Cancer Today: Origins, Prevention, and Treatment (1984)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

8 Cancer Medicine: Chemotherapy If cancer is detected early enough, when the tumor is still re- gionally localized, the patient has a nearly 50 percent chance of being cured by surgery, radiation therapy, or a combination of the two. Unfortunately, of the approximately 785,000 patients who are expected to be diagnosed with serious cancers in 1984, only 30 percent will fall into that category. For the remainder, some 550,000 patients, by the time of diagnosis the cancer will have already spread, or metastasized, to distant parts of the body, or be circulating as malignant blood cells. New advances in cancer prevention and early detection may eventually improve those fig- ures. Meanwhile, different therapeutic strategies are needed to treat those patients whose disease has spread by the time they are di- agnosed. Most of these patients will die within 5 years of their diagnosis. Once the cancer has metastasized, a surgical cure is impossible and This chapter is based on the presentation given by Emil Frei III, Dana-Farber Cancer Institute, at the 1983 annual meeting of the Institute of Medicine. 81

CANCER TODAY radiation therapy is only rarely successful. Drugs may offer the best hope, for they can pervade the body and kill disseminated cancer cells when there are still too few to be detectable. The difficulty, however, is that cancer drugs may produce ad- verse effects. Specifically, they are cytotoxic—they kill cells. The aim of chemotherapy—the use of drugs to combat disease—is to kill the cancer cells while doing minimal damage to the rest of the body. The strategy used for most cancer drugs is to target them to attack only cells that are rapidly dividing—the hallmark of a cancer cell. The antimitotic drugs, for instance, block one of the steps needed for cell division to occur. As most normal, mature cells divide slowly, if at all, they survive relatively unscathed. There are a few important exceptions, however. The cells in the bone marrow divide rapidly, as do those in the intestinal mucosa and the hair follicles. Hence, these tissues are also vulnerable to the effects of these powerful drugs. That explains the common side effects of chemotherapy—the temporary hair loss, nausea, and vomiting. Since the first of these drugs was identified, clinical researchers have been attempting to strike a balance between their antitumor activity and their toxicity. In the past 30 years, they have made impressive advances in chemotherapy. Nearly 40 cancer drugs are now in use, all effective to some extent in treating cancer. In many cases, treatment has advanced from palliative measures that brought temporary remission from the disease to outright cures. The National Cancer Institute recently estimated that some 46,000 of the 550,000 annual cancer victims whose tumors have spread elsewhere in the body can now be cured by a combination of chemotherapy and surgery and/or radiation therapy. The number may seem small, but only 30 years ago none of these patients would have had a chance of survival. For some types of cancer, the gains have been particularly dramatic. The survival rate for acute lymphocytic leukemia, a childhood cancer, has climbed from 4 to 60 percent since 1955, due largely to new chemotherapeutic regimens. For other cancers, however, few effective chemothera- peutic agents have been found. This is true for the most prevalent form of cancer, the carcinomas, or tumors that originate in the 82

CHEMOTHERAPY epithelial tissues. Surgery and radiation can provide effective treat- ment against a primary tumor, but when the cancer metastasizes, there is little hope of halting the progression of the disease. The search continues for new antitumor agents and new treatment strategies for these intractable cancers. Trial and Error The advances in chemotherapy have not come easily. The effort began in 1955, when Congress authorized the National Cancer Institute to start a Cancer Drug Development Program. From 1955 to 1975, some 40,000 drugs were selected annually for screening, largely on an empirical basis. Chemists looked everywhere—at synthetic compounds, fermentation products, and plant and ani- mal products—to find those few compounds that might be able to destroy tumor cells. By the late 1950s, this process had turned up several promising compounds. Then the testing began. In a major clinical research effort, funded and integrated by the Na- tional Cancer Chemotherapy Program, numerous investigators began trying to determine if and how these agents could be used to slow the progression of the disease or arrest it. In the early years, biochemists provided clinical investigators with agents that in the lab were capable of killing cancer cells. Yet it was at the bedside, in the clinical application, that the chemo- therapy had to be proved. Largely through trial and error, the clinical investigators learned which drugs to give in combination and at what dose—an enormously complex question. Soon after they began, the clinical researchers learned that cancer cells are notoriously recalcitrant. Indeed, the history of chemo- therapy research is replete with disappointments and false starts. Some compounds worked well against animal tumors yet proved ineffective in treating human cancers. In other cases, agents that were particularly effective in shrinking tumors had to be aban- doned because of their extreme toxicity. Later, when the clinicians thought they had developed a successful regimen and had induced a complete remission, the cancer would unexpectedly recur. As clinical researchers developed successful strategies, they be- 83

CANCER TODAY gan to gain insights into the mechanisms through which these drugs acted. That knowledge in turn provided new hypotheses for the basic and clinical researchers seeking yet more effective agents and regimens. Gradually, as the understanding of phar- macology, toxicology, immunology, and cytokinetics increased, the empiricism of the early years of chemotherapy research gave way to a more rational approach. And all the while, the survival rate for some cancers continued to climb. Childhood Leukemia Much of the early work in chemotherapy was on a childhood cancer, acute lymphocytic leukemia, as well as on some of the other leukemias and lymphomas, which are malignancies of the organs that produce the blood, such as bone marrow and lymph nodes. In the mid-1950s, clinical researchers had three compounds that they thought might be effective in treating these cancers— prednisone, methotrexate, and 6-mercaptopurine. In terms of de- ciding how to use them, however, the investigators were starting from scratch. They knew, for instance, that these three agents induced re- missions in a minority of the patients who received them. Yet they did not know what the remissions meant in the context of overall treatment strategy—that is, whether remissions increased long-term survival. They soon found that those patients who achieved complete remission did have a substantial improvement in survival over those who did not. Hence, increasing the rate of complete remission became one of the major goals of chemo- therapy. To do so, the investigators began trying various chemothera- peutic agents in different combinations. The trickiest part was to determine the optimum dose. Generally, the higher the dose of a chemotherapeutic agent, the greater the response it elicits. Yet many of these drugs have their own dose-limiting toxicity—the dose at which they become too toxic to use. Consequently, when some drugs such as methotrexate and 6-mercaptopurine were given in combination, the dose of each had to be reduced to minimize damage to the bone marrow. These reductions took their toll in 84

