MGH Internal Medicine Associates : primary care redesign

Abstract

Internal Medicine Associates (IMA) is the largest primary care practice at Massachusetts General Hospital (MGH) with over 40 attending physicians, 60 residents, and 80 support staff that deliver care to more than 30,000 patients. The IMA is structured into seven pods that act independently to serve patients. Each pod consists of patient care providers and support staff that work collaboratively in a team structure. In particular, providers and medical assistants work closely together during the clinical session to meet patient needs. A lack of standardization in the practice's operations has contributed to inefficiencies that add to a sense of overload and burnout with the medical assistant staff A detailed study of providers' clinical schedules revealed that individual clinical sessions are highly variable in terms of the number of concurrent clinical sessions per pod, session length, and number of patient appointments booked during this time. Providers in the IMA are part-time and create their clinical schedules based on personal preference and coordination with their other MGH related commitments. Variability in the schedule arises from many systematic, predictable, and unpredictable sources. Additionally, as part of a teaching hospital, IMA supports the educational training of over 60 Internal Medicine residents who hold a varying number of clinical sessions per week, depending on specific requirements of their residency program. Coordinating and supporting provider presence consumes many resources, impacts medical assistant workload, and adds to variability within the practice. The project develops an optimization model to level-load the expected workload on medical assistants and other members of the medical care team by determining the clinic schedules of providers. The expected workload is measured by the number of concurrent sessions and expected number of patient visits per hour. The project has developed an optimization model to suggest changes to the clinic schedule. Specifically in Pod 2/3, by strategically shifting 19.5% of provider sessions, we can achieve an 83% improvement in variability, as measured by the difference between maximum and minimum expected workload. Similar results are modeled for all pods in the IMA. The team has identified a pilot pod to test the model and is