Primary Care Providers’ Initial Treatment Decisions and Antidepressant Prescribing for Adolescent Depression

Abstract

Adolescent depression is a serious and undertreated public health problem. Nonetheless, pediatric primary care providers (PCPs) may have low rates of antidepressant prescribing due to structural and training barriers. This study examined the impact of symptom severity and provider characteristics on initial depression treatment decisions in a setting with fewer structural barriers, an integrated behavioral health network. A cross-sectional survey was administered to 58 PCPs within a large pediatric practice network. PCP reports of initial treatment decisions were compared in response to 2 vignettes describing depressed adolescents with either moderate or severe symptoms. PCP depression knowledge, attitudes toward addressing psychosocial concerns, demographics, and practice characteristics were measured. Few PCPs (25% for moderate, 32% for severe) recommended an antidepressant. Compared with treatment recommendations for moderate depression, severe depression was associated with a greater likelihood of child psychiatry referral (odds ratio [OR], 5.50; 95% confidence interval [CI], 2.47-12.2] p < .001). Depression severity did not affect the likelihood of antidepressant recommendation (OR, 1.58 [95% CI, 0.80-3.11] p = .19). Antidepressants were more likely to be recommended by PCPs with greater depression knowledge (OR, 1.72 [95% CI, 1.14-2.59] p = .009) and access to an on-site mental health provider (OR, 5.13 [95% CI, 1.24-21.2] p = .02) and less likely to be recommended by PCPs who reported higher provider burden when addressing psychosocial concerns (OR, 0.85 [95% CI, 0.75-0.98] p = .02). PCPs infrequently recommended antidepressants for adolescents, regardless of depression severity. Continued PCP support through experiential training, accounting for provider burden when addressing psychosocial concerns, and co-management with mental health providers may increase PCPs' antidepressant prescribing.