Nonlinear dynamics in suicidology: First steps to an early warning system

Résumé

Suicidology is characterized by a near exclusive assumption of linearity, i.e., the output (e.g., suicide risk) is proportional to a given input (e.g., risk factor levels, stress-inducing life events); the more the risk factors the greater the probability of attempting suicide. However, linear models of suicidal behavior have been limited in their ability to predict suicides and individuals often commit suicide after the occurrence of a seemingly minor life event. Furthermore, that facts that there are so many risk factors for suicide and that their dynamic interplay over time is so complex would suggest that the process is most likely nonlinear by nature. As a result, the prediction of suicidal behavior remains limited, despite decades of research. There have been repeated calls for the application of nonlinear dynamics (complexity theories, chaos theory, synergetics etc.) in suicidology, but the application of nonlinear dynamics to suicidology research has been limited due to the lack of technical possibilities and the lack of training of suicide researchers in theories of non-linear change and the mathematics of non-linear models. This session presents the results of the application of new technologies that may move nonlinear dynamics from theory to practice and to real-world phenomena. The Salzburg research group present results from their research with chronically suicidal patients using the Synergetic Navigation System, a web-based, real-time monitoring system. Mishara presents a theoretical explanation of the dynamics of suicidal behaviors and discusses the inherent limits of prediction using non-linear causal models. Goldney will discuss the presentations in relation to current challenges in suicidology research. This workshop highlights the importance of nonlinear dynamics in suicidology, outlines potential research goals, presents some research results with chronically suicidal patients, and discusses how a research agenda using nonlinear models can be developed. SUICIDOLOGIE COMPORTEMENT-SUICIDAIRE MODÈLE PRÉDICTION THÉORIE RECHERCHE