Background: Many patients admitted to the hospital are seriously ill and at risk of dying. Inpatient providers must be able to communicate to their patients the severity of their illness so that they may have realistic expectations and be able to make informed medical decisions.1 However, evidence suggests that physicians perform poorly in predicting prognosis and life expectancy for patients. 1,2 Hence, goals of care discussions between physicians and patients do not occur consistently and the content, detail, documentation, and timing of these discussions are also variable. Seriously ill patients are, thus, at an increasing risk of being exposed to potentially unwanted interventions at their end of life (EOL). 3,4

Purpose: The purpose of this project was to identify patients admitted to our hospital at high risk of 30-day mortality and to improve the quality of their advanced care planning at discharge. We offered identified patients a personalized Transitions of Care Toolkit (TOCTK) which included a consult to the palliative care team, team-based family meetings, advanced care planning documentation, and access to educational videos for the patient and family. We hoped to improve the patient’s education on their illness and address important concerns regarding end-of-life care for that patient.

Description: This project included all adult patients admitted to internal medicine hospitalist teams at Duke Regional Hospital (DRH). First, we identified patients who were admitted at DRH with a high 30-day risk of mortality by using the mortality risk predictor model developed by the Duke Institute for Health Innovation. Next, we discussed these patients at our daily Just In Time Palliative Care Huddle and developed a patient specific TOCKT. We then informed the primary inpatient provider of their patient’s 30-day risk of death and offered them the standardized TOCTK to help them support their patients in the hospital as well as at the time of discharge. If the provider agreed, we then implemented the TOCTK for the patient, together. Alongside this, we created multiple education sessions for our hospitalist team where we discussed end of life pain management, how to improve communication during advanced care planning and end of life discussions, and how to identify patients who might meet criteria for hospice care. These education sessions were done by our palliative care team members. We completed a pre and post implementation survey of need to our hospitalist team.

Conclusions: Using a mortality risk predictor model, we were able to identify 59 patients who were considered “high risk” of 30-day mortality on the hospitalist service at DRH in a period of 5 months. Of those 59 patients, 42% went on to have family meetings to discuss advanced care planning; 37% received a palliative care consult and direct advanced care planning and goals of care discussions with the multidisciplinary team; 16% had a change in their code status; 7% used educations videos as a resource to improve their understanding of disease process; 3% completed a MOST form with a member of the multidisciplinary team. In addition, we improved the awareness and understanding of advanced care planning documentation of our Hospitalist group. Given the initial success, these efforts have sustained and have now become part of routine workflow for the providers. We hope to expand this workflow to other units, including the ICU and surgical units at DRH.