Medical errors, injuries, and avoidable complications cost billions of dollars annually. Suboptimal teamwork between physicians and nurses in hospitals contributes to higher rates of errors and adverse events. In current practice, nurses and allied health professionals are assigned to a ward, while physicians treat patients on multiple wards simultaneously. This antiquated structure reflects the pattern of "rounding" by community physicians who would come to the hospital to monitor their patients. The traditional, physician-centric model of inpatient care poses obstacles to physician-nurse communication, coordination, and quality improvement. We propose to evaluate an ongoing administrative pilot, the Emory Accountable Care Unit (ACU) Demonstration Project. In ACUs, hospitalists and nurses are assigned to unit-based teams. They round together, develop patient care plans, receive unit-level feedback, and jointly manage the ward. ACUs have all of the attributes associated with successful teams: repeated interaction, substantial decision authority, peer monitoring, and group evaluation. There are three ACUs in the Emory University Healthcare system. The first two ACUs were developed in 2010 in two medical wards in Emory University Hospital, a 579-bed tertiary care facility. A third ACU began operations in March 2011 at Emory Johns Creek Hospital, a 110-bed community hospital. Using a pre/post with concurrent controls study design, we will compare trends in patient outcomes between patients admitted to ACU wards and patients admitted to other medical wards. Based on our knowledge of patient intake practices, we believe that assignment of patients to ACU or non-ACU wards within hospitals mimics random assignment.
Under aim 1, we will verify the hypothesis by comparing patient characteristics between ACU and non-ACU wards and using the Hotelling test to assess the null that there are no differences in characteristics.
Under aim 2, we will use a time-to-event model to examine the impact of ACUs on the likelihood that patients experience any one of 7 adverse events. We hypothesize that ACUs will reduce the incidence of in-hospital mortality and other markers for poor quality care (e.g., central line associated blood stream infections, hospital acquired deep vein thrombosis). Demonstration of clinical benefit will set the stage for a cost study, studies to understand the impact on patient and employee satisfaction, studies to understand the mechanism of action, and, ultimately, dissemination to other health systems.
The failure to provide-high quality, patient-centered care to hospitalized patients is a leading contributor to poor patient outcomes. This proposal seeks to evaluate a novel model of inpatient care where physicians and nurses work in teams to improve quality.