Brown University School of Medicine Dying Trajectory in the Last Year of Life: Does Cancer Trajectory Fit Other Diseases? Joan M Teno, M.D. M.S. Associate Professor Brown University School of Medicine Good Morning, On behalf of my co-authors, Ms. Sherry Weitzen, Dr. Mary Fennell, and Dr. Vincent Mor, I present the following presentation that examine the dying trajectory in the last year of life.

Dying in the US Institutionalization of Death - historical changes from home to institutional death (hospital and NH) Hospice Medicare Benefit based on cancer disease trajectory Americans fear dying in pain, surrounded by technology, and isolated from love ones. The majority of Americans now die in an institutional setting either a nursing home or an hospital. The Hospice Movement was born out of these concerns, with the goal of allowing persons to die in familiar settings with the best possible quality of life. Yet, only a minority of the dying use hospice services and a cancer diagnosis is predominant among users. In part, the fact that majority of patients in US hospice have cancer may reflect the dying trajectory that is exhibited by person dying of cancer.

Hospice Medicare Benefit Created early 1980s, as both cost containment and improving the quality care of the dying For persons with “expected six month survival”, but never defined that term. Recent debate as focused on issue of ACCESS and ABUSE In the United States, the Hospice Medicare Benefit was created in the early 1980s as both a means of cost containment and improving the quality of car of the dying. Persons with Medicare (an health insurance provided by the United States Federal Government) must sign an agreement that they accept non-curative treatment and their physician must state that patient has an expected six month survival. This term was never defined in the legislation by the US Government. Recent debate in the US has focused on the issue of Access (in that less than one in five person who die utilize hospice) and abuse. The Inspector General Office of the Federal Government has been investigating cases where person live longer than six months as care that potentially is defrauding the US Government. The US Medicare Hospice Benefit was created based largely on a cancer dying trajectory. Current debate has focused on how this benefit can appropriately be extended to include persons dying of other disease with potentially, quite different dying trajectories. Central to this debate is whether the dying trajectories of non-cancer illnesses fit the expected trajectory covered by the Medicare Hospice Benefit.

Goals Describe functional trajectory in the last year of life for cancer and leading non-cancer causes of death. Examined its association with site of death and hospice involvement The goals of this research effort is to …

Methods-Sample National Mortality Followback Survey-1993 Death Certificate and Questionnaire 1% sample of > 15 yr olds in US Deaths from chronic, progressive illnesses N(weighted)=1,980,388 Since the 1960s, the National Center for Health Statistics has conducted representative studies of all deaths in the United States based on in-depth interviews with bereaved next of kin listed on death certificates. The 1993 survey represents the sixth national survey that was a 1% sample of dedents age 1 and older. We restricted our analyses to those deaths from chronic, progressive illness. Cancer, Congestive Heart Failure, Chronic Obstructive Pulmonary Disease, Diabetes Mellitus, Stroke. These latter four diagnosis are the four leading non-cancer cause of death in the USA. The National Mortality Followback Survey is weighted to represent 1,980, 388 deaths .

Methods-Activities of Daily Living ADLs: Getting in and out of bed or chair Eating Using the toilet Dressing Bathing Ex of question: “During ___ entire life, how long did ___ have difficulty bathing” measured in days Our focus was on describing changes in the Activities of Daily Living in the last year of Life. Next of kin were asked the following question: During (NAME) entire life, how long did he have difficulty “bathing” measured in days? This question was ask for Getting in and out of bed or chair Eating Using the Toiliet Dressing Bathing We created a variable that indicated the number of ADL impairments that the decedent had for every month in the last year of life.

Sample Description The following table describe characteristics of persons who died from Cancer, CHF, DM, COPD, and CVA (cerebral vascular accident). Person were old with mean age greater than 70, with differences in gender reflecting the biology of what disease impact women as compared to me. Person who dies of cancer were more likely at home while those dying of DM, COPD, and stroke were more likely to die in the hospital. Congestive heart failure had the highest rate of dying in a nursing home, with about one in three person dying in nursing homes.

Results-Functional Decline Figure 1 Figure 1 depicts the rate of functional decline during the last 12 months of life after adjustment for age of the decedent. The X axes goes from 12 months prior to death on left end to the day before death on the extreme right. One-year prior to death, persons dying from cancer were less functionally impaired than were other decedents at a similar point in time. Only 20% of the persons dying from cancer had any ADL impairments one-year prior to death, while more than 40% of non-cancer causes of death were already ADL impaired. The mean number of ADL impairments 12 months prior to death in non-cancer causes of death varied between 1.3 and 1.6. All five diseases exhibited a functional decline that was relatively flat until the last five months of life. At that time, persons dying from cancer went from a mean of 1.0 to 3.0 ADL impairments, while those dying of non-cancer causes increased their mean ADL impairment from 1.6 to 2.3. For each ADL and mobility items, the number of day difficulty functioning was statistically significant (i.e., p<.001) for cancer compared to non-cancer decedents as test by one way analysis of variance. P <.001

Results- Functional Decline Figure 2 Figure 2 depicts the percentage of persons limited in mobility to the point that they can not transfer out of bed or chair by month preceding death. Only 13.9% of cancer patients had difficulty getting out of bed or a chair a year prior to death. This increased from 22.2% to 63.0% in the last five months of life. People with non-cancer diseases had higher rates of functional impairment one year prior to death (about 35% had difficulty getting out of bed or chair), but they experienced a gradual increase in the levels of functional impairment than did cancer decedents (about 50% had difficulty getting out of bed during the final months of life).

Results-Functional Decline Table 2 The pattern of functional decline was associated with hospice involvement and site of death. Among decedents with a sharp functional decline in the last 5 months of life, 29% had hospice involved in their care compared to 6.5% hospice involvement when there were no changes in ADL impairment (Mantel-Hanzel test of linear trend, p <.001). Similarly, persons with increased functional decline were more likely to die at home (see Table 2). P <.001, for the Mantel-Hanzel Test of Linear Trend

Study Limitations Data obtained retrospectively, from proxy respondents- Secondary analysis-limited to questions that were asked in NMFS No Questions about patient preferences for site of death The limitations of this study are as follows

Conclusions Non-Cancer Diagnoses have different patterns of functional decline in last year of life. The pattern of functional decline was associated with site of death and hospice involvement Cancer Patients are more likely to die at home--because of more predictable trajectory In conclusion, Those person who died of a non-cancer diagnoses exhibited a different pattern of functional decline in the last year of life. The pattern of functional decline was associated with site of death and hospice involvement. Cancer patients, who have a more predictable dying trajectory, were more like to die at home and have hospice involved in their care. We hypothesize that the rapidity of the functional decline may serve as a cue to the physician, patient, and family that the person is dying. Additionally, the functional decline and resulting diminished quality of life may influence dying persons and their family to accept the explicit tradeoff of the hospice (i.e., forgoing active treatment for increase supportive care). However, the majority of patients did not exhibit that classic rapid functional deterioration.