CHEMOTHERAPY effectiveness; often two drugs in combination yielded little increase in the remission rate over a single agent alone. Fortunately, in- vestigators found other agents such as prednisone or vincristine, a product of the periwinkle plant, which do not damage the bone marrow, and developed combined doses that in some cases were synergistic, that is, they produced a greater-than-additive increase in the remission rate. Combination chemotherapy was established. By the early 1960s, researchers had designed a combination chemotherapy regimen to treat patients with acute lymphocytic leukemia that seemed certain to turn around their prognosis. While receiving this therapy, the children showed no signs of the disease; even examinations of the bone marrow failed to turn up any tumor cells. Hopes were high. Unexpected Relapse Yet before long, all of the patients suffered a relapse. Moreover, they often relapsed while they were receiving chemotherapy. "That was our first insight into the refractoriness of the cancer cell, which we now know to be specific tumor cell resistance," says Emil Frei of the Harvard Medical School and director and physician-in-chief of the affiliated Dana-Farber Cancer Institute. He and the other clinical researchers realized that they would have to continue chemotherapy during remission, a procedure known as mainte- nance, using still other agents to combat drug resistance. In ad- dition, by adjusting the dose schedule—specifically by giving drugs intermittently rather than continuously—they were further able to reduce drug resistance and prolong remission. By 1965, the standard combination chemotherapy schedule for acute lympho- cytic leukemia entailed one drug daily and another weekly. As the investigators fine-tuned their procedures, an increasing number—90 percent—of the patients entered and remained in remission. Unfortunately, these patients began to show an in- creasing incidence of another cancer, meningeal leukemia. Sixty to 70 percent of them developed this disease, in which sheets of leukemia cells cover the meninges, the membranes that cover the brain and spinal cord. (The leukemic cells cause an increase in intercranial pressure, with associated symptoms such as headache, 85

CANCER TODAY nausea, and vomiting.) The patients with meningeal leukemia then suffered a complete relapse of the systemic cancer. Several investigators have determined how this second cancer arose. It seems that the meninges were already involved at the time the acute leukemia was diagnosed. However, the drugs used to combat the systemic cancer were largely unable to reach the central nervous system because they were blocked by what is known as the blood-brain barrier. Hence, despite a complete re- mission in the rest of the body, the leukemia cells in the spinal fluid or brain were able to proliferate and progress to a full-blown cancer. The problem was further compounded, it seems, because a minute amount of these drugs was able to cross the blood-brain barrier and pass into the central nervous system. Present in such a small amount, these drugs gave rise to a resistant cancer cell, which was then fed back into the systemic circuit, producing the total relapse. In response to this new threat, the investigators began a treat- ment known as central nervous system (CNS) prophylaxis, which consisted of the injection of chemotherapeutic drugs directly into the spinal fluid, followed by brain irradiation. First they used conventional chemotherapy to induce complete remission of the systemic leukemia, then applied CNS prophylaxis. As a result, the incidence of meningeal leukemia dropped from 60 or 70 percent to 10 percent. By 1965, the therapeutic approach was in place: combination therapy to induce complete remission, followed by maintenance chemotherapy and central nervous system prophylaxis. The sur- vival rate for acute lymphocytic leukemia climbed dramatically. In 1955, these children had no chance of survival, and the median life expectancy at time of diagnosis was one or two months. By 1965, half of them would be cured of the disease. The strategies learned in the work on acute lymphocytic leu- kemia have since been applied to patients with other cancers, such as advanced Hodgkin's disease and other types of lymphomas, again with remarkable success. In the last 10 to 15 years, the development and use of new cancer drugs has brought a cure rate of 70 to 80 percent for disseminated testicular cancer. Ten years ago, only 10 to 20 percent of these patients survived. 86

CHEMOTHERAPY The Toughest Cancers Unfortunately, the chemotherapeutic regimens that have in- creased the life expectancy for people with leukemias and lym- phomas so dramatically have been less effective against the com- mon epithelial tumors—the carcinomas of the head and neck, breast, lung, bowel—as well as melanoma, the deadly form of skin cancer. Investigators do not know exactly why these regimens are less effective against epithelial tumors. Some suspect that the different characteristics of the two types of cancer cells may be responsible. Those in the blood and lymph systems divide rapidly and are thus accessible targets for antimitotic cancer drugs. The epithelial tumors, however, are slower growing, which may en- able them to survive these drugs. Beginning in the 1970s, a group of clinical researchers worked toward a new strategy for these carcinomas, starting with breast cancer. For women whose tumor is limited to the breast, the cure rate from surgery or radiation therapy is near 80 percent. If the cancer has spread to the lymph nodes, the prognosis is quite dif- ferent. In most cases, surgery and radiation therapy can control the primary tumor in the breast, but about 70 percent of the patients relapse within one to three years because micrometastases, microscopic pieces of tumor, have been carried by the bloodstream to distant parts of the body. Chemotherapy is only marginally effective against a major, pri- mary tumor. Research in the 1960s revealed, however, that the same drugs that are ineffective against a major tumor can be re- markably effective, even curative, against the cancer in its micro- scopic form. At first, chemotherapy had been used only after a relapse, essentially as a last resort. In the late 1960s, several in- vestigators reasoned that chemotherapy might be effective if it were used immediately following radiation therapy or surgery to kill any disseminated cancer cells before they massed into a tumor. This approach is called adjuvant chemotherapy, and has been tested in clinical trials on women with advanced cancer who are at high risk of having a disseminated disease. According to Emil Frei, the results are encouraging, revealing a substantial improvement in short-term survival. Nonetheless, they are still tentative. In breast 87

CANCER TODAY cancer, unlike most other cancers, the danger of relapse continues past five years. Several more years of observation will be necessary to determine the effectiveness of adjuvant chemotherapy in treat- ing breast cancer. Head and Neck Cancer In the past five years, Emil Frei and his colleagues at Harvard have been trying to adapt the strategy of adjuvant chemotherapy to treat head and neck carcinoma. This is a particularly virulent cancer, afflicting some 37,000 people a year in the United States. By way of comparison, acute lymphocytic leukemia strikes 5,200 people a year. In nearly 60 percent of head and neck patients, the disease has progressed to the inoperable, and thus incurable, stage by the time of diagnosis, according to Frei. The time-honored approach to these patients has been palliative treatment—usually radiation but sometimes surgery—which can provide a temporary remission. Borrowing a strategy used suc- cessfully in the 1950s to treat Wilms' tumor and some other child- hood tumors, Frei and his colleagues decided to try chemotherapy as the initial treatment. Their hope was to reduce the primary tumor to an operable size or a size more amenable to radiation treatment. They call this neo-adjuvant chemotherapy, as opposed to adjuvant chemotherapy, which is used after radiation treatment or surgery. In their first study, approximately half of the patients responded well to chemotherapy, enabling the use of surgery or radiation therapy to control the disease. Frei and his colleagues have since been trying to increase the response rate to this neo-adjuvant chemotherapy. They have tried various chemotherapeutic regi- mens, using different drugs in different doses given at different times. They have finally worked out a strategy, Frei says, that elicits an 80 percent response rate: in 80 percent of the patients, the treatment shrinks the tumor to less than half of its original size. Moreover, in 30 percent of the patients, there is complete tumor regression. In the first study, there were no complete regres- sions. Although these results are encouraging, especially in these can- 88

