Improving Performance in Late-Life Care - A Modest Effort Starts in Philadelphia

I’m working with a small group of Philadelphia-area  hospital executives who have joined together in a campaign to improve late-life care in the region. While recognizing that many factors influence the overall performance of communities in how patients are treated in the final stage of life, these executives appreciate the enormous sphere of influence their institutions exert on their respective communities.

Until shown the DAI Palliative Performance Reports for their respective hospitals, these executives, all of whose hospitals reportedly have a palliative care service, felt  their institutions had been effectively caring for the late-life needs of their patients. Yet the hospitals’ performance lagged behind state and national benchmarks, and well behind the performance of  hospital exemplars.  The Philadelphia region’s performance similarly lagged.

Examples abound.  According to the DAI Palliative Performance Profile (P3) for Philadelphia, a Philadelphia resident will spend 20% more days in a hospital during last six months of life than the national average, and twice as many days as would a resident of the exemplar region of Portland, Oregon. The Philadelphia resident is one-third more likely to die in a hospital than his counterpart in Portland, and twice as likely to have had an ICU stay associated with that terminal hospitalization.  On a positive note, at least one indicator shows Philadelphia to be performing better than national benchmarks – while 41.9% of those who die in the US use the hospice benefit, 44.6% of Philadelphia residents received services from a licensed hospices prior to their death. Yet considerable opportunity  remains for improvement.
How much? Consider that in 10% of communities  across the nation (the exemplar benchmark)  more than 55% of their decedents utilized the hospice benefit.

While few of the executives were surprised by Philadelphia’s poor performance in late-life care, most were surprised by the enormous gaps between the region’s performance and that of exemplar communities. Yet, should we be so surprised? After all, the Dartmouth Medical Atlas (from which the DAI P3 draws its data) has for years documented such variations. In a recent study titled Trends and Variation in End-of-Life Care for Medicare Beneficiaries with Severe Chronic Illness, the authors concluded that  geography continues to play a huge role in late-life care, noting that “care patients received in the months before they died depended largely on where they lived, and widespread variations persist.”

In future posts, we’ll take a closer look at what the DAI Palliative Performance Profiles tell us, and perhaps more importantly, how they might provide sharper focus for performance improvement campaigns, not unlike the modest effort  briefly described above. In the meantime, I invite your comments on the state of late-life care in your region, or the US, in general, and welcome your ideas on the sort of performance improvement campaigns which could make a difference. We’ll also take a closer look at exemplar communities (read here for an earlier blog post on this subject) and how do they do that?

A Closer Look at an Exemplar Palliative Care Community - How Did They Do That?

Previous posts have taken a look at exemplar palliative care communities (read here , here , and here ) and the factors contributing to their exemplary performance. In this post, let's examine more closely another community (LaCrosse, Wisconsin) whose late-life care practices earned it exemplar status in the DAI Palliative Community Performance Profiles. The aim of our (DAI Palliative Care Group) studies of such communities (and their health care institutions) is to identify communities that have achieved results suggesting high performance in late-life care, have undertaken innovations designed to reach higher performance, or exemplify attributes that can foster high performance. These studies are intended to enable other hospice and palliative medicine (HPM) leadership to draw lessons from the experience of exemplar performers that will be helpful in their own efforts to become high performers. These communities, and health systems within the communities, we believe, are well-positioned to develop Accountable Palliative Care Organizations (APCOs).

In LaCrosse County, Gundersen Lutheran Health System has a market share of 59 percent of the inpatient cases. Gundersen is a physician-led, not-for-profit integrated delivery system serving an area with more than 550,000 people in a tristate region that includes parts of western Wisconsin, northeastern Iowa, and southeastern Minnesota. The population it serves, which is both urban and rural, is healthier, less transient, and more educated—but older and poorer—than the national median. The hub of the system is a 325-bed teaching hospital, which serves as the western clinical campus for the University of Wisconsin Medical School and the University of Wisconsin–Madison School of Nursing. While the hospital has an open medical staff, fewer than 5 percent of credentialed physicians are non–Gundersen Lutheran employees.

