Medicare Spending on Hospice Spending Up; Per-Patient Costs Flat

Medicare spending on hospice care rose by 52% between 2007 and 2015, increasing from $10.4 billion to $15.8 billion, and driven mainly by increases in the number of patients receiving hospice services during the period because per-patient costs remained flat, according to a new study in Health Affairs.

Study authors John Hargraves and Niall Brennan, both of CMS, wrote that they found that the recent growth in hospice spending showed significant geographic variation, with high-cost regions having average per-patient spending more than 3 times that of the spending in low-cost regions. They also found that recent growth in hospice spending varied substantially by patient diagnosis, possibly reflecting a change in diagnosis reporting rules.

“Recent research has shown that providers’ practice patterns, patients’ preferences, and patient’s characteristics may all play a role in hospice use and spending,” they noted.

Medicare’s hospice benefit provides palliative care for terminally ill people who have an estimated life expectancy of 6 months or less. Hospice providers are paid a daily rate ($159 in 2015) for all routine home care services related to a patient’s terminal illness.

The authors analyzed all Medicare hospice claims, including those enrolled in Medicare Advantage, for 2007 to 2015. The analysis included beneficiary, claim, and claim line data from the CMS Chronic Conditions Data Warehouse for the study period as of April 2016. To categorize diagnoses, the authors used the single-level Clinical Classifications Software (CCS) categories for diagnosis codes from the International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10), and hospice patients’ most recent primary diagnosis. They grouped the 285 CCS categories into 7 broad categories they developed based on their frequency of use among hospice patients (cancer, circulatory or heart disease, dementia, respiratory disease, stroke, debility or failure to thrive, and other). Hospital referral regions were used to analyze geographic variation.

Mean per-patient spending for hospice care ranged from $4683 (North Dakota) to $18,106 (California). They found that, in general, spending per patient was higher in California, Texas, the Southwest, and the South, and lower in central New York and North Dakota and South Dakota.

Regions with higher average spending per patient often have more hospice patients with dementia diagnoses compared with areas with lower per-patient spending. Conversely, regions with lower spending tend to have more hospice patients with cancer diagnoses than areas with high per-patient spending. Patients with cancer have the fewest average days of hospice care per patient, and those with dementia have the most average days.

Hospice spending growth from 2007 to 2015 was driven by an increase in patients with non-cancer diagnoses, the authors found, and the diagnosis of dementia accounted for 25% of spending for hospice care ($4 billion in 2015). The increase in non-cancer diagnoses may be a result of growing recognition of the benefits of hospice care for terminal illnesses besides cancer, the researchers conclude. Stroke spending increased markedly, they said, after 2012, possibly a result of changes in diagnostic reporting rules in 2013 and CMS changes in what hospice claims would be acceptable.

“Understanding the drivers of hospice spending and use is important to ensure that hospice care is available and affordable,” the authors concluded. Further research in this area is needed.