Prior studies have examined the predictors of poor adherence [2], [3], [4], [5] and have tested various interventions aimed at improving patient adherence with antidepressant treatment [1], [6]. Notably, collaborative care models [7], [8] have demonstrated benefit by intervening on both clinician and patient level barriers to adherence. In spite of these gains, adherence to antidepressant treatment for most patients remains inadequate. Primary care physicians need better tools to identify patient attitudes and preferences for treatment that can then inform strategies for patient-specific education about antidepressants. Treatment strategies that incorporate patient preferences increase the likelihood that a patient will enter treatment, adhere to prescribed regimens, [9] and show clinical benefit [10], [11]. The determinants of patient treatment preferences for depression, however, are less well studied.

Learning theory suggests that a patient's positive past experience with depression treatment should increase motivation to seek the same or similar treatment [12]. Similarly, social cognitive theory predicts that successfully accomplishing a task should boost self-efficacy to accomplish similar tasks in the future [13], [14], [15]. Moreover, social cognitive theory posits that vicarious experience has similar effects on self-efficacy. In this context, we are considering vicarious experience to be the experience, observed by the patient, of a friend or family member with depression. Our group reported previously that physicians' vicarious experiences with depression treatment influence their attitudes and management [16], [17]. In this study, we examined whether patients' vicarious experiences with depression treatment, i.e., having a close friend or family member who had undergone treatment for depression, would lead to a more favorable attitude towards treatment.

Separate analyses were carried out for respondents with versus without a personal history of depression. In both analyses, we controlled for current depression symptoms and socio-demographic factors. In respondents both with (coefficient 0.90, p=0.15) and without (coefficient 0.96, p=0.005) a personal history of depression, having a vicarious experience with depression, that is, knowing a friend or family member treated for depression, was associated with a more favorable attitude towards antidepressant medications.

To further delineate the relationship between vicarious experience and treatment preferences, we then performed a more detailed regression to examine separately the influence of friend and family experience. Results are presented in table 2. For respondents with a personal history of depression, having a family member who had been treated for depression was associated with having a positive attitude toward antidepressant medications (coefficient 1.08, p=0.03). However, among respondents with no personal history of depression, having a friend who had been treated for depression was associated with having a positive attitude toward antidepressant medication (coefficient 0.77, p=0.03). Additionally, a history of prior treatment with antidepressant medication is strongly associated with having a more favorable attitude to antidepressants. Interestingly, a higher PHQ-9 score is negatively associated with attitudes towards anti-depressants, suggesting that those who are more depressed are less favorably inclined towards medication.

Patients who have taken antidepressant medication in the past have a highly favorable attitude towards antidepressants. This is consistent with other work which has shown an favorable attitude towards treatment for those who have been prescribed or are currently taking anti-depressants [20],. Prior research suggests that those with positive attitudes towards antidepressants are more likely to be adherent [21] and that as treatments more closely match patient preferences, adherence is increased [10].

Interestingly, we also found that higher scores on the PHQ-9 are associated with more negative attitudes. As past work has shown that those with negative attitudes towards treatment have decreased adherence [22], this suggests that those who might benefit most from treatment, those most depressed [23], might be a group at particular risk for non-adherence. Every effort should be made to engage these patients in evidence based therapies.

Busy clinicians, especially in primary care, often face multiple demands for their time. We feel the ideal time to discuss vicarious experiences with depression treatment with a patient is when a clinician is considering initiation of treatment. This can be incorporated into routine counseling and anticipatory guidance around starting a new medication. Open-ended statements encourage patients to share details about their own experiences. For example, clinicians could say “Do you know anyone who has had depression?” or “Tell me about people you have known with depression”. Once obtained, this information can help identify patients at greater risk for non-adherence and be used to tailor patient-specific education about treatment.

It is curious that patients with a personal history of depression are influenced differently by their vicarious experience compared to those with no personal history. Patients with a personal and family history of depression may view their condition as more attributable to heredity and thus be open to a biomedical perspective, one that views pharmacological treatment favorably. In contrast, a person with no history of depression may see the depression experiences of a friend as being more informative than the experience of a family member. It should be noted, however, that although the association between attitudes toward antidepressant medications and having a family member who had been treated for depression for this group of respondents missed statistical significance, the relationship was in the anticipated direction. How vicarious experience mediates attitudes toward treatment for depressive symptoms is unknown. One possible mechanism may lie through the reduction of stigma. Prior studies of vicarious experience have noted that people who have a friend or family member with depression rate lower on measures of stigma (such as regarding depression as a ‘real’ illness, or feeling that a person with depression could “get better if they wanted”) than those without such an experience [24].

This study has several limitations. First, our outcome was a psychometric measure of attitudes towards antidepressants derived from survey data. It was not validated or assessed for reliability in other populations, and was not a clinical measure. Thus we cannot be sure how the associated positive regard will translate in the office setting. Second, our data are cross-sectional and thus can demonstrate an association between vicarious experience and treatment attitudes, but not a causal link. Third, compared to the BRFSS survey sample as a whole [25], which is representative of California, our sample was generally older and had higher income than the general population. Although we controlled for age and income in our analysis, this may limit generalizability to other settings.

Appendix S1