Are there any clinical characteristics of depression in elderly people that could be useful for case finding in general practice?

OBJECTIVES: The aim of this study was to identify clinical characteristics of depression in elderly people that could be useful for case finding in general practice. DESIGN: A cross-sectional study of clinical characteristics through review of medical records. SETTING: Herrhagen health centre, Karlstad, Sweden. SUBJECTS: Seventy-one persons with a high depressive score in a screening of depressive symptoms and an age-matched and sex-matched control group of 138 persons with a low depressive score. RESULTS: The high depressive score group had an increased relative risk for "mental health problems" (RR 3.4; CI 95% 1.7-7.2), "many contacts with the health care centre" (> or = 14/3 years) (RR 2.9; CI 95% 1.4-6.1), and prescriptions of benzodiazepines (RR 1.7; CI 95% 1.0-2.9). Two-thirds of those in the high depressive score group had at least one of these characteristics. However, three-quarters of those with any of these characteristics had a low depressive score. In our population of elderly people with an estimated prevalence of 10.2% the positive predictive value would be 21% and the negative predictive value 95%. CONCLUSION: General practitioners should suspect a possible depressive disorder in elderly patients with mental health problems, with frequent contacts with the health centre or with prescriptions of benzodiazepines. Despite the high occurrences of these prominent clinical characteristics in the high-score group, they did not unfortunately have sufficient discriminatory power to be useful for case finding