Are Hospice Services Underutilized, Abused, or Both?

Growth in expenditures for hospice services has accelerated in the last ten years, raising concerns about overuse. According to CMS, Medicare spending for hospice services has grown  by an average of  $1 billion per year, from $2.2 billion in calendar year 1998 to $12.1 billion in 2009.  Medicare covered twice as many hospice patients in 2009 than in 1998.

The number of providers of hospice services also has skyrocketed. When Medicare first covered hospice services in the early 1980s, hospice providers were nonprofit organizations staffed largely by volunteers. In 2010, there were more than 3,400 providers of hospice services, serving 1.56 million patients. More than half of the providers are for-profit entities.

 The average length of service (LOS) in hospice care has grown steadily.According to CMS, the average number of days a Medicare beneficiary received hospice services in 1998 was 48 days. By 2009 the average was 71. Cancer patients were originally the most common users of hospice services, but in the last several years patients with other diagnoses, such as dementia, stroke, congestive heart failure or kidney disease, also have made use of the benefit.

Coverage of hospice care also has been extended to frail individuals with multiple chronic conditions, no one of which is necessarily terminal, using a diagnosis of failure-to-thrive or debility—not otherwise specified (NOS). In these cases, the terminal condition required for Medicare coverage is the combination of illnesses. 

CMS tracks the top 20 primary diagnoses of Medicare beneficiaries receiving hospice services and the average length of service (LOS) associated with each. In 1999, the most common diagnosis was lung cancer (15 percent). Debility-NOS was number 13 (3 percent).  Failure-to-thrive was not among the top 20.  The average LOS were 48 and 50 days, respectively.

In 2009, debility NOS was the most common primary diagnosis  (11 percent), and the average LOS  was 83 days. Lung cancer was the 3rd most common, with an LOS of 45 days. The average LOS for other cancer diagnoses changed only by a few days. In contrast, the average LOS for Alzheimer’s, non-Alzheimers dementia and congestive heart failure grew by 50 percent or more.  Failure to thrive was the 6th most common diagnosis, with an average LOS of 84 days.

There are not yet standard measures for predicting how long a patient is likely to live, and it may be harder to do so when the patient is suffering from dementia or stroke than with cancer.  Perhaps determining life expectancy is even more difficult with more nebulous diagnoses such as debility.

 The growth in utilization of hospice services has coincided with the growth of for-profit hospice providers. Some researchers have found that for-profit hospices admit patients earlier and bill for longer stays; for-profit hospices have more patients  whose length of stay exceeds a year. News stories report that several False Claims Act and  whistleblower lawsuits have been brought against for-profit hospice chains. One story describes patients who continued to receive hospice services for four years or more. 

There is evidence that palliative care, including hospice care, actually may prolong life. In one study, cancer patients who received palliative care soon after diagnosis lived longer and did not seek high-cost hospital intervention in the last few days of life. 

Sometimes patients rally and beat the odds, as Art Buchwald did. When he was admitted to hospice, his physicians believed he had only a few weeks to live. Five months later, he was discharged, and he lived several  more months.  

So, how much of the increased utilization is overuse? To what extent are beneficiaries simply taking advantage of a benefit to which they are entitled? Medicaid home- and community-based services programs have experienced higher enrollment than anticipated, which some attribute to a “woodwork effect” — people who qualified for nursing facility care but asked for services only when the HCBS benefit became available.

A report by the National Hospice and Palliative Care Organization (NHPCO) in December 2011 distinguishes between the mean LOS, the number of patient days divided by the number of discharges, and the median LOS, the LOS at the 50th percentile. NHPCO believes that the median is a better measure of utilization because it is less subject to distortion by outliers. The report states that the median LOS for hospice patients in 2010 was 19.7 days,  decreased from 21.1 in 2009. According to NHPCO, the mean LOS for all hospice patients declined from 69.0 in 2009 to 67.4 in 2010. 

The writers called attention to the point that,a median of 20 days means that half of the Medicare beneficiaries who used hospice services received services for less than three weeks. In 2010, about 35 percent died or were discharged within  seven days of admission, and another 14 percent died or were discharged within 14 days.

It’s also worth noting that the average (mean) LOS is well under the 180-day initial benefit. Perhaps the abuses reported actually are outliers.