Suicide risk factors in alcohol abuse.

The "clinical model" approach to estimating suicide risk assumes that persons sharing specific attributes will also share certain indicators of vulnerability to suicide. This would warrant an empirically derived risk assessment scale applicable only to persons with those attributes. Nine hundred seventy eight persons at risk for suicide who met our criteria for alcohol abuse were interviewed at length and followed for two years, during which 53 (5.5 percent) committed suicide. Eleven variables which best differentiated those who suicided from those who did not in an index set were then applied to an independent validation set. Clinical implications of the findings are discussed.