The original hospice Medicare benefit created a number of problems that impact on the appropriate use of palliative irradiation, particularly in the hospice setting. The language assumes incorrectly that there is a well defined "terminal" period that is easy to define and lasts exactly six months. The reimbursement also suggests a simple definition of terminal care that does not include the use of "aggressive therapy" such as radiation therapy and therefore does not reimburse for it under the Medicare benefit. This creates a tremendous financial burden on the hospice program.
     The proper definition of the terminal phase would be a well defined period in a disease marked by a rapid physical deterioration leading to death. In cancer patients this period is only two to three months long and is marked by a dramatic exponential decline in performance score and rapidly increasing need for medical care (figure 1.) The proper use of the word "palliative care" would have to include much more than care for incurable patients. There are at least three levels of care beyond cure (table 1) and the category palliative symptomatic would appropriately include radiation therapy and might also include hospice care if the patient would benefit from entrance into a hospice program. The original National Hospice Study demonstrated a lower use of radiation therapy(figure 2)  but was not designed to determine if such care was being given inappropriately. A recent study suggested that radiation is not being use enough in the palliative setting (JAMA 1998;279:343.)

Making accurate survival estimates

     There is a large literature that suggests that health care workers are very poor at making accurate survival estimates. Since the hospice benefit under Medicare expects physicians to be able to identify patients who have only six months to live, some guidelines are appropriate to avoid accusations of 'Medicare fraud' if a patient 'lives too long.' Survival estimates are particularly difficult for nonmalignant patients but there is considerable data about cancer patients. Survival is   particularly impacted by performance or Karnofsky score, but also include extent of disease (minimal liver involvement versus advanced), source of the primary (lung to bone   worse than breast or prostate), and response to therapy. Since response to therapy cannot be assessed until a decision about treatment has been made, these survival estimates are obviously only that, crude estimates. Nevertheless some estimate of survival expectations needs to be made to help gauge the benefits of a short course of therapy (versus prolonged or no therapy at all.) Some representative data is shown in table 2a-f.

Specific sites and the value of palliative irradiation.

     Brain. Brain metastases are common in advanced cancer patients, particularly lung cancer. Survival is often very short with no treatment (1 month) and prolonged by palliative radiation therapy. Brain metastases may also be successful retreated. This data is presented in tables 3.
     Bone. The pain in bone metastases include that produced by periosteal stretching and compression or infiltration of nerve roots. Radiation is often effective in relieving pain (70 - 80% of the time) with a modest objective response (25-30%.) Patients with widespread lesions may be better treated with large fields (hemibody) or with radioisotopes (such as strontium-89.) Patients with short survival expectations can be well treated with single shots of 600 or 800cGy. This data is presented in tables 4..
     Neurologic. Spinal cord compression syndrome is a common and catastrophic complication in patients with metastatic cancer. Patients respond well do irradiation if treated early. The importance of this problem relates to the benefits of early diagnosis (tables 5.)
     Visceral Sites. Patients with advanced   disease in other sites commonly will benefit from short courses of palliative irradiation and should be considered for such therapy. Many patients can be treated with short courses (370cGy bid X 4 over two days) with excellent palliation, and patients may also respond well to retreatment (table 6).

Summary and Conclusion

     Palliative irradiation is useful in multiple sites of the body and often provides a high degree of durable palliation with minimal morbidity. The appropriate use requires an understanding of the natural history of the disease, an accurate diagnosis of the etiology of the patients pain, and understand and participation of an educated patient/family in the decision process. Newer forms of irradiation such a high dose rate catheters (HDR table 7a.b) may play an increasing role in palliation. Many patients may benefit from abbreviated course or retreatment. A bibliography of selected abstracts emphasizes a number of these concepts (table 8).