Incomplete implementation of substance use treatment referrals in U.S. hospitals: What can be learned through a lens...

Hospitals can play a key role in identifying and treating substance use disorders (SUD), and a process known as Screening, Brief Intervention, and Referral to Treatment (SBIRT) exists to assist with this process. However, SBIRT is often implemented in ways that minimize disruption to operational workflows; screening is performed but not referral to treatment. This pattern of hospital practices aligns with and may be indicative of "decoupling," the symbolic adoption of new practices in suboptimal ways to comply with legitimacy pressures. This study seeks to investigate whether hospital characteristics consistent with those theorized to relate to decoupling are linked with incomplete SBIRT implementation, as viewing SBIRT through this lens may provide new insights into hospital systems and practices that contribute to care gaps. Using the National Survey of Healthcare Organizations and Systems 2017-2018 hospital survey, we regress incomplete implementation (systems for screening but limited referral) on hospital characteristics aligned with those hypothesized to influence two unique types of decoupling: goals-system decoupling (decoupling between an organization's stated goals and the systems designed to implement them) and system-practice decoupling (decoupling between the systems designed to guide daily activities and the activities as practiced). 49.05% of hospitals reported incomplete SBIRT implementation. Having more processes to disseminate best practices was associated with lower prevalence of incomplete implementation (adjusted prevalence ratio (aPR) = 0.91, p = 0.01). A less robust association (i.e., at 0.10 level) was also observed between decentralized decision and incomplete implementation (aPR = 0.81, p = 0.06). The pattern of associations is more consistent with characteristics theorized to relate to system-practice decoupling and suggests that investment in infrastructure to disseminate best practices and, perhaps, autonomy in physician decision making may improve implementation and reduce treatment gaps.