Psychiatric diagnoses are common in up to 50% of patients with incurable conditions. 1 The Canadian National Palliative Care Survey found that 13% of palliative care patients (N= 381) had a diagnosis of major depression. When patients with mild depression, dysthymia, and other depressive disorders were included, the total proportion rose to 44%. 2 Mitchell et al. conducted a recent meta-analysis of 94 psychiatric interview-based studies that assessed cancer patients for depressive disorders. The prevalence of major depressive disorder in both palliative and nonpalliative care settings was found to be 14%, rising to 24% when all forms of depressive illness (minor depression and dysthymia) were included. 3 However, differentiating depressive disorders from an appropriate grief reaction in the setting of a terminal illness may be difficult, 4, 5 and underdetection and undertreatment of the psychological and psychiatric morbidity developed in terminally ill patients are common. 6

Anhedonia, a lack of interest or pleasure in previously enjoyable activities, is a Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) criterion for depression, but it also could be attributed to decreased functional status because of the primary illness. It may be difficult to discern if the person is withdrawing from previously enjoyable activities because of depression or declining function and strength resulting from the terminal illness make the activities less enjoyable. Furthermore, many DSM-IV criteria commonly used to diagnose depression are related to somatic complaints (weight loss, fatigue, loss of appetite, and insomnia) that also are symptoms of terminally ill patients' underlying diagnoses and thus not very helpful in distinguishing depression from symptoms of their terminal illness. 7 Endicott 8 proposed that in the assessment of depression in cancer patients, somatic symptoms should be substituted (Table 1). Akechi et al. 9 compared the Endicott substitutions with DSM-IV criteria in 728 patients with cancer. Based on their assessment, the “fearfulness” and “brooding” items were suggestive of mild major depression, the “social withdrawal” item correlated with moderate severity of major depression, and the “cannot be cheered up” item indicated severe major depression. 9 Evidence of hopelessness, helplessness, worthlessness, guilt, and suicidal ideation may be better indicators of depression in terminally ill patients than neurovegetative symptoms. 4 Another significant problem in the terminally ill population is patients' short life expectancy and the significant length of time needed (weeks) for commonly used antidepressants (ie, selective serotonin reuptake inhibitor [SSRIs]) to take effect.