Neil S. Wenger, MD
UCLA Dept of Medicine-GIM
"I’ve always loved taking care of the complex patient -- the individual who not only has complex medical issues but whose care also raises complex social and ethical questions.”
Efforts to measure healthcare quality—with an eye toward improving it—started gaining momentum more than 20 years ago. But early advocates for quality measurement were employers concerned about the increasingly hefty tab for their employees' health coverage. As a result, says Neil S. Wenger, MD, MPH, the quality measurement movement focused primarily on working people—to the exclusion of older adults.
"Yet vulnerable older adults particularly need healthcare evaluation," says Dr. Wenger, who has been at the forefront of efforts to ensure older people get just that.
An internist whose board certified in geriatrics, a professor in the Division of General Internal Medicine and Health Services Research at UCLA and director of UCLA's Healthcare Ethics Center, Dr. Wenger is also the director of the Assessing Care of Vulnerable Elders (ACOVE) project.
Over the last decade, he and fellow ACOVE investigators have developed a groundbreaking set of quality indicators (QIs) to evaluate the quality of care that vulnerable older adults receive. The indicators are based on a comprehensive review of data applicable to vulnerable older people—adults 65 and older whose health problems put them at risk of losing their ability to carry out daily activities and their independence—and the expertise of experienced clinicians. An estimated one-third of older Americans meet the description of "vulnerable," if not frail. They are the largest consumers of healthcare in the US.
Dr. Wenger and colleagues published the first set of ACOVE quality indicators in 2000. They expanded and revised the set in 2002 and again in 2007. The indicators now cover a wide range of common age-associated health problems and healthcare needs, such as dementia, malnutrition, pressure ulcers, urinary incontinence, falls and mobility problems, continuity and coordination of care, and end-of-life care.
In addition to developing, testing and updating the indicators, the ACOVE team has also created a now widely used screening tool to identify vulnerable, high-risk older adults. And it's used the indicators in studies evaluating healthcare for these adults. The team's 2003 study evaluating the medical care vulnerable older adults receive for age-related conditions such as malnutrition, dementia, and pressure ulcers found that this care fell short of recommendations a staggering two-thirds of the time. To help rectify this, the ACOVE team has also designed and assessed interventions based on the quality measures that have been shown to improve primary care for vulnerable seniors.
In recognition of their work, the RAND Corporation, the UCLA School of Medicine, Dr. Wenger and his ACOVE colleagues were named recipients of the Joint Commission and National Quality Forum's prestigious John M. Eisenberg Patient Safety and Quality Award for Research.
The American Geriatrics Society talked with Dr. Wenger shortly after the award was announced.
Q: What got you interested in geriatrics and in developing quality measures for vulnerable older patients?
A: Older adults tend to have more complex health problems than younger ones. I've always loved taking care of the complex patient—the individual who not only has complex medical issues but whose care also raises complex social and ethical questions. Quality and ethics are areas of great interest to me.
Taking care of patients in such a way that you see all the different pieces, and all the different bits of care being provided, and put them together into one seamless whole, is, in my view, what medicine is all about. And it's clearly the focus of geriatrics. It's what makes geriatrics so important.
Q: Why do vulnerable older adults need their own set of quality measures?
A: Vulnerable older people need these for several reasons. While many of the clinical conditions common in older adults are also present in the general adult population, some are not. These include cognitive impairment, pressure ulcers, falls and the need for end-of-life care. These are areas that early quality measurement work had barely touched. The few measures that did exist focused on care for those in the nursing home, and weren't applicable to vulnerable elders living in the community.
Another reason for developing these measures is that quality measures generally assume that patients' goals for care universally focus on longevity. Typical measures don't integrate quality of life considerations, or preferences for comfort over longevity, into the equation. These are important considerations for vulnerable older patients.
In addition, older patients often cannot adequately advocate for themselves. As opposed to a 20-year-old or a 50-year-old, who might insist on good care, an 80-year-old with dementia can't do the same. That's another reason we undertook this work.
Older people also receive a disproportionate proportion of healthcare services. So it's critical that these services are appropriately applied.
Q: How does ACOVE define "vulnerable elders"?
A: Typically, "vulnerable elders" would be defined as those who use a lot of care, which makes sense, since that's the case. But that definition leaves out the undiagnosed and those who aren't receiving the kind of care they should.
ACOVE developed a definition for "vulnerable elders" using a survey, rather than utilization data. The survey was developed by looking at which factors would predict functional decline in someone who's 65 or older and is living in the community, over a two year period. Our goal was to identify a group that—with good care—one might be able to keep at their current functional status, or even improve it.
This work, led by Dr. Debra Saliba, who's also affiliated with UCLA and RAND, focused on using a national data set to see which predictors would be most highly correlated with functional decline and death in this group. It turns out it's not having certain diseases that predicts decline; rather, it's factors that have to do with the person's level of function, their age and their perception of their health. The survey—called the Vulnerable Elders Survey 13 or VES 13—includes 13 items to be asked of a person 65 years or older, or their proxy. The focus is on function, with additional items on age and perceived health.