One of the keys to Gundersen Lutheran’s strategy for improving quality of care and lowering its cost is the optimal use of medical resources for patients with complex conditions and minimal social support. Through a care coordination program, the health system identifies patients who are frequently hospitalized—or who make frequent visits to the emergency care or urgent-care clinics, lack strong support at home, or simply have difficulty coping with the complexity of their health care needs—and assigns them to one of 28 registered nurses and social workers who are trained to help them navigate the health care system. The care coordination program has demonstrated significant cost savings. Charges per patient after 12 months in the program have fallen on average by $7,300 (generating net savings of $5,100 after accounting for program costs of $2,200 per patient), as patients are hospitalized less and make less frequent use of the emergency department for care. The hospital uses the program for its health plan members as well as for the fee-for-service population, though doing so reduces its hospital revenue. “This is living up to [the] mission of improving the health of the community,” says Jeffrey E. Thompson, M.D., Gundersen Lutheran’s CEO.

Lutheran has also increased coordination of care at the end of life—a time at which medical expenses rise—by implementing a comprehensive system for understanding, documenting, and honoring patient values and goals for care at the end of life in all health care settings. The documentation begins with the creation of advance directives that spell out what actions should be taken in the event that a patient is incapacitated or is no longer able to make decisions. Advance directives are embedded in the system’s electronic medical records and are made available to all providers in all care settings. Discussions are held and reviewed periodically during many types of patient encounters to make sure that plans remain current.

A strong partnership with other local providers and community groups promotes advance care planning among community members before they become terminally ill.
A training program developed in partnership with a competing local health system, Franciscan Skemp Healthcare (a division of Mayo Health System), and other community groups helps promote a consistent approach to advance care planning among social workers, chaplains, and other volunteers who carry out community education.

Partnership with other hospitals and community groups is essential to ensuring that conversations with patients about treatment preferences at the end of life—and the documentation of them—are consistent across settings and sites of care. Without such assurances, providers are tempted to dismiss documentation of treatment preferences from competitors because they are uncertain of the methods used to collect the information.

A recent study involving 400 deaths of residents of La Crosse County at all health care institutions over seven months in 2007 and 2008 found that 96 percent had either a written advance directive or a Physician Order for Life-Sustaining Treatment (POLST), a standardized medical order that reflects patient choices about key medical treatments often used at the end of life. An internal study among these patients found that those with advance directives used $2,000 less in physician and hospital services in the last six months of life.

In 2005, the health system began offering palliative care services to patients with end-stage disease, which reduced the rate of readmission by nearly two-thirds and lowered hospital-billed costs per patient by approximately $3,500 in the first 15 months of the program. Hospice and palliative providers have access to inpatient and outpatient medical records via the EHR, helping to ensure that patients who have serious and eventually fatal chronic conditions obtain seamless medical care across multiple settings, including home and hospital.

Owing in part to these programs, the cost of inpatient care at Gundersen Lutheran in the last two years of life was $18,359, or 29 percent lower than the national average of $25,860. The number of hospital days in the last two years of life was 13.5, nearly 43 percent lower than the national average of 23.6, according to data from the Dartmouth Atlas of Health Care on chronically ill Medicare beneficiaries.

For physicians who demonstrate leadership potential, the organization has created a physician learning community that helps them understand the strategic and business side of health care. The group meets monthly with the organization’s leaders to discuss institutional challenges and engage in problem-solving. The community also provides an opportunity to learn and practice leadership skills. Gundersen Lutheran does not use an incentive-based compensation system. Instead, salaries are set to be competitive in the market. Physicians are evaluated for productivity and citizenship; the latter is defined by adherence to a physician compact. They are also evaluated on measures of patient satisfaction, disease management, and patient access, which are recorded in the health system’s dashboard. The measurement feedback is critical to improvement.

To address this, Gundersen Lutheran uses data on clinical and financial outcomes to set goals for physicians to aspire to. Department chairs and administrators are also evaluated on such measures, which may include disease management targets and patient satisfaction measures, as well as measures of financial efficiency.

Gundersen has received national recognition and ratings, and The Commonwealth Fund Commission on a High Performance Health System identified Gundersen as a health sytem providing high-value care.

Several points come to mind. The power of partnerships and collaboration across the community. The influence of a palliative care network that spans acute, outpatient, and home settings. The importance of setting ambitious targets, and then monitoring and measuring performance. For LaCrosse, it took 15 years to become an overnight success. For most other communities, time is not on their side. HPM leaders would do well to accelerate their efforts to build attractive and influential palliative care "enterprises."

