Physicians' Beliefs Contribute to Geographic Variation in End-of-Life Expenditures for Cancer Patients

Healthcare expenditures in the last months of life are known to vary considerably across US geographic areas, with no link to improved outcomes. This variation is driven not by patient values, but by the types of healthcare services available in different regions and by the end-of-life care beliefs and knowledge of physicians, according to a report published in Health Affairs.

"We found that physicians' beliefs and practice styles and area-level availability of services were the primary drivers of variations in intensity of care," write the authors. "Patients' beliefs, preferences and supports did not contribute meaningfully to geographic variation in spending intensity."

Investigators analyzed patient survey data from the prospective, multiregional Cancer Care Outcomes Research and Surveillance Consortium (CanCORS) study; that data was then linked to Medicare data on 1,132 Medicare patients diagnosed from 2003 to 2005 with advanced lung or colorectal cancer who had died by 2013 (female, 43%; white race, 79%; mean age at death, 75.6 years).

Also included in the analysis were results from a 2005-2007 CanCORS survey conducted among physicians identified by patients as having key roles in their care. Hospital referral regions (HRRs) across the US (n = 26) were stratified into quintiles according to the amount of Medicare spending in the last 30 days of life.

Mean healthcare expenditures in the last month of life were $13,663, but ranged widely, from $10,131 in the lowest-spending quintile to $19,318 in the highest-spending region.

Higher-Spending Regions Had:

• Greater concentrations of physicians per capita than lower-spending regions
• A lower proportion of primary care physicians
• Fewer hospices per 10,000 people

Physicians in Higher-Spending Areas Reported Feeling:

• Less prepared to treat end-of-life symptoms than physicians in lower spending areas (33.4% vs 40.7%; P < 0.001)
• Less knowledgeable about discussing end-of-life options (49.5% vs 57.8%; P < 0.001)
• Less likely to discuss DNR status with a patient they estimated had four to six months to live (19.3% vs 30.3%; P < 0.001)
• Less likely to seek hospice care for themselves if terminally ill (54.4% vs 71.4%; P < 0.001)
• More likely to recommend chemotherapy for a Stage-IV lung cancer patient with poor performance status and pain (47.8% vs 29.4%; P < 0.001)

In multivariate analysis, patient demographic and clinical variables had a negligible effect on the variation in end-of-life spending. However, availability of healthcare services explained 39% and physician beliefs explained 26% of expenditure variation in the final month of life.

"What we really need are interventions that help physicians feel more comfortable taking care of patients at the end of life, along with better training about the lack of efficacy and potential harms of some intensive treatments for patients with advanced cancer," says lead author Nancy L. Keating, MD, professor of healthcare policy and medicine at Harvard Medical School and a physician at Brigham and Women's Hospital, Boston.

Source: "Factors Contributing to Geographic Variation in End-of-Life Expenditures for Cancer Patients," Health Affairs; July 2018; 37(7):1136-1143. Keating NL, Huskamp HA, Kouri E, Schrag D, Hornbrook MC, Haggstrom DA, Landrum MB; Department of Health Care Policy, Harvard Medical School; Division of General Internal Medicine, Brigham and Women's Hospital; and Dana-Farber Cancer Institute, all in Boston; Center for Health Research, Kaiser Permanente Northwest, Portland, Oregon; Indiana University School of Medicine and Indianapolis Veterans Affairs Medical Center, both in Indianapolis.