Impact of Serious Illness on Patients’ Families: Findings of the Support Project

The SUPPORT project was a landmark study of prognoses, preferences, and decision making in patients with one of nine serious life-threatening illnesses. One of the goals of the SUPPORT project was to develop a better understanding of the impact of serious illness on the families of patients. Surrogate and patient interviews about the impact of illness on the family were obtained for 2129 patients who survived an index hospitalization. The interviews were obtained between 2 and 6 months after the hospitalization. Patients represented a wide age range (15% < 45 years; 45% > 65 years).

Table 47-1 describes the frequency of burdens described by SUPPORT subjects. Families noted several severe burdens. For example, 20% reported that a caregiver made a major life change to care for the patient, and this often included quitting or taking time off from work. In 31% of cases, most of the family savings was lost, whereas in 29% of cases, a major source of family income was lost. Seventeen percent reported a major change in family plans, such as moving to a less expensive home, delaying medical care for another family member, or altering educational plans for another family member. At least one major burden was reported by more than half (55%) of families.

Although rates of burdens were high for all patient subgroups, several characteristics defined families most likely to suffer adverse financial impacts. For example, loss of most of the family savings was more common in families of patients with severe functional dependence than in those without severe functional dependence (37% vs. 29%) and in families with incomes higher than $25,000 a year than in those with lower incomes (35% vs. 22%). Young patient age was, however, the strongest predictor of loss of family savings. Loss of savings was reported for 43% of families of patients younger than 45 years compared with 25% for patients older than 75 years. There are probably several reasons families of younger patients were at higher risk than families of older patients. Because most older patients receive Social Security, their income is less likely to be threatened by severe illness. In contrast, it is more likely that younger patients and their caregivers will need to give up work as the result of severe illness. Moreover, older patients are more often eligible for social and other support services than younger patients, and it is possible that some of these services offer some measure of protection.

The SUPPORT study did not identify the mechanisms of these burdens. However, because 96% of patients in SUPPORT had health insurance, it is likely that many of these burdens were the result of expenses not covered by insurance. Undoubtedly, a major source of financial burden is the loss of employment income for patients and caregivers. Many illness-related costs associated with terminal illness are not covered by most health insurers. This includes the costs of home health aides and assistance with basic activities of daily living (ADLs), home modifications, and long-term care costs. Although some of these costs are covered by hospice, many terminally ill patients are never enrolled in hospice, and those who are enrolled usually enter hospice for the last few weeks of their illness.

Commonwealth-Cummings Study

The Commonwealth-Cummings Project shed some more light on the mechanisms of burden that the SUPPORT project did not identify. Within 6 selected regions, 988 terminally patients (identified by their physicians as having a life expectancy of less than 6 months) and 893 of their caregivers were chosen for interviews. Fifty-nine percent of the subjects were more than 65 years old. Overall, 35% of patients had at least moderate need for nursing home care, homemaking, or personal care. Eighteen percent had unmet needs for nursing care, and 23% had unmet homemaking needs. Greater functional burdens, older age, and low income were associated with greater care needs. Substantial care needs, female gender, and African-American ethnicity were associated with unmet care needs.

The study strongly suggested that substantial care needs were a key contributor to economic and other burdens. For example, significant depressive symptoms were reported by 31% of caregivers of patients with high care needs, compared with 25% of caregivers of patients with few needs. Similarly, caregivers of patients with high care needs were more likely to report that caregiving interfered with their personal needs (36% vs. 24%). Interestingly, having an empathic physician seemed to buffer these burdens. Among caregivers of patients with high caregiving needs, those who reported that the needs and opinions of the caregiver were addressed by the physician were less likely to be depressed (28% vs. 42%) and less likely to feel that caregiving interfered with their personal life (32% vs. 48%).

The results of this study suggest an explanatory model for the burdens critical illness places on caregivers. In this model, factors such as poor performance status, older age, incontinence, and low income lead to high caregiving needs. These high caregiving needs lead to both economic burdens and health burdens such as depression.

Economic Impact of Dementia

Dementia differs from other end-of-life conditions because the later stages of the disease are often protracted, and the prognosis is unpredictable. Dementia also imposes special challenges on a patient’s family because patients often require extensive caregiving assistance. This caregiving assistance is generally required because of a combination of memory impairments, behavioral disturbances, and impairments in physical functioning. Although dementia is the leading cause of nursing home placement, most patients with dementia are cared for at home by family members. However, although it seems probable that dementia care would impose high costs on families, only a limited number of studies have addressed the economic costs of dementia care.

One of the most important economic costs of dementia caregiving may be on the ability of caregivers to work. Because many caregivers need to provide continuous supervision, the ability to be employed is often compromised. In a study of patients newly enrolled in the Program of All-Inclusive Care for the Elderly (a new capitated benefit authorized by the Balanced Budget Act of 1997 that features a comprehensive service delivery system and integrated Medicare and Medicaid financing), 25% of patients with dementia had at least one caregiver who either quit working or reduced the number of hours worked to care for the patient. Loss of employment was considerably more common in patients of African-American and Hispanic ethnicity, as was the case in the Commonwealth-Cummings Study of terminally ill patients, a finding suggesting that this economic burden of caregiving may be disproportionately endured by minority communities. Loss of employment was also more common when the caregiver was a woman.

