Decoding the Crime Scene: Exploring the Contribution of Psychopathology to Homicidal Behaviour

Background. In Italian criminal proceedings, forensic psychiatric assessment of criminal responsibility requires establishing a causal nexus between a mental disorder and the specific offence. Crime-scene behaviour may provide information relevant to reconstructing the offender's mental state at the time of the offence. Research on the crime-scene behaviour of mentally disordered homicide offenders has mostly drawn on psychiatric samples without a comparison group, and the Italian context remains largely unexplored. Aim of the study. The first aim was to investigate whether crime-scene behavioural patterns differed between offenders who underwent a forensic psychiatric evaluation of criminal responsibility and offenders who did not undergo such an evaluation. The second aim was to compare crime-scene behaviour across groups defined by forensic-psychiatric evaluation status and criminal responsibility. The third aim was to compare offenders with psychotic disorders and offenders with other psychiatric disorders. Materials and methods. The study included Italian cases of completed homicide, attempted homicide, and strage (killing at least two victims). Offenders who underwent forensic psychiatric evaluation formed Group 1. Offenders for whom no evaluation was requested formed Group 2. Crime-scene, post-offence, testimonial and clinical indicators were coded from forensic psychiatric reports and from judicial, investigative and autopsy documentation. Univariate analyses were followed by binary logistic regression. The regression included crime-scene and behavioural variables only, since clinical variables are inherently related to the likelihood of undergoing evaluation. The discriminative ability of the model was explored with ROC curve analysis. Secondary analyses compared crime-scene behaviour across groups defined by forensic-psychiatric evaluation status and criminal responsibility, and between evaluated offenders with psychotic-spectrum and other psychiatric disorders. Results. Eighty-nine cases were included, 51 in Group 1 and 38 in Group 2. Chaotic or incomprehensible scene, absence of flight from the scene, absence of actions directed at covering the offence, and sudden and unmotivated assault were independently associated with forensic psychiatric evaluation, whereas increasing duration of the assault was inversely associated with evaluation. The model showed high apparent discrimination (AUC 0.963, 95% CI 0.916–1.000). After Bonferroni adjustment, no crime-scene variable distinguished fully responsible offenders from those with diminished or absent criminal responsibility. Among evaluated offenders, absence of flight from the scene was the only variable more frequent in the psychotic-spectrum group. The composite indices did not differ between groups. Conclusions. Offenders who underwent forensic psychiatric evaluation differed from those who did not mainly in their conduct after the assault, and the intensity of violence did not in itself indicate psychopathology. Crime-scene behaviour did not reflect the outcome of the evaluation, consistent with a legal standard that requires a causal link between disorder and act. Single variables proved more informative than composite indices. The crime scene is best regarded as complementary evidence in forensic psychiatric evaluation. Its association with Group 1 membership reflects, in part, the circumstances under which an evaluation is requested.