Spreading the Message and Science of Palliative Care

A recent study published in the New England Journal of Medicine identified the merits of early palliative care intervention, for patients with a lung cancer diagnosis. Among the benefits from such interventions were improved quality of life and longer survival. This study has received wide coverage in the mainstream press -New York Times (read here) , Wall Street Journal Health Blog (read here) , Boston Globe ( read here) , USA Today (read here) , and Associated Press (read here) and medical blogs, such as this blog post by the Deputy Medical Officer of the American Cancer Society).

While an important study, to be sure, it is hardly the first to tout the benefits of palliative care.
A study published in an August 19, 2009 issue of JAMA (read here) found that patients with advanced cancer who received a palliative care intervention focused on addressing physical and psychosocial issues and care coordination provided AT THE SAME TIME as cancer treatment had improved quality of life and mood. Interestingly enough, these patients did not experience a significant change in the number of days in the hospital or the severity of their symptoms compared to patients who received usual care.
In an earlier article, published in the June 2007 issue of Critical Care Medicine, (read here)
researchers at the University of Rochester Medical Center found that early palliative care interventions reduced the length of stay for seriously ill patients in the medical intensive care unit (MICU) by more than seven days without having an impact on mortality rates.

And, in early 2008, a Kaiser Permanente study (read here) was the first multi-center randomized trial to show that hospital-based palliative care results in improved quality of care, including higher patient satisfaction, improved communication, and fewer ICU admissions.

So is this most recent study a “game-changer”? Could be. Just like previous similar studies have been/could have been.

Evidence that palliative care intervention (regardless of timing) improves quality-of-life has been available for some time. Whether as a result of such evidence, or other factors, unquestionably there has been growth in the availability AND use of palliative care services. Most would also agree that the “spread of science” around palliative care has been slow, that is to say, programs whose results could replicate the benefits of these studies have been slow to produce similar results. Why?

It certainly hasn’t been for lack of effort. So then, one might ask, what are the obstacles? Simply put, two of the most formidable barriers are organizational and financial. Let’s take a closer look at this most recent study.

It was performed at a single, tertiary care site with a specialized group of thoracic oncology
providers and palliative care clinicians, thereby limiting generalization of the results to other
care settings or patients with other types of cancer. And I suspect that reimbursement for these palliative care services was not an issue.

Such care requires organizational models tightly integrated in order to produce similar results. As has been discussed in previous posts on this blog, such organizational models exist only in a few areas across the country. In most communities, palliative care has been viewed as an alternative to curative care, not as concurrent care. Moreover, current delivery models within today’s health care system (more on this in future posts) do not encourage nor facilitate the close coordination and collaboration among providers (individual and organizational) necessary to replicate this study’s findings. Nor do they encourage the adoption of evidence-based best practices. And, of course, the current reimbursement system, which rewards more procedures and aggressive care, surely does not encourage coordination. In future posts, we’ll explore the make-up of those communities (and its health care providers) that encourage tightly coordinated and bundled care.

A final thought on the study. Of course, more studies are needed to identify the value of palliative care. The efforts to promote study results should focus on VALUE (most significantly, quality), rather than longer survival (“Cancer Patients Live Longer With Palliative Care” screamed the headline from the consumer website Medicinenet.com). As Drew Rosielle, MD reminds us, “surveying the 'average' advanced cancer patient will reveal that most of them rate QOL (Quality of Life) as more important than survival" and we “should not spend too much time basking in the survival benefit”. And as Lyle Fettig, MD posts in Pallimed blog, “the PRIMARY (emphasis added) outcomes of the study stand on their own as important--it's yet more evidence that palliative care interventions improve quality of life and reduce psychological morbidity.”

David Weissman, MD puts it bluntly and best :
_____________________________

"While three cheers are certainly in order, I would be very cautious about saying that palliative care improves survival. Although randomized, the study was small and from a single institution. I love the results, but think we all need to be careful about over-interpreting single study results. When the next study shows no change in survival, which I fully expect, the fall for the field could be hard."
_____________________________

Tim Kirk , assistant professor of philosophy at the City University of New York where he specializes in philosophy of nursing and healthcare ethics with an emphasis on hospice and palliative care, writes “ All of these organizations have been focusing on the fact that patients in the study who received palliative care lived a bit (2.7 months) longer than patients who did not. While I understand that the hospice and palliative care communities have long suffered from the stigmatized perception that they welcome (or, even, hasten) death, and that this study seems to provide a rebuttal to such a perception, I believe the way the study is being embraced and promoted is misleading. He continues, "The emphasis on time to death perpetuates exactly the belief that hospice and palliative care have been working to change for decades: that what is important is how long you live rather than how well you live. If we begin promoting hospice and palliative care under the premise that such care extends life, we tacitly accept that extension of life is the primary goal of healthcare. And, we tacitly send that message to our colleagues, patients, and the public at large.”

