Palliative care and hospice (a type of palliative care) are associated with less aggressive treatment at the end of life but under Medicare, beneficiaries have to forego expensive treatment like radiation and chemotherapy. The Veteran's Health Administration (VHA) does not require this of Veterans, whether receiving VA provided or Medicare reimbursed hospice, allowing us to ask whether the large increase in the availability of hospice in the VHA over the past decade without restrictions on the use of medical treatment has altered Veterans' end of life experiences, particularly the "burdensome health care transitions" that often accompany patients receiving aggressive medical care at the end of life.
The objectives of this study are to show the relationship between hospice and less aggressive treatment at the end of life for veterans newly diagnosed with non-small cell lung cancer (NSCLC). Specifically, we investigate if allowing veterans to access cancer treatment concurrently with hospice has altered veterans' end of life experiences, particularly burdensome events (i.e., burdensome health transitions and treatments) that often accompany patients receiving aggressive medical care at the end of life and whether this access results in higher overall medical care costs.
Using VA inpatient, outpatient and pharmacy claims matched with similar Medicare data, we created VA Medical Center level aggregates per year (2006-2012) characterizing all cancer decedents' use of hospice, cancer treatment and/or the simultaneous receipt of these two types of service. Using a facility fixed effect model and controlling for patient demographics, co-morbidities and pre-diagnosis hospital use, we tested the effect of the facility level estimate of the changing proportion of cancer decedents receiving hospice in the last six months of life in a given facility year on the likelihood that non-small cell lung cancer (NSCLC) patients experienced burdensome transitions (3+ hospital admissions, Tube feeding, mechanical ventilation, ICU admission) in the last month of life and whether their overall medical care costs increased in relation to exposure to more concurrent care.
There was a 50% increase in the VHA facility average percentage of all cancer decedents to receive hospice care during the study period but no change in the proportion of all cancer patients who received radiation or chemotherapy in the last 6 months of life. Over the study period 11.9% of NSCLC decedents experienced a burdensome transition (9.0% ICU admission, 5.1% mechanical ventilation, 7.4% feeding tube and .5% 3+ hospital admissions) in the last month of life. Controlling for patient demographics and co-morbidities, for every one percent increase in facility-level hospice use, the relative risk of a NSCLC patient experiencing a burdensome transition in the last month of life declined by almost 80% (AOR .22; 95% CI .07-.67), despite the fact that over the study period there was a 3 percentage point absolute increase in burdensome transitions in the newly diagnosed NSCLC patient population. We observed no association between 6 month mortality and increasing hospice exposure nor was there a statistically significant increase in VA plus Medicare health care costs.
Consistent with several small, single institution randomize trials, these findings suggest that allowing patients to simultaneously receive palliative care and anti-tumor treatment, concurrent care, actually reduces the aggressive end-of- life care associated with poor quality of life at no significant increase in medical care expenditures.
External Links for this Project
Grant Number: I01HX000956-01A1
None at this time.