CHEMOTHERAPY cers that are the most resistant to chemotherapy, Frei cautions that this is only the first step. It will take several years to determine whether this approach will actually increase the cure rate for head and neck carcinomas and perhaps for other types of epithelial tumors as well. The Future Empiricism played a major role in the early years of chemo- therapy. It still contributes—the recent recognition of the anti- tumor activity of platinum, for instance, was largely serendipi- tous—but to a lesser extent. Research in the past 30 years has provided a strong body of knowledge from which to draw. In addition, chemotherapy is increasingly influenced by advances in cancer biology and medicinal pharmacology. The methods of se- lecting agents for evaluation have improved, and the search is not nearly as random as it once was. Now some 15,000 compounds are screened each year, as opposed to 40,000 a decade or more ago. And for those agents that warrant further study, new model- ing systems allow better preclinical evaluation of their anticancer activity. As the recent work on head and neck cancer has shown, progress in chemotherapy has and will undoubtedly continue to come from the more imaginative and effective use of established agents. Ad- vances in molecular biology have enabled investigators to deter- mine the precise mechanisms of action of some antitumor agents. This knowledge can then be used to design more effective regi- mens. It also promises to provide insight into the differences be- tween normal and cancerous cells. Nonetheless, drug resistance and toxicity continue to plague chemotherapy. Investigations on a molecular level are beginning to reveal the biological basis of drug resistance, and new strategies are under development to combat it. In addition, investigators are learning how to modulate the activity of cancer drugs with nor- mal metabolites (substances produced by natural chemical changes in the body), a promising approach to increasing the effectiveness of a drug while reducing its toxicity. For some cancers, it is already possible to treat effectively with little or no toxicity. For example, 89

CANCER TODAY prostate cancer and some cancers of the breast are influenced by the endocrine system, and there has been some success to date in treating them with nontoxic hormones and antihormones. Biotechnology also promises to have a profound effect on the treatment of cancer. Cells produce a number of substances, ge- nerically known as biological response modifiers, to regulate their own growth and to defend themselves against disease. A number of these may also affect tumor growth. As most of these substances are proteins, they can be produced through recombinant DNA technology. Several are currently being studied. These include growth factors and antigrowth factors. Perhaps the best known of these substances is interferon, the glycoprotein that cells release when they are invaded by a virus. Interferon increases the cell's resistance; the hope is to harness it and use it to combat cancer instead of viruses. Similarly, immunologists are beginning to understand the molecular basis of the immune system. The rela- tively new area of immunotherapy seeks to manipulate the pa- tient's immune system to increase its ability to fight cancer cells. Hopes are also pinned on another offshoot of the new genetic technologies, monoclonal antibodies. These are antibodies of un- paralleled specificity, developed and cloned through immunologic techniques. It may be possible to target them to destroy specific cancer cells or carry existing cytotoxic drugs to only one type of cancer cell, leaving all other cells unharmed. On a more speculative level, the identification of oncogene prod- ucts—the proteins thought to be responsible for cancerous growth— may open up new possibilities for chemotherapy, such as agents that can destroy or interrupt the action of these proteins. Both conventional research and biotechnology promise more effective, less toxic agents to treat cancer. The medical armamen- tarium is clearly growing, giving hope that it may eventually be possible to reduce the large number of incurable cancers. 90

The Psychological and Social Effects of Cancer With the founding of the National Cancer Institute in 1937, the country embarked on a major program of cancer research. Clinical researchers developed improved radiation therapies, surgical tech- niques, and new chemotherapeutic drugs capable of arresting and even eradicating the disease in some patients. In the past 20 years, the 5-year survival rate for all cancers combined has climbed stead- ily. On another front, basic researchers have begun to unravel the biological basis of cancer—the molecular changes within a cell that give rise to malignant growth. Their work promises still more effective therapies to treat and cure more types of cancer, as well as new approaches to prevention. Now, after two decades of en- couraging advances in medical research, attention is turning to the psychological and social consequences of cancer—the toll the dis- ease exacts from the patient, the family, and the medical staff. This chapter is based on the presentation given by Jimmie C. Holland, Memorial Sloan-Kettering Cancer Center, at the 1983 annual meeting of the Institute of Medicine. 91

CANCER TODAY Throughout the 1960s, there was a growing realization among the medical community that cancer care, to be effective, must address the emotional impact of the disease as well as its physical effects. At that time, the clinicians' understanding of these prob- lems was based largely on observation and experience, not re- search. Some of the first controlled studies were undertaken in the early 1970s. Since that time, the field known as psychosocial re- search has flourished, attracting an increasing number of research- ers and embracing new concerns. The obvious starting point, when research began in the early 1970s, was to find ways to relieve the emotional turmoil of the cancer patient—the anger, fear, and the often debilitating depres- sion. Since this work began, counseling and other psychological services have become an integral part of cancer care. Yet that is just the first step, according to Jimmie C. Holland, chief of psy- chiatry at Memorial Sloan-Kettering Cancer Center. Despite the increase in supportive services, she says, many patients remain seriously troubled, and additional forms of medical intervention may be necessary. Moreover, other problems only now are being identified. In- vestigators have begun to examine the effects of cancer on the patient's family and medical staff—the depression, the chronic stress, the increased alcohol consumption. In addition, it now seems that prolonged or suppressed grief may have serious bio- logical as well as emotional consequences, leading to an increase in illness and death among those who have lost a loved one to cancer.1 And finally, there are special issues confronting the cured cancer patient. As the survival rate has increased, it has become clear that patients' troubles do not disappear along with the phys- ical symptoms. The disease and the treatment may leave emotional scars that hamper the individual's efforts to resume an active, normal life. All of these problems may be amenable to intervention of some kind—to counseling, education, or medication. Psychosocial re- 1 For more on the health effects of grief, see Committee on Health Consequences of the Stress of Bereavement, Institute of Medicine, Bereavement: Reactions, Con- sequences, and Care. Washington, D.C.: National Academy Press, 1984. 92