One method of estimating the value of the time family members spend caring for patients with dementia is to estimate the opportunity cost of caregiving. Opportunity costs of caregiving generally represent the estimate of how a caregiver’s time would be valued if they could provide their services within the labor market. Using this approach, Langa and colleagues used the Asset and Health Dynamics (AHEAD) study, a nationally representative sample of community-dwelling elderly persons, to conduct a study of the cost of informal caregiving for elders with dementia. The market value of caregiving was estimated using the average wage of a home health aide. In 2001, Langa and colleagues determined that patients with mild, moderate, or severe dementia required an additional 8.5, 17.4, and 41.5 hours, respectively, of caregiving per week compared with elders without cognitive impairment. The yearly opportunity cost of this caregiving was $3600, $7420, and $17,700, respectively, in these three groups. This represented a national annual opportunity cost of more than $18 billion. This estimate is striking because the total national expenditure for all paid home services (for all conditions) was $29 billion in 1998. Because the $18 billion cost estimate considers only one condition, it seems almost certain that the economic cost of family caregiving (generally unpaid) exceeds the current economic cost of paid home care.

Caregiving also has substantial personal costs. Although these costs are incurred in many end-of-life illnesses, they are best documented for caregivers of patients with dementia. In particular, caregivers of patients with dementia have repeatedly been found to be at higher risk for depression. In a study of more than 5000 patients with advanced dementia, for example, 32% of caregivers were classified as depressed based on a high number of depressive symptoms. Rates of depression varied depending on patient and caregiver characteristics. Patient characteristics associated with higher rates of caregiver depression included younger age, white or Hispanic ethnicity (compared with black), low education, dependence on the caregiver for activities of daily living, and behavioral disturbances. Caregiver characteristics associated with high rates of depression included low income, female gender, more hours spent caregiving, and poor functional status. Thus, caregiver depression appears to be a complex process that is influenced by ethnicity, as well as by patient and caregiver characteristics. Depression is also associated with labor productivity losses. In one study, workers with depression had 5.6 hours per week of lost productive time compared with an expected 1.5 hours per week. Most productivity costs were explained by reduced performance while at work.

Some evidence suggests that caregiving for a frail elder is sometimes associated with more global health effects. For example, Schulz and colleagues demonstrated that caregivers who reported emotional strain while caring for disabled spouses had a 63% higher risk of death than nonspousal caregivers. Because spousal caregivers who did not have strain were not at higher risk for death, this study illustrated the importance of identifying caregivers who are experiencing adverse emotional outcomes from caregiving. There is also evidence suggesting that dying a “good death” by receiving palliative care services in the form of hospice may also have positive effects for spouses. Christakis and Iwashyna retrospectively matched 30,838 couples, in which the decedent in one group used hospice while the matched decedent in the other group did not. After 18 months, bereaved wives whose husbands had used hospice showed a significantly lower mortality than those wives whose husbands did not use hospice. This finding suggests that high-quality palliative care has the potential to soften the blow of widowhood.

Economic Impact of Cancer

Cancer is a leading cause of death, particularly among elderly persons. Although cancer affects all ages, the studies on the economic costs of cancer to patients and their families have focused on the elderly. Although initial studies used small convenience samples to examine the out-of-pocket costs related to cancer, more recent studies have exploited the data available from the 1995 AHEAD study. A study by Langa and colleagues examined the relative out-of-pocket medical expenditures of community-dwelling elderly persons who were more than 70 years old and who had cancer. In this sample of 6370 respondents, 84% reported no cancer or history of cancer, and 13% reported a history of cancer but no current treatment regimen, whereas 3% reported a history of cancer and current treatment. Adjusted mean annual out-of-pocket expenditures were significantly different at $1220, $1450, and $1880 for each of the groups, respectively. The major sources of costs were prescription medications and home care services. Low-income individuals undergoing cancer treatment spent 27% of their yearly income on out-of-pocket expenses, a demonstration of the staggering effect that cancer can have on low-income individuals and their families.

Pain is a frequent and feared symptom of patients with cancer. Optimizing pain management remains an important clinical goal. However, considering pain as a mere physical symptom ignores the social and economic effects it can have on the patient and the family. In a study of 373 outpatients with cancer, investigators estimated the direct and indirect costs of cancer pain. Sixty-nine percent of patients had a direct medical cost associated with cancer pain, be it an emergency room visit or cost of analgesic medication. These costs are absorbed by insurance after copayments and deductibles; however, patients without insurance shoulder the whole cost, which was, on average, $825 per month, or almost $10,000 per year. Excluding productivity losses, 57% reported at least one indirect cost associated with cancer pain. These expenses are borne almost entirely by the patient and family and totaled $61 per month, or more than $700 per year. The most frequently cited expense was transportation, followed by the costs of purchasing over-the-counter medication. The largest predictors of both types of costs were pain intensity, pain interference, and breakthrough pain. This finding indicates that high-quality pain management in palliative care not only frees the patient from physical ailments but also aids in the reduction of economic costs.

Although the previous study excluded labor costs, further analysis of the first wave (1993) of AHEAD data using similar methods to calculate the costs of dementia caregiving provides estimates of these numbers. Even though the informal caregiving costs for elderly patients with cancer are less than those with dementia, they are still significant at $1200 a year per patient, for a total of nearly $1 billion annually.

Illness and Bankruptcy

Filing for bankruptcy involves petitioning a federal court for protection from creditors under bankruptcy laws. It is not a step to be taken lightly, as it has negative consequences for access to credit, reputation, and employment prospects. Himmelstein and colleagues have shown the strong and increasingly powerful link between illness and bankruptcy. In their most recent study, they surveyed a random national sample of 2,314 bankruptcy filers from 2007, finding that 62.1% of bankruptcies were medical in nature. This is a nearly 50% increase in medical bankruptcy since 2001. The primary medical contributors to bankruptcy included unpaid medical bills and income loss as a result of illness. Most bankruptcy filers were middle-class, well-educated, privately-insured, and homeowners, revealing that even those who are financially well-off can be subject to the negative economic effects of illness.