ATLANTA—End-of-life care in cancer patients became increasingly aggressive in the 1990s, according to a study presented here at the ASCO Annual Meeting.
“We found further evidence of a steadily increasing propensity to utilize aggressive interventions very near death over the 1990s,” said Craig C. Earle, MD, MSc, Medical Director of the Lance Armstrong Foundation and Associate Professor of Medicine at Harvard Medical School and the Center for Outcomes and Policy Research at Dana-Farber Cancer Institute.
“There are significant regional variations in these trends, but the reliability we observed over time suggests that these measures are stable properties of the health care system.”
The study (Abstract 6004), presented at an oral session, used linked data from the Surveillance, Epidemiology, and End Results program and the Medicare database to examine trends in cancer care at the end of life in patients over age 65. Dr. Earle and his colleagues examined data on more than 215,000 patients treated between 1991 and 2000 in 77 health care service areas across the country.
During that decade, it became strikingly more likely for patients to still be receiving care within 14 days of dying, with the rate rising from 9.7% in 1991 to 11.6% in 2000. However, a previously detected trend for an increasing number of patients to start a new chemotherapy within 30 days of death did not continue. Rather, the fraction of patients receiving a new chemotherapy regimen in the last month of life stabilized after 1996 at approximately 3%.
Patients were more likely at the end of the decade than at the beginning to visit the emergency department in the last month of life, with the percentage rising from 23.9% to 28.2%. Similarly, the fraction of patients admitted to the ICU in their final 30 days rose from 7.8% to 11% and there was an increase in a trend for multiple hospitalizations from 6.7% to 8.7%.
Finally, there was an increase in the fraction of patients who entered hospice care within three days of death from 12.1% to 14.7% over the course of the decade.
One of the major goals of the study, Dr. Earle said, was to determine the reliability of administrative data for such analyses. As a measure of this, the research team compared the year-to-year patterns of care within a health care service area and found that regions where aggressive care was common one year also tended to have aggressive care the following year.
The consistency of such data indicates that the methods and data sources are reliable, Dr. Earle concluded.
The Discussant for the study, Bruce E. Hillner, MD, Professor and Eminent Scholar in the Department of Internal Medicine at Virginia Commonwealth University, commented, “All of these elements of increasingly aggressive care require physician action. Are any of us surprised? I think the answer is no.”
Among the questions asked by the study's Discussant were: Does managed care have an impact on end-of-life choices for treatment? Is there an adequate supply of hospice beds or is the delay in entering hospice due to inadequate availability? And how do the financial pressures of care affect such decisions?
The more interesting questions perhaps examine how the system affects such care, he continued. Does managed care have an impact on end-of-life choices for treatment? Is there an adequate supply of hospice beds or is the delay in entering hospice due to inadequate availability? And how do the financial pressures of care affect such decisions?
Dr. Earle said that it is important to distinguish between giving aggressive treatment when there is a hope of a cure and giving aggressive treatment to a terminal patient where there is little hope of even extending life or palliating symptoms.
“Clinical medicine in oncology is not as simple as ‘more is better,’” he said. “There is a time to stop.”