Understanding patterns of behavioral health referrals for chronic disease in integrated care settings.

BACKGROUND: Multiple chronic conditions affect one quarter of Americans, with prevalence rising to 75% among those aged 65 and older. Early identification of behavioral comorbidities and the integration of behavioral health (BH) interventions are associated with improved outcomes, yet BH integration remains non-standard in primary care settings. METHODS: Secondary data analysis examining behavioral health (BH) referral patterns across four primary care clinics with integrated BH care. Adult patients (18 +) with a BH referral and/or a diagnosis of one or more chronic condition(s) are included. Relevant metrics are summarized using raw numbers, proportions, means/medians, odds ratios, and incidence rate ratios. RESULTS: A total of 31,317 patients and 13,157 BH referrals were included for analysis. 47% (n = 6,135) of referrals were submitted for patients with at least one chronic condition, yet only 9% (n = 1,170) specifically indicated medical condition management as the primary reason for the referral. Treatment initiation rates were relatively high (57%) among referred patients, with a median referral-to-completion time of 32 days. The majority of patients with one or more chronic conditions received no behavioral health referral (85%, n = 22,358). Compared with White/Non-Hispanic patients, African American and American Indian/Alaska Native patients had higher odds of a chronic condition with a BH referral than a chronic condition alone (OR = 1.2 and 2.0, respectively). Hispanic patients had higher odds of a BH referral alone or with a chronic condition (OR = 1.2 and 1.3, respectively) than a chronic condition alone compared to their Non-Hispanic counterparts, as did females compared to males (OR = 1.6). Patients with behavioral elements had higher overall healthcare utilization (IRR = 1.8) and were more likely to have COPD, weak/failing kidneys, arthritis, and hepatitis (OR = 1.8, 1.6, 1.4, and 1.4, respectively). CONCLUSIONS: Our findings suggest significant underutilization of integrated behavioral health services for chronic condition management, despite high treatment initiation rates among referred patients. Results highlight important differences in referral patterns based on specific demographic characteristics. Significant differences suggest that the added burden of behavioral elements is more likely to affect certain minority groups in our sample.