A multivariate approach to the interpretation of patterns in homicidal and suicidal sharp force fatalities

Sammanfattning: A multivariate approach to the interpretation of patterns in homicidal and suicidal sharp force fatalities By Thore Karlsson, Karolinska Institute, Institution of Oncology Pathology, Department of Forensic Medicine. Sharp force injury is in Sweden as in many other European countries the leading cause of homicidal deaths. Sharp force suicide, constitutes almost as many fatalities and sometimes a difficult differential diagnosis. An objective of this thesis is to identify and quantify variables indicating suicide or hornicide. Further objectives are the detection and quantification of variables indicating conditions concerning the perpetrator and the emotional relation between perpetrator and victim. A consecutive study of all 316 sharp force homicides and 178 suicides submitted to medicolegal examination in Stockholm, Sweden in the 20 years period 1973-92 showed that sharp force homicide had increased both quantitatively (regarding male victims) and qualitatively in the respect that the mean number of sharp force injuries per victim had increased. Multivariate projective techniques ("forensiometrics") were used to create a model for the differentiation between sharp force homicide and suicide. This model, based on all such fatalities 1983-92, was validated by use of a test set. The multivariate model's sensitivity to detect homicide was 100 % and its specificity was 93 %. Numbers of homicidal sharp force injuries inflicted were found to vary with the interpersonal relationship between victim and perpetrator: a single injury was often inflicted in relationships based on alcohol, ten or more injuries among close relations (spouses or farnily) and 2-9 injuries in other relations. In sharp force homicides offender and victim as a rule were "alike", i.e. in a majority of cases a male killed a male, eight out of ten perpetrators and seven out of ten victims were inebriated by alcohol and immigrants were over-represented both as victims and perpetrators. About a quarter of the offenders had sharp injuries on their hands when apprehended. Victims of sharp force homicide and suicide differed significantly regarding age, blood alcohol concentration (BAC), venue, sharp objects used, presence of additional kinds of violence, reported suicidal ideation or written suicidal message, earlier parasuicide and anatomical area injured. Stab wounds to the anterior aspect of the chest with vertical or upwards/right slanting axis of entrance wound were significantly more frequent in homicide than in suicide, whereas no significant difference was seen regarding stabs with a horizontal axis of entrance wound. Key words: Forensic medicine, homicide, suicide, Multivariate statistical analysis, PCA, PLS-DA, trauma pathology, sharp injury. ISBN 91-628-2556-9