or the first time, Medicare officials Wednesday posted quality scores for some 3,800 hospice providers on its new website, Hospice Compare, aimed at helping people select hospice facilities for themselves or others.
In a press briefing Wednesday, Kate Goodrich of the Centers for Medicare & Medicaid Services said the effort will provide a “snapshot on the quality of care delivered by each provider” that will “help consumers make informed decisions.”
Scores for the vast majority of hospices were near the top end of the quality range — so good, in fact, that some observers questioned whether consumers will find the data useful for comparison shopping.
The Hospice Compare website allows individuals to compare up to three hospice facilities at a time along with national averages for each of the seven measures. They also can see the data in table or graph format. For profit or not-for-profit status, Medicare certification date and phone number of each hospice are also shown.
By reporting these data publicly — which was required as part of the Affordable Care Act — CMS hopes to encourage hospice providers to improve their services, Goodrich said.
The scores rate hospice companies — those providing care to patients with life expectancy of six months or less — on whether their services met the following seven measures: pain screening; pain assessment; shortness of breath assessment; shortness of breath treatment; whether patients’ opioid-related constipation was treated with a bowel regimen; whether patients and caregivers were asked about treatment preferences such as hospitalization and resuscitation; and whether patients and caregivers were asked about their beliefs and values at the start of hospice care.
Data were self-reported for care provided Oct. 1, 2015, to Sept. 30, 2016. Reporting is voluntary, although failing to report incurs a 2 percent reduction in Medicare payments. Despite that, hundreds of hospices declined to report. On the pain assessment measure, for example, nearly 4,500 hospices are listed, but data were unavailable for nearly 1,400, although the exact reasons were not given.
Hospice experts have raised two complaints about the system. One is that the measures chosen for reporting may not fully reflect care quality. The other may seem odd: it’s that, while the chosen measures aren’t inappropriate for evaluating hospice care, nearly all hospices perform very well on them, and thus the data don’t provide any differentiation when consumers try to compare different facilities.
Let’s look at the second complaint first.
Too much of a good thing?
The problem, some hospice experts say, is that hospices have had a long time to perfect CMS’s chosen measures and, as a result, all hospices in many communities scored at least 90% on all measures.
Providers have known for several years which measures were being scored, and got a chance to review their data prior to publication early this summer to correct errors.
In March, the Medicare Payment Advisory Commission also noted the lack of variation in hospice scores. For all measures except pain assessment, “at least three-quarters of hospices performed the process appropriately more than 91 percent of the time.” For documentation of treatment preferences and shortness of breath screening, scores averaged 98 percent.
Scores for one measure, assessment of pain, did show wider variation, with 907 hospices scoring below 70 percent, and 1,388 lower than 80 percent. The national average was 77.7 percent — a rate suggesting considerable room for improvement nationally.
But the national average for the other six measures was 93.6 percent or higher.
Goodrich, director of the CMS Clinical Center for Standards and Quality, acknowledged during the press briefing that many of the measures do have “fairly high” performance scores.
“These are measures that hospices have gained a fair amount of familiarity with; they’ve been in the program for a little while, so we knew we had — based on our own work with the data — fairly valid and reliable data, which of course is important for anything that’s going to be publicly reported,” Goodrich said.
Charles Padgett, also of the CMS clinical center, added during the briefing that the website is particularly helpful if one is looking at a hospice that is at the lower end of quality. “There is differentiation among facilities on the lower end,” he said.
More measures to come
In managing pain, it’s not enough just to screen for pain and assess pain, said Dr. Joan Teno, a national hospice quality expert at the University of Washington, who wrote about the issue in the journal Health Affairs. Providers need to choose the right medicine, and then do the right follow-up to make sure the medicine is achieving the right results.
That’s one of the concerns that CMS’s quality rankings are not comprehensive enough.
More measures are needed, acknowledged CMS’s Goodrich, and some are in the works to score hospice care in other important ways. The agency plans to publicly report survey results on family and patient experiences of care.
The survey’s 10 questions ask family members such questions as: How often did the hospice team keep you informed about when they would arrive to care for your family member? Or, how often did the hospice team explain things in a way that was easy to understand?
And, how often did anyone from the hospice team give confusing or contradictory information about a family member’s condition or care?
Other measures were added for hospices to start reporting last April, but are at least a year away from being publicly reported on Hospice Compare.
According to the MedPAC report, one of those scores the percentage of patients who received a visit from a registered nurse, physician, nurse practitioner or physician assistant during the last three days of life.
Another counts the percentage of patients receiving at least two visits from a social worker, chaplain or spiritual counselor, licensed practice nurse or hospice aide in the last seven days of life.
Those are expected to be rolled out in winter of 2018 on Hospice Compare.
Yet another measure under consideration would score whether symptoms such as pain are brought under control within 48 hours and whether or not the patient has a “comfortable death” in the eyes of the family members and how that should be defined.
Goodrich also noted that CMS is working on a composite score that will roll several hospice measures together, which she said will be more meaningful to patients and “will also further differentiate performance across providers.”
“By next winter, we’re going to have a much more complete picture for consumers and payers to use,” said Carol Spence, vice president of research and quality for the National Hospice and Palliative Care Organization.
She noted that measuring care for hospice services has been difficult — “it’s come late to the table when it comes to quality reporting” — in part because one-third of patients die within seven days of admission to hospice care.
Some providers welcomed the release. Mary Zalaznik, senior vice president of VITAS, a large national for-profit hospice provider, said the data gave them insight on how to create quality improvement plans. “As the site matures, Hospice Compare will include measurements regarding the patient’s reported experience along with quality, outcome-based measurements,” she said.
Dr. Karl Steinberg, who sits on the National Quality Forum standing committee which debates and approves end of life measures for use in healthcare settings, said he hopes that all of the hospices will benefit from Hospice Compare. “They can see how they compare to local and national parameters, and take stock to see what quality improvements they can put in place.”
That includes his own. As medical director of Hospice By the Sea in Solana Beach, Calif., he acknowledged room to improve in pain assessment, treatment for patients who are short of breath, and asking patients about their beliefs and values, measures on which his hospice scored between 80 percent and 84.8 percent.
But he too expressed concern about how high so many hospices scored this time around. “The bar is so high, for most other measures, 85 percent would be considered really good,” he said.
“I think you have to take it all with a grain of salt, especially when there’s not a lot of scatter,” Steinberg added. “If one hospice is 94 percent and one is 92 percent, you don’t really know that one hospice is somehow a lot better.”