Geography Is Big Variable in Cancer Care

The cost of care during the last month of a cancer patient's life varied by as much as 90% depending on where the patient lived, an analysis of more than 1,000 cases showed.

Average expenditures during the last month of life ranged from about $10,000 in some areas to more than $19,000 in others. Investigators found that the wide variation in end-of-life expenditures was not associated with patient or family preferences but instead with treating physicians' attitudes, beliefs, and knowledge about end-of-life care, as reported in Health Affairs.

"Numerous studies have shown that greater spending and more care at the end of life do not contribute to better outcomes," said study author Nancy L. Keating, MD, of Harvard Medical School and Brigham and Women's Hospital in Boston, said in a statement. "Given that more care and greater spending also do not stem from patient preferences, much of these additional services can be considered wasteful or even harmful."

An unrelated study documented similar geography-related wide variation in the use of imaging studies during follow-up of patients with low-risk breast cancer. In almost a third of the 36,000 cases reviewed, physicians requested advanced imaging studies not recommended for monitoring low-risk breast cancer, such as PET, as reported in the Journal of the National Comprehensive Cancer Network.

End-of-Life Care

The study by Keating's group added to a large volume of evidence that healthcare spending at end of life varies across geographic areas and is not associated with improved outcomes. The authors sought to extend the evidence by investigating potential explanations for the variation, including patient-related socioeconomic and demographic factors, clinical factors, availability of services, and patient and physician beliefs about end-of-life care.

Data for the analysis came from the Cancer Care Outcomes Research and Surveillance Consortium (CanCORS) and involved 1,132 patients, ages ≥65, with diagnoses of advanced-stage lung or colorectal cancer during 2003-2005. Follow-up continued to 2013. The data comprised medical facilities in 26 hospital referral regions (HRRs), which were stratified into quintiles on the basis of end-of-life spending.

The data included responses to a CanCORS survey conducted during 2005-2007 among physicians who patients identified as having key roles in their care.

Overall, healthcare expenditures during a patient's last month of life averaged $13,663. Mean values ranged from $10,131 in the lowest-cost HRRs to $19,318 for HRRs in the highest of the five spending categories. In general, HRRs with higher levels of end-of-life spending had larger populations, more nonwhite residents, more physicians, fewer primary care doctors, and fewer hospital beds and hospice facilities.

The analysis showed no consistent associations between patient-related factors and HRR-specific spending, including beliefs about cure, the toxicity associated with treatment, or general knowledge about cancer. The data showed a slight trend for an association between higher HRR-level spending and patient preference not to receive treatment that would extend their lives even if it caused more pain.

Analysis of the treating physicians' views revealed several factors associated with HRR-level spending at end of life. Physicians in higher-spending areas reported feeling:

• Less prepared to treat end-of-life symptoms
• Less knowledgeable about discussing end-of-life options
• Less comfortable discussing DNR status
• Less comfortable discussing hospice care and less likely to enroll in hospice themselves if they had terminal cancer
• More likely to recommend chemotherapy when patients were unlikely to benefit from it

Statistical modeling showed that physician beliefs explained 26% of the HRR-level variation in expenditures during the last 30 days of life, whereas patient beliefs did not influence the geographic variation. The availability of services, combined with patient demographics and clinical variables, explained 39% of the variation.

The findings emphasized the need for education and training "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," said Keating.

Breast Ca Imaging

The study of imaging during follow-up of patients with low-risk breast cancer compared clinical practices within the context of recommended imaging. The American College of Radiology, American Society of Clinical Oncology, and the NCCN have published recommendations about surveillance after treatment for breast cancer, and the guidance limits or discourages use of advanced imaging techniques in most cases.

Additionally, several large randomized trials and systematic reviews consistently showed that imaging studies beyond mammography, particularly nonbreast imaging, has no effect on survival or quality of life for patients with early breast cancer, according to Benjamin Franc, MD, of the University of California San Francisco, and colleagues. Inappropriate use of imaging can add thousands of dollars to the cost of care, which can be especially burdensome to patients with high-deductible insurance coverage.

The study included data for 36,045 women, ages 18-64, who had surgical treatment of one breast from 2010 through 2012. The data came from a large commercial health claims database. The principal objective was to examine geographic variation in rates of recommended postoperative breast imaging and high-cost whole-body imaging (including CT, MRI, PET, and bone scans). Geography was defined by standard metropolitan statistical areas (MSA).

The data showed that 70.8% of the women had at least one recommended dedicated breast imaging study. However, 31.7% had at least one advanced, whole-body imaging study, including 12.5% (one in eight) who had a PET scan. Across MSAs, the proportion of women who had at least one advanced-imaging study ranged from 18% to 46%.

Patient age and type of therapy (surgery alone versus surgery plus radiotherapy) topped the factors influencing the use of recommended imaging, but MSA came in third (OR 1.42, 95% CI 1.35 to 1.51). Use of PET/CT imaging was influenced most by the type of surgery a patient, followed by MSA (OR 1.82, 95% CI 1.70 to 1.97).

"Age and therapy make sense as predictors of breast imaging, but it doesn't make sense that where you live makes a difference in whether you wee likely to get a follow-up mammogram or high-cost imaging," Franc said in a statement. "What's actionable here is that we have these guidelines, but doctors aren't following them."

The nature of the data used for the analysis did not provide insight into the reasons for the geographic variation. In an email response to MedPage Today, Franc cited several possibilities.

"In part, doctors in local areas may develop cultures of medical care that emphasize certain habits, essentially saying, 'That's how we do it around here.' These habits can persist even as guidelines change."

Another potential factor might be local awareness or access to information related to imaging tests, such as the national impact on prophylactic mastectomy and genetic testing that followed actress Angelina Jolie's personal experience with breast cancer.

"This phenomenon can occur at much more local levels, where the 'celebrity' could be a local radio personality or even someone vocal in a support group," said Franc. "These factors can influence patients to ask their physicians for tests that are recommended by guidelines, which is a positive outcome. Likewise, patients may insist on tests that are not necessarily indicated and have associated drawbacks, such as unnecessary exposure to medical radiation and financial toxicity."