AND SOCIAL EFFECTS searchers are now trying to understand these and other problems confronting the cancer patient in order to develop therapeutic strat- egies for relieving the burden of this disease on patients and their families. Crisis and Resiliency Cancer remains a terrifying disease. Fear sets in at the time of diagnosis and often persists throughout the ensuing months or years. On learning they have cancer, or that the cancer has re- curred, patients experience shock and denial. They typically be- come anxious, depressed, or both. They lose their appetite and have trouble sleeping and concentrating. This "acute stress reac- tion," as it is known, is a normal, expected response to a crisis. Its severity and duration are somewhat difficult to predict, varying from patient to patient and influenced by many factors, including age, the site and course of the disease, the required treatment, and the level of support from family and friends. The reaction is usually self-limiting, subsiding to a more manageable level within a week or two. For some patients, however, this does not happen; severe depression and anxiety continue. Holland and a number of other clinical researchers from several institutions have collaborated to study this and other problems. Recently, these researchers, known as the Psychosocial Collabo- rative Oncology Group, set out to measure the prevalence of severe emotional distress among cancer patients. Their yardstick was whether the patients were sufficiently distressed to warrant di- agnosis of a psychiatric disorder. In a study of patients in three cancer centers—The Johns Hop- kins University, the University of Rochester, and Memorial Sloan- Kettering—they found that 47 percent were severely enough dis- tressed to have a recognizable psychiatric disorder. Most were psychologically stable individuals suffering from a combination of reactive depression and anxiety, a transient but severe response to crisis, Holland says. Thirteen percent of these distressed patients were diagnosed with a more severe, or major, depression. For the majority, these reactions were directly related to their illness; few had unrelated psychiatric disorders. 93

CANCER TODAY On the positive side, says Holland, slighty over half of the patients were coping well with their illness, reflecting the "re- markable courage and resiliency of emotionally healthy human beings facing a major crisis." Therapeutic Nihilism This and other studies suggest that for almost half of all cancer patients, some form of intervention to relieve their emotional dis- tress may be indicated. Many of the patients respond well to coun- seling; for others, antidepressants or other psychotropic drugs may be necessary. Yet remarkably few patients in the past have been given antidepressants, according to Holland. In a study conducted seven years ago, the Psychosocial Collaborative Oncology Group found that only 1 percent of the drugs ordered for 800 cancer patients were antidepressants. This reflects, in part, what Holland calls "therapeutic nihilism" among the medical staff: "The general feeling has been, 'Well, if you have cancer, of course you are depressed, so why bother to treat it?' " In addition, she says, physicians tend to avoid prescribing other psychotropic drugs and pain killers for fear that they will be abused. Holland has found, however, that cancer patients are "highly responsible" in their use of drugs. Another factor in the negligible use of antidepressants is that they have generally been considered ineffective in treating depres- sion related to physical illness. Holland and her colleagues have evidence to the contrary. They have found that certain tricyclic antidepressants, such as nortriptyline or desipramine, are partic- ularly effective in treating depressed, withdrawn patients. For pa- tients who have a mixture of anxiety and depression, the research- ers have had success with another group of tricyclics, amitriptylin and imipramine, as these drugs have sedative and analgesic effects. Given these encouraging results, Holland suggests that controlled clinical trials be conducted to determine the effectiveness of these and other drugs in treating emotional disturbances related to can- cer. Different forms of intervention may be appropriate for other 94

PSYCHOLOG/CML AND SOCML EFFECTS psychological problems. In some patients, for instance, part of the severe physical reaction to chemotherapy may be emotionally me- diated. Chemotherapeutic agents can have extremely unpleasant side effects, including fatigue, fever, chills, and nausea. In addition to these effects, nearly one-third of patients receiving chemother- apy develop anticipatory vomiting, a conditioned, Pavlovian re- sponse: Just thinking about chemotherapy, or smelling the alcohol as they enter the hospital, is enough to start these people vomiting. For some patients, the anticipatory reaction is so severe that chemotherapy has to be discontinued. Antinausea drugs and coun- seling help, although in a recent study, the best results were achieved from a behavior modification technique known as desensitization, in which a person is exposed in small, cumulative steps to the feared object. This, too, warrants clinical trials, says Holland. Chronic Stress When the acute stress accompanying diagnosis subsides, it is often replaced by lower-level, chronic stress that lasts throughout the course of treatment. Patients are slightly more depressed than normal, slightly more anxious. They are often prey to sudden swings of emotion, alternating between moments of hope and despair. The physical manifestations are insomnia, fatigue, and loss of appetite, although it is often difficult to determine if these are stress- or disease-related. The family is also subject to chronic stress. They, too, feel anger, fear, grief, and sometimes guilt over the plight of a loved one. And as they are coping with the illness, they often confront added financial burdens or perhaps greater responsibilities, such as child care. Moreover, their hardships often go unrecognized, as most of the attention and sympathy is directed toward the sick member of the family. For both patient and family, individual or group therapy can help. In many parts of the country, there are special cancer groups in which patients can discuss their problems and gain insight, advice, or just sympathy from others going through the same ordeal. 95

CANCER TODAY The Physician Most of these studies have focused on the patient, and to some extent, the family. Recently, psychosocial researchers have begun to consider the physician and other medical staff as well. The few studies that have been conducted to date suggest that oncologists (the physicians who treat cancer patients) may also experience chronic stress, Holland says, probably more severely than phy- sicians in other specialities. They may describe the problem as "burnout," but Holland suspects that it is a reflection of the special emotional burden of cancer care, not just the normal, and by no means insubstantial, stresses a physician encounters. A diagnosis of cancer is devastating to the patient. It can also be distressing to the physician, who often must convey this news several times a day. During the course of treatment, the oncologist must often give additional bad news and observe the suffering of the patient and the family. There are problem patients who have unrealistic demands, there are the tragic childhood cancers, and above all, there is the constant stream of deaths. Inevitably, some patients come to mean more to the physician than others, and the impact of their death is greater still. These stresses may surface several ways. Physicians may become overinvolved, working long hours past the point of peak ability. They may become tense, irritable, and overly critical of the work of others. Or they may become uninvolved, detached from work, patients, and family. They may also turn to alcohol or drugs. According to Holland, easy access to prescription drugs is an occupational hazard for physicians. A recent study by Steven Schreiber and Brian Doyle found that alcoholism is a problem for 8 percent of physicians; drug abuse, for 2 to 3 percent. Other figures are more disturbing: some 100 physicians, the equivalent of one average medical school class, commit suicide each year. Indeed, suicide is the major cause of death among physicians. It is not clear whether these problems are more pronounced among cancer specialists than among phy- sicians as a whole. If so, special education and intervention strat- egies might be useful. 96

