Dlin/Fischer Clinical Research Award: Impact of Co-locating Mental and Behavioral Health Providers with Primary Care Providers and Specialists on Patient Outcomes

2021 Dlin/Fischer Clinical Research Award

Background: There has been a movement towards co-locating/integrating physical health with mental and behavioral health (MBH) services since the Institute of Medicine advocated for such an integration (Medicine et al., 2006). Evidence from the literature supports that colocation improves health outcomes (Williams et al., 2006). However, a thorough investigation of the role of colocation and its impact on patient health is missing from the literature. We examine how colocation of MBH providers with primary care providers (PCPs) and specialists impacts health outcomes, where colocation is defined as sharing the same physical space (hospital or clinic). We also investigate how patient severity and continuity of care (CoC) moderate the relationship between colocation and outcomes.

Methods: We used data on diabetic patients, who suffer from chronic kidney disease and MBH illnesses from the Veterans Health Administration between October 1, 2002, and September 30, 2013. We used multivariate linear regression models to study the association between colocation and the following outcomes: number of MBH hospitalizations, associated length of stay (LOS) and 30-day readmissions; and suicide ideation/attempt. The models adjust for demographics, comorbidities, medications, labs, main events, and hospital characteristics. We aggregated data on a quarterly basis; each patient-quarter combination is considered an observation.

Results: The study sample included 577,813 observations and 73,055 unique patients, with an average age of 65 years and a high baseline prevalence of MBH conditions, with 47.8% of patients suffering from depression, and 67.8% using one or more MBH medications. We find that colocation is negatively correlated with all four outcomes; that is, the more often the patient sees their MBH provider that is colocated with their treating PCP or specialist, the better their outcomes. In examining the moderators, we find that colocation benefits sicker patients more, while those not severe benefit little. However, the moderating impact of continuity is mixed. We observe that increasing colocation is beneficial (in improving outcomes) when CoC is low but detrimental when CoC is high. In other words, colocation and continuity are “substitutes” when continuity is low but “complements” when continuity is high.

Conclusion: Colocation of MBH providers with PCPs and specialists is associated with improved outcomes. Our results offer insights into how healthcare organizations can develop targeted interventions, including strategies for co-locating providers that are aimed at specific set of patients, to yield the highest expected benefit.

References:
1. Institute of Medicine, Committee on Crossing the Quality Chasm, Adaptation to Mental Health, Addictive Disorders. Improving the quality of health care for mental and substance-use conditions: Quality Chasm Series. 2006.
2. Jane Williams, Steven E Shore, and Jane Meschan Foy. Colocation of mental health professionals in primary care settings: three North Carolina models. Clinical pediatrics, 45(6):537–543, 2006.