Nonlinear dynamics and probabilistic behavior of adolescent suicidal outbreaks

There is increasing evidence of suicide contagion among teenagers. This phenomenon is further exacerbated by the availability of long-range communication channels like the Internet. The aim of this work is to develop a quantitative model linking contagion to the nonlinear dynamics underlying the imitation phenomena at work; to explore different scenarios and predict trends to be expected; and to propose prevention strategies. To this end, the methods of population dynamics and epidemiology are augmented by incorporating individual variability by means of Monte Carlo simulation. It is shown that contagion may take an explosive form which is dramatically accelerated in the presence of long-range interactions. Situations of undecidability can also be reached. The relevant time scales are identified and linked to two fundamental parameters. Applications in prevention in hospital, in school and in use of Internet are outlined. It is suggested that interpretations of surveys based on purely random scenarios need to be reassessed.