Background: Misalignment of goals of care at the End-of-Life exposes patients to risk and the health care system to considerable costs. A lack of end of life conversations can lead to unrealistic patient expectations, patient harm, and multiple hospital readmissions. This study aims to identify physician perspectives on the barriers to initiating end of life discussion in the acute inpatient setting.
Methods: This was deemed IRB exempt as part of an institutional quality improvement effort. Twenty-two general medicine attending physicians and hospitalists were interviewed about patients that they discharged from the University of Virginia Health System during a one month period. Each physician was asked to identify whether or not they would be surprised if the patient in question were to pass away in the next twelve months. If the physician was not surprised, they were asked whether they had a goals of care, hospice or palliative care discussion with the patient. If they did not, they were prompted to identify barriers to having this discussion. Each physician response was noted, and then sorted into thematic groups. Seven umbrella barrier terms emerged. Responses were analyzed using the Pareto Principle.
Results: A total of twenty-two physicians were interviewed about a total of 222 discharged patients. The average age of the patients was 60.27 years old, with 112 males and 110 females. Of the 222 discharged patients, physicians would not be surprised if 33.3% (74/222) of those patients were to pass away in the next 12 months. Of the aforementioned 33.3% of discharged patients, physicians did not have a goals of care, hospice, or palliative care discussion with 54.1% (42/74) of these patients. Physician identified barriers for this 51.4% were then sorted into thematic groups, the frequency of these thematic groups were as follows: patient’s acute illness took priority of hospitalization, 28.6% (12/42), the inpatient setting was deemed not appropriate for discussion, 21.4% (9/42), team identified psychosocial barriers, 19.0% (8/42), not enough clinical information, 14.3% (6/42), transition of care between teams/hospitals, 11.9% (5/42), the patient lacked medical insight, 2.4% (1/42), and the conversation was confirmed to occur prior to admission 2.4% (1/42).
Conclusions: A high prevalence of medical inpatients are perceived as being near the end of life by their attending physician. Despite this, physicians elect not to initiate end of life discussions in over half of these patients. Commonly, physicians prioritize treatment of the acute illness over eliciting the context and patient goals in the final year of life, and perceive that acute hospitalization is often not the best place to initiate these discussions. Improvement efforts will need to account for specific drivers of these behaviors to maximize chances of success.