PSYCHOLOGICAL AND SOCIAL EFFECTS Cancer Survivors Twenty years ago, the survival rate for all cancers was 25 per- cent. This year, nearly half of the individuals who will be diag- nosed as having cancer will be cured. "Cured" is a difficult concept in cancer care: some tumors can recur after long periods of re- mission. It is hard to say with accuracy when an individual has been cured. Nonetheless, for most types of cancer, 5-year survival represents a cure. By this definition, there are now some 3 million Americans alive who have survived cancer, and there are a total of 5 million alive who have a history of cancer but who have not yet had five disease-free years. These numbers reflect one of the great successes of contemporary medicine—of diagnostic techniques, surgery, radiation therapy, and chemotherapy. Nonetheless, these patients' troubles do not end when they leave the hospital or when they cross the 5-year line. They are con- fronted by many medical, psychological, and social problems dur- ing the transition back to an active, healthy life. Psychosocial researchers are now beginning to study the special problems of the cancer survivor. One of the first priorities, ac- cording to Holland, is to determine the delayed and long-term effects of both the disease and the treatment. For instance, a small percentage of patients develop secondary malignancies, a distress- ing consequence of radiation and chemotherapy. Research is needed to determine the prevalence of these secondary malignancies, how they arise, and specifically, how they can be prevented. Both radiation and chemotherapy are relatively new treatments. Although their long-term effects on the body are generally not known, it is clear that both may damage the reproductive cells, leading in some patients to infertility. This is of particular concern, as many of the successful cancer cures have been for childhood and adolescent cancers. As these patients enter their childbearing years, they want to know about possible reproductive damage or risk. Few data are available at this time, although several studies are under way. Similarly, there is evidence of central nervous system damage 97

CANCER TODAY following some cancer treatments. In an effort to prevent a relapse in the central nervous system, many acute lympocytic cancer pa- tients have been given cranial radiation and chemotherapy (see Chapter 8). Julia Rowland and her colleagues in the Cancer and Leukemia Group B in Scarsdale, New York, have recently found that the average IQ of children who received this treatment was 10 points lower than in children who received other treatments, but not cranial irradiation. Of course, damage must be weighed against the benefits of this treatment. In response to such studies of long-term effects, treatment programs can be altered to reduce undesirable consequences. For instance, in most cases today, the radiation dose used in therapy has been reduced, except for children at extremely high risk of a relapse in the central nervous system. Fear of Recurrence Many psychological problems relate to the patients' uncertainty about their physical condition and about the future. One of the most prevalent problems among cancer survivors, for instance, is fear of recurrence. This fear is deeply rooted in reality; physicians can provide no guarantee that all the cancer cells have been erad- icated. Consequently, cancer patients can never be sure when they have crossed the line from illness to health. Although the possibility of recurrence is always present, in some patients fear becomes excessive and debilitating. Some individuals feel extremely vulnerable, as if a sword is hanging over them, ready to drop at any time. These excessive fears are known as the Damocles syndrome. Fear of recurrence can lead to a debilitating preoccupation with the disease. Every cough, every ache is thought to signal that the cancer has returned. Some former cancer patients repeatedly seek medical reassurance. Others go to the opposite extreme, avoiding physicians entirely and neglecting essential health monitoring and maintenance programs. Perhaps the best way to combat this fear is for physicians to provide solid information about the patient's chances of recurrence or secondary malignancies, according to Fitzhugh Mullan, who is both a physician and a cancer survivor.2 It would also help, he 2 Fitzhugh Mullan has written a book about his experiences: Vital Signs: A Young Doctor's Struggle with Cancer, New York: Farrar, Straus, Giroux, 1982. 98

PSYCHOLOGICAL AND SOCIAL EFFECTS says, to warn recovering cancer patients that they will feel these fears and that they are an expected part of the process of emotional healing. Living With Compromise Cancer often leaves a physical mark. Some patients lose a limb, a breast, a colon. Other disabilities, such as weakened lung capacity or infertility, are less visible. Adjusting to this physical compro- mise can be difficult. According to Mullan, the problem can be particularly severe for younger patients who are attempting to return to active lives that challenge their strength, endurance, and sexuality. Again, education and counseling can help. Indeed, as early as the 1940s, the American Cancer Society sponsored a program in which patients who had gone through particularly radical surgery, such as the removal of a colon or larynx, counseled those who had recently been diagnosed as having cancers of those organs. Since that time, such programs have expanded considerably. Some of the best known are Reach for Recovery and Lost Chord. For some patients, there is also the possibility of physical re- habilitation, such as breast reconstruction. In a study conducted with her colleagues at Sloan-Kettering, Holland has found that breast reconstruction can significantly reduce the emotional dis- tress resulting from mastectomy and may help to overcome the reluctance of some women to have this life-saving but emotionally distressing operation. They found that the women who opted for breast reconstruction were psychologically healthy and well ad- justed and had reasonable expectations about the surgery, realizing that it will not produce a truly normal-looking breast. Early con- cerns were that women who sought reconstruction would un- realistically believe that the reconstruction would solve other life problems, as is sometimes the case with other forms of plastic surgery. After surgery, most of the women—85 percent—were happy with the reconstruction even when the results was consid- ered by the surgeon to be less than desirable. Perhaps most im- portant, they were more satisfied with their sexual lives, had higher self-esteem, and took more pleasure in their appearance following breast reconstruction. 99

CANCER TODAY The Transition to Health For the cured cancer patient there are social issues as well. As people make the transition from sick to healthy status, they have to confront the perceptions and reactions of friends, family, em- ployers, and others. Families typically respond by overprotection; friends may tend to see the patient as someone special, either a hero or a victim. This "Lazarus syndrome" hampers a former patient's efforts to take up where he or she left off. Cancer survivors may also be shunned by those people who are afraid or uncomfortable to be with someone who is or has been seriously ill. This reaction also surfaces frequently as job or in- surance discrimination, according to Holland. Some employers are reluctant to hire someone with a history of cancer—even after a childhood cure. Some insurance companies charge former cancer patients a higher premium or reduce their coverage. One unfor- tunate consequence of these two forms of discrimination is that the individual becomes reluctant to switch jobs, and personal and career growth may suffer. Bereavement Despite the remarkable progress in cancer care in the past few years, slightly more than half of all cancer patients succumb to the disease. They leave behind family members who appear to be more prone to both physical and psychological illnesses. Some preliminary studies on the effects of grief have shown that both alcohol and cigarette consumption is elevated in families who have just lost a member to cancer, as is the suicide rate. A recent study by Knud Helsing and Moyses Szklo of The Johns Hopkins Uni- versity found that widowers between the ages of 55 and 74 showed increased mortality rates from a range of illnesses, especially during the first six months after their wife's death. This increase in mor- tality rate was not evident in women of the same age group who had lost their husbands, however. Whether the increase in male mortality is related to a breakdown in normal health practices or occurs by some as yet unknown cellular route is not clear. "Re- 100