Q: Critics often argue that quality measures aren't patient-centered and don't apply to complex patients with multiple conditions. How does ACOVE address this?
A: One of the major concerns about QI measures is that they are cookie cutter, and assume that every patient should receive the same care. Our response to this has been to do as much as possible to take into account all relevant information, and that means including information that goes beyond what a particular claims data set would include.
A typical quality measure, for example, would say, "If patient has a disease that we know about because of claims from a doctor or hospital for that disease, then the patient ought to be taking a certain medication, and we know whether they are taking it by looking at pharmacy data." Now, what a measure like that doesn't consider is: Has this medication been tried in the past, but was unsuitable for the patient? Has the patient had side effects they can't tolerate or prefer not to? That's why we've been reluctant to use claims data alone to look at quality.
We believe you need a much better sense of what's happening in the clinical setting. To date, the best way to do this is to carefully review medical records describing the clinical encounters. Those records may note that a patient did not want to receive a certain drug or certain care process. Or it may identify subtle reasons that this treatment was not optimal for this patient. Taking this step leads to higher - and more clinically accurate - evaluation that better reflects the actual care being provided. This has been a cornerstone of ACOVE measurement.
Yes, it's thought that quality measures are all well and good if a patient just has one condition or two, but if you take a patient who's really complicated, then is it fair to apply quality measures to her? We studied this. We investigated the relationship of quality to the number of conditions that vulnerable older patients have. And, interestingly, we found that quality, as measured by the ACOVE measures, does not go down in more complex patients. If anything, you get a gently up-sloping curve, with the QIs indicating that more complex patients are getting higher quality care. And this is not entirely because they receive care more often or because they see specialists more often. The ACOVE measures, and others built in the same way, are clinically detailed enough that they really can fairly evaluate complex care, and won't necessarily penalize providers caring for patients with multiple conditions.
Q: So, does ACOVE set a bar for clinicians to meet?
A: In the development of these quality measures, we've attempted to define a low bar. We're trying to define the elements of care for complex older patients that absolutely must be provided. Our findings suggest that doctors are able to provide these necessary elements of care. But often they don't. And figuring out why that's the case is the next thing we need to do.
Q: Could you talk about the variety of ways in which the ACOVE measures are being used?
A: They're used in quality of care research, but also in clinical practice and in other ways. Medical groups and health plans use the measures to evaluate the care they provide and to identify areas to improve their care.
Another way to use the instrument set is to translate it into a quality improvement guide. Right now we're finishing up a study of medical practices that wanted to improve the care they provide to patients who have fallen. Based on the quality measures, the ACOVE team, led largely by UCLA's Dr. Dave Reuben, has developed a chart-based tool to guide them in providing the best possible care for these patients. It isn't a care pathway, but, rather, a framework for the best way to approach care for someone who's fallen. We've similarly developed tools for patients with cognitive impairment and incontinence.
I'm told that bits and pieces of the ACOVE measures are used by a wide variety of groups to measure particular aspects of care. For example, a medical group has measured its depression care by employing the ACOVE depression set of measures. Others have used it to look at their falls care or end-of-life care. It's difficult to get a handle on how often the measures are being used, but my understanding is that elements are used commonly.
One of our hopes is that the literature that's the basis of the ACOVE measures will also be used as an educational tool. And a variety of residency and fellowship programs do use it in this way. The measures are also starting to be used, to a degree, by some groups that promulgate national standards.
Q: Could you talk a bit more about how the ACOVE indicators have been used to develop interventions for vulnerable older adults?
A: There have been a number of attempts to change the way primary care doctors approach older community-living patients, especially those who have problems with falls or cognitive impairment or urinary incontinence. We're now completing three studies of implementing such interventions—basic interventions—for these problems. In collaboration with the American College of Physicians, the ACOVE QIs are being used in a national study focused on falls and incontinence. The ACOVE set is also being used in a collaborative project with the Alzheimer's Association to improve care for patients with cognitive impairment.
The results are in for one study so far and it's showed that both falls and incontinence care improved substantially with these interventions. But care was so bad at baseline that even doubling the quality score only got us up to a point where 40 to 50% of desirable care processes were being done. The results of the cognitive impairment study are about to come out.
Our goal is to find ways to better employ these measures to develop the skills of practitioners and change the way practices approach older people so ACOVE measures are always satisfied. The dream is to become obsolete.
Q: You mentioned the Vulnerable Elders Survey 13—can this be used as a screening tool to identify the vulnerable elderly?
A: Yes, it can. In fact, many places - from internal medicine practices to cancer clinics - are using it to identify vulnerable older people who ought to be targeted for special care. There are an increasing number of individuals using it for screening. Even though it's simple, the survey identifies people at risk of decline - even in time periods as short as a year. Like all the ACOVE measures and everything else we've developed, it's in the public domain and available online, at http://www.rand.org/health/projects/acove/.
Q: What have been ACOVE's most important contributions to patient care thus far?
A: The project's most important contribution is getting the world to recognize that vulnerable older people are not like typical adults. To care for them we need special attention, special skills, and special measures.
