Psychiatric aftercare and suicide risk: a case-control study using blind rating

Blindly abstracted records of last episode of care together with aftercare records of 39 psychiatric patient suicides and their matched controls were rated blind for suicidal talk during aftercare, reduction of aftercare at last appointment and high-low suicide risk. Neither suicidal talk nor reduction of aftercare was confirmed as a predictor of suicide, although trends in the predicted directions were observed. Blind estimates of high-low risk correctly identified a significant minority (40%) of suicides but at the cost of misclassifying 60% as controls. Evidence is still lacking that clinicians blind to case identity may, from records, reliably distinguish a majority of suicides from their matched controls. Some implications for practice and research are discussed.