Later-life depressive disorders are a major public health problem in primary care settings. A validated screening instrument might aid in the recognition of depression. However, available findings from younger patients may not generalize to older persons, and existing studies of screening instruments in older patient samples have suffered substantial methodological limitations.
One hundred thirty patients 60 years or older attending 3 primary care internists' practices participated in the study. Two screening scales were used: the Center for Epidemiologic Studies—Depression Scale (CES-D) and the Geriatric Depression Scale (GDS). The Structured Clinical Interview for the Diagnostic and statistical Manual of Mental Disorders, Third Edition, Revised, was used to establish "gold standard" diagnoses including major and minor depressive disorders. Receiver operating curve analysis was used to determine each scale's operating characteristics.
Both the CES-D and the GDS had excellent properties in screening for major depression. The optimum cutoff point for the CES-D was 21, yielding a sensitivity of 92% and a specificity of 87%. The optimum cutoff point for the GDS was 10, yielding a sensitivity of 100% and a specificity of 84%. A shorter version of the GDS had a sensitivity of 92% and a specificity of 81% using a cutoff point of 5. All scales lost accuracy when used to detect minor depression or the presence of any depressive diagnosis.
The CES-D and the GDS have excellent properties for use as screening instruments for major depression in older primary care patients. Because the GDS's yes or no format may ease administration, primary care clinicians should consider its routine use in their practices.Arch Intern Med. 1997;157:449-454