISS 1-4

-! SS = 5-8

= ISS=9-17

9

Cffencesranked 16-30 inseriousness

-Offences Offences Offences

ranked 31-45 ranked 46-60 in ranked 61-75 inin seriousness seriousness seriousness

Offence seriousnessFigure 3 Relation between severity of offence and injury severity score.

_20

0az000

4-015 10.0Ez

0Not reported

am004-0

0

Ez

10

0

5IL

15 mostsevereoffences

17 leastseriousoffences

206

20r- on February 1, 2022 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.4.204 on 1 July 1997. Dow

nloaded from

Seriousness of injury and outcome after assault

10r1 = 15 Most severe offences2 = Offences ranked 16-30 in seriousness3 = Offences ranked 31-45 in seriousness4 = Offences ranked 46-60 in seriousness5 = Offences ranked 61-75 in seriousness6 = 17 Least serious offence

Outcome

Figure 4 Relation between offence seriousness and outcome.

convicted, and in only three cases (5%) were

assailants imprisoned (fig 1).

DiscussionThis study is the first to correlate objectiveinjury severity scores with outcome in thecriminal justice system. There was no statisti-cally significant relation between injury sever-

ity and outcome of the criminal justice processand only a weakly significant link betweenoffence seriousness and outcome. The formerfinding is perhaps not surprising in view ofother important and legitimate dimensions ofoffence seriousness. This means that manyassailants were not brought to justice and, pre-sumably, rarely deterred from causing futureinjury. Therefore, as with other non-accidentalinjury, mechanisms involving A&E depart-ments need to be developed to ensure that vic-tims, if they wish, are protected from furtherharm and are allowed to report violence at an

early stage. Although the study reported heredid not investigate the roles of A&E depart-ment staff, it appeared that they maintained a

passive role.Through gaining access to victims of violent

crime in a setting outside the criminal justicesystem, this study provides evidence that injuryseverity plays little part in determining crimi-nal justice response to violence. Although the"injury against society" model is pre-eminentin the criminal law, this research found littleevidence that injury severity per se was

influential. Although this research has shownthat injury severity has little influence in thecriminal justice system, it has found thatthroughout the criminal justice process there isgreat dependence on the victim to make a

complaint."2 The distinction between the needto complain and the need to report offences isoften lost on victims. In police recording, pros-ecution, and conviction, the people who carryinjuries and their physical and psychologicalscars are crucially important. In this context,from a medical perspective it is surprising thatthe importance of injured victims is notacknowledged. One eminent lawyer has re-

cently written that "whether the particular vic-tim's interest should count for more than thoseof any other member of the community mustremain open.""8The findings of this research have important

implications, in particular relating to the rightsand responsibilities of the injured. In parts ofthe United States, the burden of taking theircases through the justice system has been takenfrom the shoulders of victims in cases ofdomestic violence, though whether this changeis increasing deterrence and victim protectionis still being debated.'9 In Britain the nationalcharity Victim Support has introduced victim/witness support in many Crown Courts, whichvictims have found effective. A comprehensivesupport service based in A&E departmentsmight help increase the rate of conviction ofoffenders, empower those who find it hardestto seek the law's protection, and also result inthe "injury against society" concept beingmore consistently applied.There are parallels here between the crimi-

nal justice system and public health medicine.Both are concerned with deterrents (healthwarnings) incapacitation (isolation) and reha-bilitation. Both are concerned with future as

well as past behaviour. Although it is easy totake the parallel too far, it does suggest ways inwhich reporting might be enforced. Althoughdoctors have a legal and ethical responsibilityto report some diseases, so that the publichealth and individual citizens can be benefited,there is no such legal responsibility in relationto the reporting of injuries, however serious.The research reported here has shown howsome injuries which were not reported, or theassailants prosecuted, are nevertheless lifethreatening. In public health terms, if thecauses of other such serious illnesses were notcompulsorily reported and investigated therewould be a public outcry. It is surprising in thiscontext that similar procedures-underpinnedwith appropriate legislation-do not apply tothe reporting of serious physical and psycho-logical injuries occasioned by violence, particu-larly where there is a risk of future harm.

8

co0co

a0'4-0.0Ez

6

4

2

0

207

on February 1, 2022 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.4.204 on 1 July 1997. Dow

nloaded from

208 Shepherd

Key messages* Violence which results in the need for hospital treatment is often

not investigated by the police and assailants are rarely brought tojustice.

* For this group of patients there is a poor correspondence betweeninjury severity, offence seriousness, and outcome in the criminaljustice system.

* This is because the criminal justice process depends almostentirely on the injured themselves, who are often afraid of repris-als, habituated to violence, and have a continuing relationship withtheir assailant.

