Graham Center Research

More Primary Care Involvement Improves End-of-Life Care

January 18, 2017 03:16 pm Michael Laff

A recently published study indicates that greater primary care involvement at the end of life contributes to improved outcomes and lower Medicare costs for patients in their final years.

Researchers at the Robert Graham Center for Policy Studies in Family Medicine and Primary Care investigated how the degree to which primary care physician are involved with patients at the end of life influences the quality of their care. In the study,(www.annfammed.org) published in the January/February issue of Annals of Family Medicine, they conclude that patients in regions with more primary care involvement experience less intensive end-of-life care.

To determine the level of primary care involvement, researchers calculated the ratio of primary care physician visits to subspecialist visits. In regions with a high ratio of primary care visits, patients at the end of life recorded fewer ICU visits, less fragmentation of care and lower overall Medicare spending.

The average stay in the ICU during the last six months of life was 2.9 days for patients in regions with the most primary care involvement compared to 4.3 days for patients in regions with the least primary care involvement. In particular, patients with lung cancer whose primary care physicians were involved during their hospitalization were 25 percent less likely to require ICU care.

Story Highlights
  • In a study published in the January/February issue of Annals of Family Medicine, researchers at the Robert Graham Center for Policy Studies in Family Medicine and Primary Care investigated how the degree to which primary care physicians are involved with patients at the end of life influences the quality of their care.
  • In regions with more primary care physician involvement, such patients recorded fewer ICU visits, less fragmentation of care and lower spending.
  • Medicare spending during the final two years of life for the patients studied was $65,160 for those with the most primary care involvement and $69,030 for those with the least involvement.

Because greater primary care involvement also means improved care coordination, patients at the end of life who lived in regions with the most primary care involvement were less likely to see 10 or more physicians.

"Primary care really helps with care coordination but can be overwhelmed if too many specialists are involved," said Claire Ankuda, M.D., M.P.H., a former Graham Center fellow and a co-author of the study.

The difference in care is reflected in costs. Medicare spending during the final two years of life for the patients studied was $65,160 for those with the most primary care involvement and $69,030 for those with the least involvement. Ankuda noted the study's cost estimates are conservative, so the variance researchers identified is significant.

Although previous research has found that patients with more primary care physician visits preceding the last six to 12 months of life saw fewer hospital days at the end of life, as well as lower rates of in-hospital death, the Graham Center researchers found no difference in the number of hospital deaths or days spent in hospital during the last six months of life for those with more versus less primary care involvement.

Ankuda explained that research on end-of-life care is largely focused on specific illnesses, such as cancer, or settings of care, such as the ICU.

"Mortality rates are high in those settings, but most people are not dying from those diseases. They have multimorbidities," she said. "If we are really thinking about population health and end-of-life care, we need to think about primary care."

Regions with the highest rates of primary care involvement in end-of-life care did report lower rates of hospice enrollment -- a finding the authors sought to explain. Hospice growth in select areas could be a result of fewer available primary care physicians, they wrote, but they considered other possible causes.

"Primary care physicians may hesitate to refer to hospice if doing so is perceived as terminating a longstanding relationship," authors wrote. "Alternatively, primary care physicians may not recognize patients as having a prognosis of less than six months because of their longstanding relationship."

Data were drawn from 306 hospital referral regions (HRRs) where a total of 1.1 million Medicare Part B beneficiaries with chronic disease died. HRRs are based on the nearest hospital where a patient would go for highly specialized care such as brain surgery.

The researchers emphasized that their findings are not a comprehensive look that can be interpreted as a predictor of similar results nationally. Regions with greater primary care involvement had fewer urban residents and fewer African Americans, for example. Ankuda said comparing the patient experience in Boston, where a range of subspecialists are available, with that in rural Minnesota would not yield much insight.

"This study is important because there is so little data about primary care and end of life," she said. "By looking at regions, it provides a bird's eye view of the landscape, but more work needs to be done to understand what is happening in the doctor's office and inside people's homes."

Related AAFP News Coverage
AAFP, Conversation Project Partner to Aid End-of-Life Care Conversations
(12/5/2016)

Health is Primary
Campaign Promotes Family Physicians' Role in End-of-Life Care

(2/26/2016)

More From AAFP
Policy on Ethics and Advance Planning for End-of-Life Care

Recommended Curriculum Guidelines for Family Medicine Residents: Palliative and End-of-Life Care(10 page PDF)

American Family Physician: AFP by Topic: End-of-Life Care

Family Practice Management: Discussing End-of-Life Care With Your Patients
(March 2008)

FamilyDoctor.org: End-of-Life Issues(familydoctor.org)