PSYCHOLOGICAL AND SOCIAL EFFECTS search in this basic and common emotion—grief—is remarkably absent," Holland says. Similarly, little is known about the health effects on the oncology staff who experience repeated losses. Nonetheless, Holland and others believe that sufficient evidence exists to extend the concept of cancer care to the family and staff, as well as to the patient. 101

10 Alternative Care for the Dying: American Hospices In the early 1960s, Cicely Saunders spent long hours with a friend who was dying of cancer in a busy London hospital. They talked often about the type of place where he would rather be during the last weeks of his life. The hospital, geared to treat acute diseases of the body, seemed cold, impersonal, and poorly equipped to handle the more chronic problems confronting a terminal cancer patient. He and Saunders envisioned a different kind of place, a special center for treatment of dying patients that would tend to the emotional and spiritual needs of the patients as well as the physical; that would relieve their pain, assuage their fears, and help them come to terms with dying. In his will, he left Saunders, who was then a medical social worker, £500 to start a home for the dying. She went on to get her medical degree, then worked for several years in an existing This chapter is based on the presentation given by David S. Greer, Brown University School of Medicine, at the 1983 annual meeting of the Institute of Medicine. 103

CANCER TODAY center for the terminally ill, called a hospice after the church- sponsored institutions of the Middle Ages that provided for the dispossessed and disabled. Finally, in the late 1960s, with funding from the National Health Service, Saunders opened her own center in London, St. Christopher's Hospice, and set about changing the nature of medical care for the dying. Relieving Fear and Pain She calls her center a home, not a hospital, and has described it as "high people, low technology." To counter loneliness, several patients share a room, which is typically decorated with family mementos and personal belongings. The hospice is open all day to visitors, and family and friends are usually in abundance. The intravenous drips, diagnostic machines, and monitoring equip- ment typical of a hospital ward are missing. Dying is confronted openly. The hospice does not try to post- pone the inevitable by heroic, and often traumatic, life-saving measures such as surgery or chemotherapy. Patients can be trans- ferred to a hospital for such treatment at any time, however. Instead, the hospice staff, which includes physicians, social work- ers, psychiatrists, and a large contingent of volunteers, provides emotional support and concentrates on minimizing pain as the disease takes it course. Indeed, pain relief is the hallmark of Saunders' approach to care for the terminally ill. She objects to the practice of giving pain medication only as needed, the standard hospital procedure that arose from fear of fostering drug dependency. By contrast, Saun- ders' approach is to try to prevent the patient from experiencing pain. In some cases, that means providing regular doses of the Brompton cocktail, a mixture of heroin, cocaine, gin, sugar syrup, and chlorpromazine syrup. Despite the often high doses, patients remain alert, she reports, able to face their last days without fear or debilitating pain. American Hospices Saunders' ideas spread quickly through England and Europe. The hospice movement came to represent both a philosophy and 104

ALTERNATIVE CARE FOR THE DYING a system of care for the terminally ill. It reached this country in 1974, with the founding of the first hospice in Connecticut. The movement struck a particularly responsive chord in the United States; many health professionals welcomed hospices as an alter- native to what they saw as the impersonality and overemphasis on technology of the American terminal care system. In the following 10 years, hundreds of hospices were started in the United States; now some 1,000 institutions offer hospice care. Perhaps even more significant as a measure of their acceptance in the American health care system is the legislation passed by Con- gress in 1982, as part of the Tax Equity and Fiscal Responsibility Act (TEFRA), that provides Medicare coverage for hospice serv- ices. It is the hospice philosophy, more precisely, that has taken hold; the exact implementation varies from country to country, and sometimes from hospice to hospice. In the United States, hospices have a decidedly different orientation from their counterparts in Europe and Canada. In most U.S. hospices, the concept of a homelike environment has been expanded to an emphasis on treat- ing the patient at home whenever possible. By contrast, English hospices are principally inpatient facilities. Even within the United States, hospices vary one from another in many ways. Some are affiliated with hospitals, some with home health care agencies such as visiting nursing programs; some with churches. Some are small, community-based programs, treating only a handful of patients, others are associated with large agencies that may treat hundreds. Those affiliated with hospitals often have inpatient facilities within the hospital, although they are usually decorated with homelike furnishings and are run by a separate hospice staff. Some of the freestanding hospices, unaffiliated with hospitals, also have their own inpatient facilities. Most do not, and instead are designed exclusively to provide care in the home, although they can arrange to provide hospital care if needed. Sim- ilarly, most hospices, including those with inpatient facilities, pro- vide home care for those who wish to spend some time at home, or die there. Underlying these differences in shape and size, however, is a common approach to care for the dying. As in England, U.S. hospices are used almost exclusively by terminal cancer patients; 105

CANCER TODAY only about 10 percent of hospice patients are dying from other diseases. (One of the qualifications for Medicare reimbursement under the new law is a prognosis of less than six months to live.) Care is planned and provided by a medically supervised team that consists of several types of professionals and paraprofessionals, including nurses, social workers, and volunteers. In all hospices, the family or close friends also play a major role in patient care. The hospice staff teaches the family how to care for the patient at home, and both family and patient are educated on death and dying. Support services are available to the family as well as to the patient, and staff members are available after the patient dies to help the family with their grief. The prevention and alleviation of pain is a major goal of Amer- ican hospice care, as it is in Saunders' St. Christopher's Hospice. Heroin is not used, as it is not approved for medical use in the United States. Instead, other narcotics and analgesics are admin- istered to combat the persistent pain that afflicts nearly 50 percent of all cancer patients as they near death. Assessing Hospice Care In the past decade, many claims have been made about hospice care. Hospices promise better relief of pain than can be achieved by conventional cancer treatment, more emotional comfort and relief for patients and their families, and, given the reliance on home care and the minimal use of medical intervention, lower costs. Only recently, however, has there been any attempt to evaluate these claims scientifically; to determine whether hospices indeed offer superior care for terminally ill patients. As momentum gath- ered behind the hospice movement, Congress in 1979 mandated a major study of hospice care. The National Hospice Study, com- pleted in 1984, was directed by David S. Greer, dean of medicine and professor of community health at Brown University, and his colleagues at Brown. It was supported by the federal Health Care Financing Administration, as well as by the Robert Wood Johnson Foundation and the John A. Hartford Foundation. For one-and-a-half years, the investigators tracked some 2,000 106