More aggressive care is not necessarily associated with longer survival. That’s an important message in the efforts to lower cultural barriers to palliative care access. But as important, perhaps even more so, in achieving the “spread of science” in palliative care, will be how effectively we reconfigure organizational and reimbursement models that will improve access to, and delivery of, palliative care.

Certainly, there is a need to replicate this study in other settings. And since this intervention is probably very dependent on the skill of the palliative medicine providers, we need to better understand how various models of palliative care delivery improve outcomes.

Finally, I do not consider as insignificant the matter of prolonging life. Nonetheless, it is an issue with more relevance to other populations for whom palliative care intervention can be important. Alex Smith, MD describes it this way in the blog GeriPal,
“I would hazard a guess that those who would have the greatest survival benefit are geriatric patients with multiple co-morbid illnesses and no dominant terminal condition. Nothing like tailoring of medication, attention to physical, psychological, social, and spiritual concerns for frail elders. I think that's why so many patients with "general debility" and "failure to thrive" are discharged from hospice alive, only to fall apart again after discharge.”

As always, I invite your feedback.

Leadership Competencies in Exemplar Late-life Communities

What competencies are common among the leaders in exemplar palliative care communities, I'm asked from time to time. At the risk of oversimplifying, I'll suggest three

These community leaders, whether professional managers, physicians, or nurses, are particularly skilled at envisioning, energizing, and stimulating a change process that coalesces communities, patients, and professionals around new models of late-life care. These leaders have an uncanny ability to align their own priorities with those of the organization and the needs and values of the community. Call this a transformation competency.

These leaders display the ability to use metrics and evidence-based techniques to hold stakeholders to high standards of performance, using force of personality rather than the power of one's position. These leaders also understand the formal and informal decision-making structures around late-life care. In other words, they are adept at execution, translating vision and strategy into optimal organizational AND community performance.

And, these leaders are competent at building and sustaining relationships that evolve into networks, and take a personal interest in coaching and mentoring others. Put another way, these leaders possess exceptional people skills.

What competencies have I overlooked? I'd like to hear from this blog's readers.

Exemplar Late-Life Communities Five Years Later - The Role of the HPM Physician

In previous posts, we've featured interviews with physicians currently practicing Hospice and Palliative Medicine (HPM) in exemplar late-life communities, as identified by DAI Palliative Care Group in its study of end-of-life care data from the Dartmouth Medical Atlas. We're also followers of the Institute for Healthcare Improvement (IHI), and have followed closely IHI's identification of regions where high-value (low cost, high-quality) medical care is provided. One of our present studies is taking a closer look at the role of the HPM physician today in those communities identified both as exemplar late-life regions AND high-value medical care regions.

For more background on this issue, have a look at earlier posts for the regions of Grand Rapids,
Asheville, and Grand Junction. And for a look at exemplar late-life care hospitals, read here. More will follow in future posts. And, of course, your comments are invited.

How Do They Do That? - Providing High-Value Late-Life Care

More than 10 years ago, the Robert Wood Johnson Foundation funded a study that analyzed Americans' values, opinions and attitudes concerning end-of-life care. Based on in-depth focus group discussions involving 385 Americans in thirty-two cities as well as two sets of follow-up interviews, The Quest to Die with Dignity identified how Americans then thought about death and dying, how they want to be treated, and how they viewed planning documents such as living wills.

Some suggest that those participating in the study share similar, clearly articulated concerns, hopes and beliefs about the process of dying in America today. Dying well, for most of these Americans, means dying pain-free, relieved from all suffering. For many, it means dying at home, surrounded by loved ones, and untethered to machines.

Are there communities that do a better job of complying with patient preferences? Communities that could be considered "bellwethers" or "exemplars", communities where high-value late-life care is provided? Using data drawn from the Dartmouth Medical Atlas, such communities can be identified.

Recently, the Institute for Health Care Improvement(IHI), an organization led by Don Berwick, MD, who will soon be leading CMS, identified ten communities where high-value care is being provided. We've taken a look at those communities, to analyze whether there's a connection between their success stories in providing high-value health care and their performance around late-life care. and we've found that some regions do provide high-value health care and patient-centered late-life care.

