Anxiety Treatment Preferences Among Veteran Primary Care Patients: Demographic, Mental Health, and Treatment-Related Correlates

Public Significance Statement Patient preferences for anxiety treatment in primary care are largely consistent with integrated behavioral health models of care. Introduction: Anxiety symptoms are common, yet undertreated, among primary care patients. Accommodating patient treatment preferences improves engagement and retention. In contrast to depression, little is known about primary care patients' preferences for anxiety treatment. Method: Participants were 144 veterans experiencing anxiety symptoms but not receiving psychotherapy who were recruited from primary care. Preferences for 11 anxiety treatment attributes (method; location; type; format; provider; frequency, length, and number of appointments; psychotherapy orientation; symptom focus; and topic/skill) and demographic, mental health (e.g., anxiety symptom severity), and treatment-related (e.g., psychotherapy history) variables were assessed via mailed survey. We used chi-square goodness of fit tests to identify patient preferences for each attribute and multivariate multinomial logistic regression models to explore demographic, mental health, and treatment-related correlates of treatment preferences. Results: Patient preferences were largely consistent with integrated primary care models, particularly Primary Care Behavioral Health, with a few exceptions. Patients preferred longer appointments (e.g., 45-60 minutes) and a longer duration of treatment (e.g., >= 13 appointments) than is typically offered in primary care. Several variables, particularly education level, perceived need for help, anxiety symptom severity, and attitudes toward psychotherapy, were repeatedly associated with preferences for various anxiety treatment attributes. Discussion: Results from this study suggest that patients tend to have distinct preferences for anxiety treatment in primary care that are largely consistent with common integrated primary care models. Results also identify several variables that may be associated with specific preferences, which may help match patients to their preferred type of care.