Article Archive
Spring 2011

Becoming ADEPT at Predicting Mortality

By Juliann Schaeffer
Aging Well
Vol. 4 No. 2 P. 24

Patients with advanced dementia could benefit from hospice care yet due to Medicare admission criteria some researchers say is flawed, many of these patients have gone without such care at the end of life. However, in an effort to improve access to such comfort care, researchers have developed a new tool to more accurately reflect a six-month prognosis for advanced dementia patients: the Advanced Dementia Prognostic Tool (ADEPT).

Barriers to Hospice Care
Through her previous research, Susan Mitchell, MD, MPH, a senior scientist at Hebrew SeniorLife and an investigator at the Institute for Aging Research, found that nursing home residents with advanced dementia (classified as those who are bedbound, nonverbal, and incontinent and who no longer recognize family members) were often not receiving the palliative care so important for this population, in part because advanced dementia was not perceived as a terminal condition by providers, families, or caregivers.

Recognizing that shortcoming, Mitchell focused much of her research over the past two decades on gaining recognition for dementia as a terminal illness. “In looking at statistics, [dementia is] now the fifth leading cause of death for people over 65, and I believe that’s probably an underestimate,” says Mitchell. “Studies over the last five or six years are beginning to convince not only families but providers and policy makers that this is a very common and costly end-of-life condition. Advanced dementia is a terminal illness, and research data has really shown that these patients are at the end stage of a disease and surely meet the criteria in terms of time frame for an end-of-life condition.”

“They also suffer from high rates of the types of burdensome symptoms that characterize end-stage disease, like pain and dyspnea,” Mitchell adds, noting that much in the palliative care field could assist with treating such symptoms as well as provide support for families dealing with the end stage of disease. “Just like metastatic cancer or AIDS, end-stage dementia certainly meets those criteria.”

However, hospice, a Medicare benefit program, has specific admission criteria that primarily center on an estimated prognosis of less than six months, which for various reasons is especially difficult to predict for dementia patients. Mitchell explains the flaws behind such criteria from her perspective: “The current criteria for hospice were never designed empirically. When people have tested it, including myself, they found it did not accurately predict six-month prognosis. And because hospice providers are reluctant to admit people who may stay for long periods of time, the challenge of prognostication has been a barrier for enrolling people in hospice.”

“Very few individuals with dementia receive the hospice care needed, and when they do, it is not in a timely fashion,” says Shellie Williams, MD, an assistant professor of medicine in the Section of Geriatrics and Palliative Medicine at the University of Chicago Medical Center. Besides dementia’s lack of recognition as a terminal illness, she points to a “lack of recognition that the natural progression of dementia is the state of diminished mobility, oral intake, and diminished communication, which cannot be reversed or cured with interventions such as feeding tubes, IV fluids, or recurrent hospitalizations, which often delay the initiation of needed hospice care.”

New Tool, Better Hospice Access?
Mitchell sought to rectify this access problem by fixing what she sees as the current Medicare criteria flaws. She and colleagues used the Minimum Data Set (MDS) of the national data set of nursing home patients, which consists of more than 1 million people, and attempted to “use the most sophisticated statistical models to try to make a predictive tool that was practical,” she says. Through their research, Mitchell and colleagues were able to identify patients with advanced dementia and, through analysis, selected the 12 variables that together best predicted survival, leading to the formation of ADEPT.

“ADEPT is a prognostic tool for assessing six-month mortality of individuals with dementia,” explains Williams. She notes that ADEPT differs from the traditional criteria in various ways, with criteria “broader in the sense that they include many of the traditional symptoms seen in all patients with limited prognosis and are not solely focused on secondary manifestations of severe dementia [anorexia, aspiration, mutism, immobility].”

The main difference between the tools? “ADEPT is based on 12 empirically chosen variables from the MDS vs. hospice guidelines, which are based on the FAST [functional assessment staging criteria] and randomly chosen conditions over the past 12 years,” according to Mitchell.

After formulating ADEPT, Mitchell tested it by applying the tool at the bedside of more than 600 nursing home residents with advanced dementia and then compared these criteria with the current hospice eligibility criteria. The study found that the ADEPT tool performed slightly better than the hospice guidelines. “Using the ADEPT tool on a [scale] of 0 to 13, it was able to classify people who died or not in six months 68% of the time compared to Medicare guidelines, which was able to classify people who died or not 55% of the time,” says Mitchell. “So we were a little bit better, but we weren’t 100% predictive.”

So is this tool the answer to getting advanced dementia patients better access to hospice care? Maybe partly, but Mitchell recognizes the limits of this or any tool in estimating six-month prognosis and believes the true answer may be found through a different direction. “Frankly, I don’t think this moves things a whole lot forward except to know that it’s very, very hard to predict with any accuracy whether someone with advanced dementia is going to die in six months. [ADEPT] does it better than hospice guidelines … and to the extent that they’re going to stick with the six months [eligibility criteria], ours works a bit better.

“But what I think the study really showed is that we can’t predict [end of life for dementia patients] really well and that access to hospice and palliative care for dementia patients should be based on people’s preferences for comfort as the main goal of care rather than expected six-month prognosis,” she says.

Keys to Better Hospice Care
Mitchell and Williams agree the most significant issue relates more to the goals of care for dementia patients and educating caregivers about the overall benefits of palliative care.

“Many of the healthcare power-of-attorney [advocates] for individuals living with dementia often feel that focusing on hospice is giving up on or limiting care rather than the truth, which is that hospice is escalation of care focused on treating the pain and the constellation of end-of-life symptoms [such as] nausea, depression, and shortness of breath that can be treated and cured,” explains Williams. “When there is lack of this realization by advocates for those living with dementia, they are often subject to recurrent hospitalizations, procedures, and side effects of interventions, which cause worsening pain and symptoms progression during the final days and weeks of these individuals’ lives.”

Mitchell believes that with proper patient and family counseling, most people with advanced dementia would opt for comfort as the primary goal of care (ie, hospice or palliative care). Yet she explains that one reason patients don’t receive such care is a dearth of clearly established goals of care.

“For example, if a patient gets pneumonia, we pretty well know they can be just as comfortable or do just as well in a nursing home. But if the family doesn’t have adequate counseling and discussion, then they may end up in the hospital, which is a very traumatic experience for both patients and families and very expensive for the healthcare system,” she says. “A lot really ties in with the work that’s done before the actual illness, which involves counseling families, making them understand that there is a limited prognosis, and then looking at the most common problems that will arise and trying to align the treatment options with their goals.”

Mitchell says if this proper counseling were conducted with families, “There’d be a lot less burdensome treatments for patients, a lot more satisfied family members, and a lot lower acute care costs.”

While Williams says this tool is “a good start to considering where we are with individual patients and considering what areas of documentation may help support a referral to hospice,” the real take-home message for physicians is in educating dementia patients and their caregivers about the reality of the disease process, so goals of care are aligned with the best interests of all involved.

“We as physicians should serve as advocates to patients and families affected by dementia by informing them of the fact that this disease is progressive and terminal. We should be open about this during the beginning of disease identification to allow patients and families to assess their goals, establish their wishes when this terminal stage arrives, and establish a plan of care that is consistent with the individual’s wishes,” she says.

— Juliann Schaeffer is an associate editor at Great Valley Publishing Company and contributing writer for Aging Well.