A previous post on this blog took a closer look at Asheville, North Carolina, one of 10 communities identifed by IHI in their How Did They Do That? study. Among the lessons from the IHI study:

Some patterns: a culture of collaboration to put patients first; considering finances as a constraint, not a goal; the importance of physician leadership; the real or virtual integration of delivery systems across the continuum of care; the importance of strong primary care; the value of electronic health records and the information-sharing it makes possible; and an emphasis on measuring and reporting data on quality and utilization.

We'll take a closer look at some of these communities, in an effort to better understand why some regions' late-life practices produce outcomes that match patient preferences better than other regions.

Accountable Palliative Care Organizations (APCOs)

Recently I was asked by a client to describe the role of Accountable Palliative Care Organizations (APCOs) in creating exemplar practices around "late-life" care within a community (see here for a previous post describing exemplar communities). The client is situated near a shopping mall that had been struggling until it recently brought in a major department store - a topic which had been a subject of an earlier conversation that day. An APCO, I replied, is like an "anchor tenant" of the palliative care community, setting norms to encourage the free-flow of ideas and collaboration, producing enduringly successful communities.

Within these APCOs, physicians , hospices, hospitals, and long-term care facilities adopt measures to blunt harmful financial incentives, thus taking collective responsibility for improving care for those with advanced illnesses (what I'm terming late-life care). Much has been written and commented of late about the role of financial incentives in the health care system, and what provisions in the health care reform bill could bring about better outcomes while containing costs.

I'm curious to learn your thoughts and experiences, as we explore this subject in greater depth in future posts.

Why Organizational Delivery Models Matter to Effective Palliative Care

There are multiple examples of health systems -- President Obama and health policy makers have cited Mayo, Geisinger, Cleveland Clinic, and others as models for health-care reform -- that consistently and reliably achieve similar results: providing good care at low cost, with high patient satisfaction. Bassett Healthcare, serving the Cooperstown, New York region, is among those others. What these systems have in common, we're advised, is that they are integrated systems that employ their physicians, emphasizing patient-centered care, better outcomes, and prudent stewardship of health-care resources, with accountability for results.

We've looked closer into these hospitals and communities to understand better if this model also produces desirable results around late-life care. We've learned previously that Geisinger has seen exemplay results in its late-life care practices. So we took at look through the Dartmouth Medical Atlas at the formerly named Mary Imogene Bassett Hospital, now known simply as Bassett Medical Center. The Hospital's results in the seven DAI Palliative Outcome Measures earned it an exemplary grade. Patients loyal to Bassett were 20% less likely to die in the hospital, spent almost 40% fewer days than state average in a hospital during the last six months of life , and were more likely to have been under hospice/palliative care. Yet perhaps the most telling is that in their final months, 30% fewer patients were seen more than ten physicians during late- life care. You read this correctly, TEN is the benchmark.

Thus, we find that in another community a tightly integrated health system produces desirable outcomes in late-life care. What remains unclear is what attributes of a tightly integrated health system are most responsible for these results. With that knowledge, these attributes might be replicated elsewhere (as in an Accountable Palliative Care Organization).

I'm curious to learn your experiences and thoughts.

Grand Junction - an Exemplar Palliative Care Community

President Obama held a town hall recently in Grand Junction, Colorado. To healthcare reformers, Grand Junction, CO., is one of the areas of innovation -- a place that provides high-quality healthcare at a fraction of the costs in most other communities. And, according to some experts, Grand Junction chalks up some impressive statistics. Only 12% of Medicare patients required readmission 30 days after a hospital visit, as opposed to the nationwide rate of 20%. Children on Medicaid in the HMO are four times as likely as other Colorado Medicaid children to receive all immunization treatment -- and adults on Medicaid were up to 10 times as likely to get comprehensive diabetes care. How does Grand Junction stack up for its palliative care practices? The Community Palliative Performance Profile, compiled by DAI Palliative Care Group, graded Grand Junction as an A-plus (an exemplar community). Why did Grand Junction earn this superlative? For starters, less reliance in the final months of life upon intensive care (less than half of national average) and one of lowest percentages of deaths occurring in a hospital (20.7%). And hospice enrollment nearly 30% greater than the national average. Is it mere coincidence? We think not, as a profile of one of the senior palliative care physicians in Grand Junction will show in future posts.