ALTERNATIVE CARE FOR THE DYING patients and their families as they moved through hospices and conventional care facilities, and analyzed data on over 13,000 pa- tients. Specifically, the investigators tried to assess how hospice care actually differs from conventional care, what effect it has on the patient and family's quality of life, and its costs relative to conventional care. The study also examined the possible effects of Medicare reimbursement on the structure, staffing, and cost of hospices. Two Types of Hospices Although American hospices assume many forms, the investi- gators found that there are essentially two types: those with in- patient beds and those without. In the study, those with beds were called hospital-based hospices, regardless of whether the inpatient facility was in a hospital or not. Those without beds were called home-care hospices. The researchers studied a total of 40 hospices, comparing one type to another, and both types to 14 facilities offering conventional cancer care. Ninety-three percent of the patients in the hospices had been diagnosed as having terminal cancer, and most of them died while in the hospice program. Over half of the patients spent a month or less in the hospice program; about 20 percent spent less than one week. Eight percent, however, remained in the hospice for over six months, some for as long as two years. The investigators found a number of differences in the pattern of care patients receive in the two types of hospices. The most obvious difference was the site of care. As might be expected, patients in home-care hospices spent most of their time within their own homes. On average, home-care hospice patients spent only 5.2 days in an inpatient setting, usually in an acute care hospital. By contrast, hospital-based hospice patients averaged 18.2 days in an inpatient facility. Examined in another way, over 99 percent of the home-care hospice patients spent some of their time in the home, while that was true of only 67 percent of the hosoital- based hospice patients; that is, one-third of all hospital-based hos- pice patients spent all their time in an inpatient setting. This trend was also reflected in the site of death. Sixty-two percent of those 107

CANCER TODAY patients in home-care hospices died in their homes. By contrast, slightly more than 20 percent of the patients in hospital-based hospices died at home. These patterns have obvious cost impli- cations, which will be discussed below. Greer and his colleagues also discerned certain differences among the patients who entered each type of program. Those entering hospital-based hospices tended to have weaker family ties—for instance, they were more likely to live alone—and thus might have fewer relatives or friends able to provide home care. In ad- dition, they tended to be more functionally impaired at the time of admission than are those entering a home-care hospice program. Nonetheless, Greer says, these differences are not sufficient to explain the different pattern of care the patients receive in the two types of programs. He concludes that care is determined largely by the structure of the hospice. "When they have beds, they tend to fill them," he says. Medical Intervention Both home-care and hospital-based hospices espouse a philos- ophy of palliative care rather than curative treatment. The study revealed that, indeed, during the last two to five weeks of life, hospice patients were significantly less likely to receive any kind of intensive medical intervention—such as radiation, chemother- apy, or surgery—than were patients in conventional care facilities. Hospice patients were also less likely to receive blood tests, X rays, and other diagnostic tests, regardless of the type of hospice (see Figures 10-1 and 10-2). The investigators also looked at other clinically relevant aspects of care, such as the use of radiation therapy to reduce bone pain, or blood transfusions and other intravenous therapy to treat weight loss or hemorrhage. Patients in both conventional care facilities and hospices were equally likely to receive this type of radiation therapy. There was a major difference in the use of intravenous treatment, however. At all times, patients were more likely to receive intravenous therapy in a conventional care setting than in a hospice. In addition, that difference became more pronounced as the patient neared death: in a conventional care setting the use 108

ALTERNATIVE CARE FOR THE DYING HOSPICE HOME-CARE HOSPICE HOSPITAL-BASED CONVENTIONAL CARE 13.9% Includes: Surgery Chemotherapy Radiation Therapy 37.7% 20 30 PERCENT FIGURE 10-1 Patients receiving "intensive" interventions as last measure prior to death. (SOURCE: National Hospice Study, Preliminary Final Report, 1983.) HOSPICE HOME-CARE HOSPICE HOSPITAL-BASED CONVENTIONAL CARE 37.7% 135.7% 25 60 PERCENT 72.4% 75 FIGURE 10-2 Patients receiving diagnostic blood tests, X rays, or scans as last measure prior to death. (SOURCE: National Hospice Study, Preliminary Final Report, 1983.) 109

CANCER TODAY of intravenous therapy increased, while in the hospices it de- creased. By contrast, the percentage of patients receiving psychological and social services was greater in the hospices than it was in con- ventional care, reflecting the different allocation of resources in the hospice and the hospital. The same difference existed even prior to hospice admission, however, and may partially reflect the counseling that cancer patients received during the process of ap- plying for hospice care. Pain About 30 percent of the patients in the study were free of pain during the last weeks of their lives. For the remainder, narcotics and analgesics were used in various combinations to provide relief. How well this drug therapy worked is difficult to assess. For many patients, the pain during the last few weeks was severe; many could not be interviewed during this time. Instead, the investi- gators relied on the judgment of the principal provider of care for each patient, for example, a family member. Their reports indi- cated that pain control may be better in the hospital-based hospices than it is in either the home-care hospices or in conventional set- tings (see Figure 10-3). Given the hospice's objective of preventing pain, it was not surprising to find that pain control was better there than in a conventional hospital ward. A comparative analysis of the use of narcotics and analgesics confirmed that hospitals provided drugs largely on an as-needed basis when the patient is in pain, according to Greer. By contrast, a higher percentage of patients in the hos- pital-based hospice were on around-the-clock narcotics. What was unexpected was the apparent difference in the degree of pain con- trol in the two types of hospices. According to Greer, a smaller amount of pain medication was used in the home. He suspects that families are reluctant to administer narcotics. As a result, home-care hospice patients received fewer and smaller doses of pain medication than did their counterparts in the hospital-based hospice. Interestingly, the level of patient awareness was similar in the 110

ALTERNATIVE CARE FOR THE DYING HIGHEST PAIN 4.0 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 FREE FROM PAIN 0.0 — Home-Care Hospice — — — — Hospital-Based Hospice —-—— Conventional Care WEEKS PRIOR TO DEATH FIGURE 10-3 Mean score on pain index for "average" hospice patient as death approached. (SOURCE: National Hospice Study, Preliminary Final Report, 1983.) hospices and the conventional facilities during the last five weeks of life (see Figure 10-4). In all three systems, awareness declined at roughly the same time, apparently regardless of the particular pain therapy received. This suggests that, contrary to many fears, pain medication can be used liberally without impairing functional abilities. Quality of Life Hospices strive to make the patient as content and satisfied as possible during the last weeks or months of life. To evaluate overall quality of life in the hospices and conventional care settings, the National Hospice Study investigators used several different indices that measured, among others, perceived health, outlook, family support, functional performance, loneliness, depression, emo- tional comfort, and patient awareness. Contrary to some of the anecdotal claims about hospices, they found that the quality of life 111

