A preliminary statistical model for identifying repeaters of parasuicide

Publisher Website
Google Scholar
Abstract
This paper presents a statistical model constructed using logistic regression to identify those at high-risk of repeating parasuicide. The subjects in the study are Cork city residents who exhibited parasuicidal behaviour between I January and 30 June 1995. Repetition of the behaviour within six months of the index episode distinguishes repeaters from non-repeaters. The model was designed so that it could be used by non-clinicians and hence does not require information relating to psychiatric diagnosis or use of psychiatric services. The proportion of subjects correctly classified remained stable across a range of cut-point probabilities (mean - 86%. range: 83.9-87.5%). Using a cut-point of 0.2. 96% of repeaters and 81% of non-repeaters were correctly classified. Using 0.45 led to the correct identification of 81% of repeaters and 90% of non-repeaters. If these high levels of sensitivity and specificity are maintained in validation tests on future cohorts in Cork city then the model could form the basis of an intervention programme designed to prevent the repetition of parasuicide.