Fuzzy logic detection of medically serious suicide attempt records in major psychiatric disorders.

Clinical prediction of suicide is a complicated task. The focus for improved suicide risk detection is on the subgroup of individuals whose high suicide risk remains unrecognized by clinicians. We sought to evaluate the accuracy of Fuzzy Adaptive Learning Control Network (FALCON) neural networks, a nonlinear algorithm, in identification of this subgroup. The study sample included the Computerized Scale for risk of Suicide, including 21 suicide risk factors (including the target variable) drawn from 987 patient records, completed by staff clinicians during face-to-face interviews of hospitalized patients. FALCON evaluated all records in two steps: a) 612 for training and 375 for validation, and b) 887 for training and 100 for validation. The existence of previous medically serious suicide attempts (MSSAs) was chosen as the target variable because it is generally recognized as the strongest suicide risk factor. Sensitivity, specificity, and unknown answers among MSSA and non-MSSA were as follows: 612/375 FALCON, 91%, 85%, 11%, 15%; 887/100 FALCON, 94%, 82%, 20%, 14.5%, respectively. Trained FALCON, a nonlinear neural network, achieves respectable accuracy in detecting MSSA patients based on 20 suicide risk factors. Trained FALCON may therefore assist in identification of subgroup of individuals who remain unrecognized by clinicians and contribute to prevention of suicide.

Suicide risk factors and suicide vulnerability in various major psychiatric disorders.

PRIMARY OBJECTIVES: There are many known suicide risk factors (SRF) common to major psychiatric disorders, but their impact on suicide vulnerability remains unclear. We used FALCON (Fuzzy Adaptive Learning Control Network) to evaluate those impacts. METHODS: Staff psychiatrists completed computerized suicide risk scales (CSRS-III) including 21 SRF for 612 patients. Diagnoses were: schizophrenia, schizoaffective, major depression, anxiety disorder, bipolar affective disorder, personality disorder, organic brain syndromes, delusional disorder and other diagnoses. An optimal trained FALCON was obtained by running the network 10 times with 552 CSRS-III, validating with the balance. Medically serious suicide attempts (the vulnerability factor) served as the target variable. The significance of each variable in the trained network was determined by the magnitude of the change in output as affected by the consecutive change in all points of the variable input, then calculating the mean variance of all cases. The direction of influence was determined by the input on the entire scale of each variable, point by point, across all cases, then calculating the mean of all outputs. RESULTS: The impact and direction of influence of the various SRF differed for each diagnosis. CONCLUSION: Evaluation of the individual patient with his/her specific impact profile, determination of direction of influence of the corresponding SRF's may assist in increasing the accuracy of individual suicide risk assessment.