CANCER TODAY MOST AWARE 1.0 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 LEAST AWARE 4.0' ^—^— Home-Care Hospice ———— Hospital-Based Hospice ^- Conventional Care WEEKS PRIOR TO DEATH FIGURE 10-4 Mean score on awareness for "average" hospice patient as death approached. (SOURCE: National Hospice Study, Preliminary Final Report, 1983.) was uniformly high, with no significant difference in quality of life across the three systems (see Figure 10-5). They looked in particular at social involvement—contact with friends, family, and staff—because it is an important component of quality of life. Again, patients in all three systems reported a high level of social involvement. While patients in home-care hos- pices actually received more hours of instrumental care and visits, the reports of the primary care persons suggested that patients in the conventional setting had a slightly higher social quality of life than did hospice patients. Similarly, although treatment varied significantly in the hospices and the conventional care facilities, all patients interviewed re- ported a high level of satisfaction with their care. Again, many could not be interviewed in the terminal period. The uniform 112

ALTERNATIVE CARE FOR THE DYING satisfaction, regardless of which type of system the patient is en- rolled in, may reflect preselection, Greer speculates: in other words, those patients who are inclined to want more intensive medical treatment probably stay in the conventional care system. In addition, satisfaction with care was also high among the families of the patients, although it was significantly higher among the families of hospital-based hospice patients. The data indicate that home care is a significant emotional burden on the families. For the families of the patients in home-care hospices, the number of hours each member devoted to caring for the patient was stag- gering, Greer reports, and it appeared to take a toll on their well- being. There was a rise in illness among family members; in the first few months after the patient's death, the family made twice the average number of doctor visits. Nonetheless, despite the signs of strain, these families did not seem to regret their decision; they BEST 10 r 6 —- Home-Care Hospice -—— Hospital-Based Hospice -^— Conventional Care WORST 0 5 3 WEEKS PRIOR TO DEATH FIGURE 10-5 Mean score on quality of life index for "average" hospice patient as death approached. (SOURCE: National Hospice Study, Preliminary Final Report, 1983.) 113

CANCER TODAY derived great satisfaction from their ability to fulfill the wish of their loved one to die at home. Costs The recent TEFRA legislation allowing Medicare reimburse- ment of hospice care was predicated on the assumption that it is less costly than conventional care. The National Hospice Study investigators found that such comparisons are not a simple matter: patterns of utilization differ considerably between hospitals and hospices and between types of hospices. For example, the average length of time spent in a home-care hospice setting was longer than the average stay in a hospital- based hospice. Further, when home-care hospice patients were hospitalized, the hospice had less control over the ancillary medical services. Consequently, the cost of inpatient care for home-care hospice patients was higher than it was for hospital-based hospice patients. For home-care hospices, the inpatient rates were $278 a day, for hospital-based hospices, $218 a day. In total, however, because of their greater reliance on inpatient care, hospital-based hospices were more expensive than home- care hospices. The hospital-based hospice cost an average of $5,980 per patient; the home-care hospice, an average $4,758 (in 1982 dollars). The major question, says Greer, is whether hospices offer a savings over conventional care. The study shows that hospices do provide significant savings, at least for typical stays of less than one month. For the last week of life, for instance, the cost of a home-care hospice was $910 lower, and a hospital-based hospice was $657 lower than conventional care. As the length of stay increased, home-care hospices continued to be less costly than conventional care, but hospital-based hospices became signifi- cantly more expensive than conventional care for stays above two months (see Figure 10-6). As mentioned earlier, the majority of patients spent fewer than two months in a hospice.1 1 All dollar figures represent findings of the first year of the study and have been published in the American Journal of Public Health. However, the final figures vary from these no more than plus or minus five dollars. 114

ALTERNATIVE CARE FOR THE DYING $3000 W z $2000 CO $1000 0 $1000 $2000 I- M 8 0 $3000 $4000 $5000 Home-Care Hospice Hospital-Based Hospice 210 Days 6 Months 3 Months 2 Months 1 2 Month Weeks 1 Week HOSPICE LENGTH OF STAY FIGURE 10-6 Cumulative Medicare savings/costs of hospice relative to com- parable conventional care. (SOURCE: National Hospice Study, Preliminary Fi- nal Report, 1983.) Something Intangible Hospices do offer a different type of care for terminal patients. They are less likely to use intensive or heroic medical techniques, and are more likely to provide social services. There is some evi- dence that pain control is better in the hospital-based hospice than in conventional care setting, although such information is difficult to obtain from critically ill patients. However, despite claims of hospice advocates, neither type of hospice offers an advantage in terms of the patient's quality of life. It appears that the quality of life in a conventional care setting is not as bad as has often been portrayed, Greer says. Perhaps this reflects a recent and widespread shift towards a more humanitarian, less technologically intensive treatment of the terminally ill, he speculates. For stays of one month or less, the hospice system can offer significant economic savings. Home-care hospices are less expen- sive than conventional care, even for longer stays. For that reason, the government might want to provide incentives to encourage the use of home-care hospices, Greer says. 115

CANCER TODAY He believes that both types of hospices are valuable alternatives to the conventional care system. Certain patients may be better suited to each type. For instance, patients with a high level of pain or weak family support may be better served in a hospital-based hospice. Moreover, there may be other, less tangible benefits that were not revealed by the study, says Greer, who practiced community medicine 17 years before joining the university. During that time, he learned that many of the decisions made in medical care are based on "experience, impressions, and intuition," not hard data. "Our study team spent many hours developing methods of measurement and inquiry designed to uncover all the salient ele- ments of the hospice experience. Much of this agonizing was done in collaboration with hospice representatives. Yet as a man who has visited hospices across the length and breadth of the nations, I can say that there is something going on in them, something emotionally and spiritually inspiring and invigorating, that is not measurable using current techniques and that cannot even be de- scribed, except perhaps by the best of our poets." Medicare reimbursement will make hospice care available to a wider number of people. Yet government participation, which means new federal regulations and restrictions, will also increase administrative costs and the complexity of management, perhaps driving out the volunteer-dominated community hospices. "Their loss would have a great impact on the hospice movement, both philosophically and programmatically," Greer says. Restrictions in Medicare reimbursement may prompt hospices to change their structure or policies, for instance, by increasing the numbers of short-stay patients. In the face of increasing regulation, Greer says, the overriding challenge to the hospice system will be to maintain those elusive elements that make hospices unique. 116