Provider characteristics related to antidepressant use in older people

To determine whether the characteristics of the usual medical care providers of older antidepressant users changed between 1986 and 1997 with the introduction of selective serotonin reuptake inhibitors.

Longitudinal study.

Five-county Piedmont area of North Carolina.

Stratified random sample of African-American (n=2,261) and white (n=1,875) community residents aged 65 to 105.

Sample members provided information on prescription medications, demographic and health status, and usual medical care provider (matched to North Carolina Health Professions Data Systems files to ascertain provider characteristics) in 1986/87, 1989/90, 1992/93, and 1996/97. Most (77.5%) named a provider (name unmatchable for 4.1%). Sample member characteristics were aggregated into probability (propensity) scores summarizing predisposing (demographic), enabling (medical care access), and need (health status) categories. Along with wave of study and whether a provider was named, these were entered as control variables in generalized estimating equation models that examined the association between provider race (white vs nonwhite), sex, age, location of practice, and primary versus specialist care and antidepressant use.

The characteristics of the usual medical care providers remained stable over the decade, although prevalence of antidepressant use increased. Two provider characteristics--race and area of practice (but not the interaction between them)--were significantly associated with patients' use of antidepressants. Patients of white physicians and of physicians with urban practices were more likely to use antidepressants.

Although use of antidepressants has increased over time, there has been little change in the characteristics of users' usual medical care providers.