• geriatric depression;
  • suicide risk;
  • depression in primary care

Objectives: To determine whether depression is treated differently in older and younger patients in primary care clinics.

Design: Administrative data were used to identify patients with a depression diagnosis code. The sources of data were baseline and 3-month follow-up surveys, the health plan electronic database, and chart audits.

Setting: Nine primary care clinics owned by a health maintenance organization in the Midwest.

Participants: The study sample (N=1,023) consisted of adult patients, aged 19 to 93, and was divided into six age groups, from young adult, under age 35, to old old, 75 or older.

Measurements: Independent variables were a series of dummy variables: age groups, baseline depression severity, sex, and incident depression. Outcomes were defined as care processes (assessment, resources) and improvement in depression symptoms (Center for Epidemiologic Studies—Depression scale short form). Univariate and multivariate logistic regression analyses were used to analyze patient characteristics, depression symptoms, and care process variables. Significance level was reported based on the chi-square test of probability, P≤.05.

Results: The adjusted response rates to the baseline and follow-up surveys were 69% and 82%. Elderly depressed patients were more likely than younger depressed patients to be widowed, have low levels of education, have fair or poor health, and have three or more comorbid health problems. There were no differences, by age, in number of depressive symptoms, antidepressant treatment, or recommendation for a follow-up appointment. Younger patients were more likely than older patients to have a new diagnosis of depression and to report being depressed most of the past year. Providers were only 6% as likely to ask old-old depressed patients about suicide risk, about one-fourth as likely to referral them to a mental health therapist, about one-fifth as likely to ask if they felt depressed, and one-twentieth as likely to ask about a problem with alcohol as they were with young-adult depressed patients. The old-old were about one-third as likely to report improvement in depression symptoms after 3 months as the young-adult patients.

Conclusion: This study suggests a possible pattern of underattention to the oldest depressed patients. Failure to ask about suicide risk is especially of concern because of the high suicide rate of depressed geriatric patients. The finding that physicians are less likely to focus on depression with old-old patients may be explained, in part, by the concept of invisibility of the ordinary.