Skip to Content, Navigation, or Footer.
Support independent student journalism. Support independent student journalism. Support independent student journalism.
The Dartmouth
February 26, 2024 | Latest Issue
The Dartmouth

Atlas reveals gap in end-of-life treatment

The type of medical care received by cancer patients near the end of their lives varies based on geographic location and the focus of local health care systems, according to research published by the Dartmouth Atlas project on November 16. In many cases, care does not reflect patient needs or desires, the study found.

The study "Quality of End-of-Life Cancer Care for Medicare Beneficiaries: Regional and Hospital-Specific Analyses" was authored by Dartmouth Medical School professor David Goodman.

"We know that the care [patients] receive has less to do with what patients want and is more about the hospitals they happen to get care at," Goodman said in an interview with The Dartmouth. "Generally, the care often doesn't represent their preferences."

The study examined almost 240,000 Medicare records of patients with aggressive cancers who died between 2003 and 2007, according to a DMS press release.

An average of 29 percent of cancer patients died in hospitals across the United States as opposed to elsewhere, including hospice care with the number varying widely depending on geographic location, the study said. The largest proportion of hospital deaths 46.7 percent occurred in Manhattan, while only 7 percent of patients died in hospitals in Mason City, Iowa, according to the study.

The type of treatments received by cancer patients near the end of their lives also varied widely, Goodman said.

Manhattan patients received the highest rates of aggressive cancer treatments, while hospitals in cities like Des Moines and Minneapolis offered these treatments at far lower rates, the study said.

Although research has shown treatments like chemotherapy to have little effect on patients with aggressive cancer, many such patients continually receive chemotherapy during their last two weeks of life, the study found. Over 10 percent of patients with aggressive cancer received chemotherapy treatments in Olympia, Wash., for example, according to the release.

Use of "palliative" methods like hospice care, which are designed to treat symptoms without attacking the underlying cause of a disease, occurred in more than half of cases dealing with aggressive cancer. The percentage varied according to type and location of treatment facilities, with at least 50 academic medical centers sending less than half of their patients into hospice care, according to the release.

In some treatment centers, the hospice referral came too late in the patient's life to be effective, the study found.

Although it has long been an established problem in the health care industry that end-of-life care does not match patients' needs, data from the Dartmouth Atlas study can show where the discrepancy is most significant, according to Goodman.

Goodman attributed the geographic variation to the tendency of different health care systems to focus on different aspects of patient care. Nuances in the dialogue between physicians and their patients about care options could also be a reason for the difference, he said.

"[End-of-life care matters because] we're all going to die, and how we're cared for is terrifically important to each of us," Goodman said. "This has long been a concern of physicians and patient advocacy groups and health systems."

Goodman said he hopes the study will compel hospitals and health systems to reexamine their practices and look for ways to improve care. Patients and families should also feel inclined to ask more questions when discussing end-of-life care with their physicians, he said.

"Progress has been made," Goodman said, "but there is still work to be done."

DMS professors Elliot Fisher and Nancy Morden co-authored the study, along with Chiang-Hua Chang, a research instructor at the Dartmouth Institute, and medical researchers from several institutions across the Northeast.