Higher Medicare spending for advanced cancer not linked to survival

As the concept of value ripples through various healthcare domains, advanced cancer treatment requires a closer look. When researchers from Dana Farber Cancer Institute reviewed spending and survival data, they observed substantial regional variation in Medicare spending for advanced cancer, but no association between spending and survival, according to a study published online March 12 in Journal of the National Cancer Institute.

“Most new advanced cancer therapies provide survival gains of weeks to months, and many are associated with high costs … improving the value of medical spending for advanced cancer is increasingly recognized as a priority,” wrote Gabriel A. Brooks, MD, from Dana Farber in Boston, and colleagues.

The researchers sought to determine the association between mean regional spending by quintile and survival in advanced cancer. They mined the Surveillance, Epidemiology and End Results-Medicare linked database to identify 116,523 patients aged 65 and older with advanced non-small cell lung, colon, breast, prostate or pancreas cancer from 2002 to 2007.

A total of 61,083 patients formed an advanced cancer (incident) cohort, and data were analyzed from diagnosis until six months after diagnosis. A total of 98,935 patients died from cancer and formed the decedent cohort. Data were analyzed for the six-month period prior to death. Thirty-seven percent of subjects were in both cohorts.

Brooks and colleagues identified 80 hospital referral regions (HRRs) and calculated mean spending by cancer site, ordered the regions by spending and divided them into quintiles.

Mean six-month spending was $33,727 in the incident cohort and $33,099 in the decedent cohort. Mean spending increased by 32 percent from quintile 1 to quintile 5 in the incident cohort, and by 41 percent in the decedent cohort.

Despite the spending variation, researchers found no consistent association between advanced cancer spending and survival outcomes.

Brooks et al also observed that patients in high-spending regions were more likely to be hospitalized and admitted to the intensive care unit and consistently received more aggressive end-of-life care than those in low-spending regions. However, spending was similar in all regions in the use of chemotherapy and other outpatient services.

Spending and hospice use were inversely associated in the decedent cohort, which supports previous findings of lower end-of-life costs for patients in hospice care.

“Our findings suggest that healthcare providers should be incentivized to develop strategies aimed at reducing potentially avoidable hospitalizations and increasing timely access to palliative care for patients with advanced cancer—goals that are consistent with patient-centered care,” concluded Brooks et al.