To the Editor: Dr. Warner's letter highlights critical issues in elderly suicide prevention. Despite efforts in prevention at a national level, one American is lost to suicide every 16 minutes. The toll is particularly high for men over the age of 75 years, who have a suicide rate 3.3 times higher than that of the general population, reaching 35.7 per 100,000 individuals (1). However, these numbers are underestimates because of underreporting.

Despite the alarming statistics, little is directly known about prevention interventions in later life. Dr. Warner points out several risk factors for suicide. However, there is limited direct evidence that addressing these factors (e.g., treating depression and pain, enriching the social network of persons at risk, removing firearms) reduces the incidence of suicide. The PROSPECT study focused on patients 60 years of age and older (75—84 years: N=155; ≥85: N=30) and used suicidal ideation and depressive symptoms as a proxy for suicide prevention. We selected this strategy for two reasons. First, depression and suicidal ideation are risk factors for suicide. Second, there is face value in ameliorating depressive symptoms and thus reduction of suffering and family disruption and improved outcomes of several comorbid medical disorders. Showing that the PROSPECT intervention works in primary care patients is important because two-thirds of depressed older adults are exclusively treated in the primary care sector. Nonetheless, the PROSPECT study provides no more than indirect evidence of the effectiveness of an intervention for suicide prevention.

Errors in the assessment and treatment planning of elderly suicidal patients can be fatal. Among elderly persons, there are only four suicide attempts for every completed suicide, but there are 100 to 200 attempts per completed suicide in individuals aged 15—24 years (1). Assessment difficulties arise, to a large measure, from the older patients' reluctance to share thoughts on suicide and from poor acceptance of suicide risk by patients' families. Much of the difficulty in treatment planning arises from the absence of direct knowledge of the effectiveness of clinical interventions specific to elderly suicide prevention. Another reason is the expectation by clinicians and families that suicide can be prevented in most, if not all, cases. This unrealistic expectation is in part generated by an emotional reaction caused by stigmatization of suicide and often leads to treatment measures that are either idiosyncratic, overly restrictive, or both.

We believe that direct studies of interventions for elderly suicide prevention and systematic efforts to destigmatize suicide are two initiatives likely to make a difference. Currently, our approach in the reduction of suicide risk must rely on indirect findings of factors associated with high suicide rates. However, definitive guidance to clinicians can only come from effectiveness studies of specific interventions targeting reduction of suicide itself rather than reduction of suicide risk factors. Destigmatization of suicide and mental illness at the community level and during clinical interactions with patients and families is critical. It can increase the accuracy and promptness of reporting suicide thoughts and related symptoms by patients and families, set appropriate outcome expectations, and allow clinicians to make informed and rational treatment plans.