According to a recent Dartmouth Atlas study:
“The nation’s most elite cancer care centers performed only modestly better than community hospitals at meeting recognized quality standards for treating dying cancer patients, displaying similar patterns of relatively aggressive, high-intensity treatments in the final weeks of these patients’ lives…
The Dartmouth researchers also found that even among hospitals with a specific clinical focus on cancer care, such as those in the National Comprehensive Cancer Network and at designated National Cancer Institute centers, there were significant variations in how they treated patients at the end of life. The analysis found two-fold differences among these institutions in the rates of intensive care unit use in the last month of life, chemotherapy in the last 14 days of life, deaths occurring in the hospital, and the use of hospice care for fewer than three days.”
This begs the question of whether most variation is due to random noise due to unobservable factors not accounted for by the paper’s risk adjustment model and how much represents true, hospital-specific findings.