* There is an urgent need to develop AED/criminal justiceapproaches to support victims who attend hospital for treatment,to allow them to report offences if they wish so that assailants canbe brought to justice and to reduce the likelihood of further harm.

A recent ethical debate highlights some ofthe problems associated with A&E depart-ments taking a more proactive role.20 Increasedviolence in the A&E department itself, how-ever, and the realisation that many of theinjured also have criminal records,2" is leadingto a more organised police presence-albeit forthe protection of staff-giving rise to increasedopportunities for patients to report offencesaffecting them. It is important, however, thatthe development of police outposts in A&Edepartments is accompanied by safeguards toprotect patients' interests and confidentiality.This is an opportunity to revisit guidelinesabout the reporting of violence to the policeand, with police liaison, develop codes of prac-tice to fit local needs, particularly in relation toknife wounds, firearm injuries, and life threat-ening injuries. Since so many violent offencesare not recorded and since the police dependon the injured making a complaint, it is impor-tant for A&E department receptionists andnurses to give patients an opportunity to reportoffences at an early stage if they wish to,perhaps by means of a police "hot line"situated in the A&E department waiting area.

This research was supported by a project grant from theEconomic and Social Research Council. Mrs Maralyn Wood-forde and Professor David Yates of the Major Trauma OutcomeStudy kindly coded injury severity and Dr R Newcombeprovided statistical advice. I thank my coworkers Gwyn Davis,Antonia Cretney, and Chris Clarkson for advice.

1 Clarkson C, Cretney A, Davis G, Shepherd JP. Assaults: therelationship between seriousness, criminalisation and pun-ishment. Criminal Law Review 1994;jan:4-20.

2 Cretney A, Davis G, Clarkson C, Shepherd JP. Criminalis-ing assault. Br J Criminol 1994;34:15-29.

3 Shepherd JP, Shapland M, Scully C. Reporting of violentoffences to the police. Med Sci Law 1989;29:251-7.

4 Mayhew P, Elliot D, Dowds L. The 1988 British crime survey.Home Office research supplement. London: HMSO, 1989;111:16.

5 Ali T, Shepherd JP. Injury severity assessment. Br J OralMaxillofac Surg 1994;32:13-18.

6 Gustafson DH, Fryback DG, Rose JH. An evaluation ofmultiple trauma severity indices created by different indexdevelopment strategies. Med Care 1983;21:674-91.

7 Shepherd JP, Richmond P, Miers D. Assessing criminaldamages: a medical model. New Law J 1994;4 Feb: 162-4.

8 Law Commission. Damages forpersonal injury: non-pecuniaryloss. London: HMSO, 1995:122.

9 Baker SP, O'Neil B, Haddon W. The injury severity score: amethod for describing patients with multiple injuries andevaluating emergency care. J Trauma 1974; 14:187-96.

10 Brickley MR, Shepherd JP. Relationship between alcoholintoxication, injury severity and Glasgow coma score.Injury 1995;26:311-4.

11 Von Hirsch A. Past or future crimes: deservedness anddangerousness in the sentencing of criminals. Oxford: OxfordUniversity Press, 1985.

12 Sellin T, Wolfgang M. The measurement of delinquency. NewYork: John Wiley, 1964.

13 Sparks R, Genn H, Dodd D. Surveying victims. Chichester:John Wiley, 1977.

14 Lurigio A, Resick P. Healing the psychological wounds ofcriminal victimisation. In: Lurigio A, Skogan W, Davis R,eds. Victims of crime. Newbury Park, CA: Sage, 1990.

15 Miers D. The responsibilities and the rights of victims ofcrime. Modern Law Rev 1992;55:2-14.

16 Henderson L. The wrongs of victims' rights. Stanford LawRev 1985;37:956-99.

17 Kelly D. Victim participation in the criminal justice system.In: Lurigio A, Skogan W, Davis R, eds. Victims of crime.Newbury Park, CA: Sage, 1990.

18 Ashworth A. Punishment and compensation. Oxford J LegalStud 1986;6:1-8.

19 Morley R, Mullender A. Hype or hope? The importation ofpro-arrest policies and batterers' programmes from NorthAmerica to Britain as key measures for preventing violenceagainst women in the home. Int J Law Fam 1992;6:265-88.

20 Shepherd JP, Morley R, Ashead G, Gillett G, Knight MA.Ethical debate: should doctors be more pro-active as advo-cates for victims of violent crime. BMJ 1995;311:1617-21.

21 Rivara FP, Shepherd JP, Farrington DP, Richmond P, Can-non P. Victim as offender in youth violence. Ann EmergMed 1995;26:609-14.

on February 1, 2022 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.4.204 on 1 July 1997. Dow

nloaded from