A recent study in Health Affairs assessed cost and utilization outcomes for 3 innovative care models for Medicare beneficiaries with cancer: oncology medical homes, patient navigator programs, and palliative care initiatives.
A recent study in Health Affairs assessed cost and utilization outcomes for 3 innovative care models for Medicare beneficiaries with cancer: oncology medical homes, patient navigator programs, and palliative care initiatives. The medical home and patient navigator models were associated with significantly lower costs and fewer hospitalizations at the end of life (EOL).
As Medicare attempts to lower its spending, a logical target for savings is the costly last year of life for beneficiaries with cancer. In 2010, the costs for cancer care in the last year of life totaled $37 million, mainly due to frequent hospitalizations, emergency department (ED) visits, and intensive care unit stays. Not only do these outcomes drive up spending, they also represent a diminished quality of life for patients with terminal cancer, who in previous studies have reported that they would prefer palliative care and spiritual support instead of aggressive treatment and inpatient hospitalization. While hospice care is an option, CMS has encouraged the development of innovative care models that provide cancer care outside of hospice.
In the Health Affairs study, researchers analyzed the cost and utilization outcomes for beneficiaries who died in 3 different oncology care models. Each of these models received a Health Care Innovation Award from CMS.
The first, the Community Oncology Medical Home (COME HOME) model, opened 7 oncology medical homes across the US based on 3 key principles: triage pathways, enhanced access to care, and evidence-based diagnosis and treatment pathways. Though not specifically a palliative care model, it encouraged patients to manage their illness at home or in outpatient settings when possible as opposed to hospitalization.
The second model, the Patient Care Connect Program (PCCP) from the University of Alabama at Birmingham, introduced patient navigation at 12 sites in 5 southern states by providing nonclinical navigators to empower patients and their families, guide them to appropriate care resources, and communicate their concerns to the care providers. A main component of this model was the Respecting Choices program, which facilitated advance care planning and EOL goal setting between patients and family members.
Last, the CARE Track model from the University of Virginia administered patient-reported outcomes surveys to identify patients in need of more intensive palliative care, then connected them to such services. The researchers then compared outcomes for patients who died in each of these 3 models to similar patients at matched practices.
Both the PCCP and the COME HOME models demonstrated significantly lower costs per patient in the last 30, 90, and 180 days of life as compared with the matched comparison patients. The CARE Track model also had lower costs per participant relative to the comparison group, but the differences did not reach statistical significance, potentially due to the small sample size.
In the last 30 days of life, PCCP participants had significantly fewer hospitalizations and ED visits than the comparison patients, while the COME HOME group had reduced hospitalizations compared with the matched group. The CARE Track participants had slightly more hospitalizations and emergency visits than the comparison patients, but these differences were not statistically significant.
Finally, the researchers assessed the use of hospice care and chemotherapy to signify quality of life. All 3 models saw increased hospice use compared with their comparison groups, although the difference only reached significance for the PCCP patients. The CARE Track participants had a low rate of chemotherapy use, which the study authors wrote “was consistent with the program’s emphasis on EOL quality over intensive treatments with limited clinical value.”
The study authors were encouraged by the positive results from the models, even though they did not always reach statistical significance. “A comprehensive approach to cancer care could include features from all three interventions described in this study to improve end-of-life outcomes, appropriately adapted to the target population and setting,” they concluded.