likely that one single database can meet all of the various objectives of such systems, for example, cancer surveillance, research, and quality monitoring. Data systems need to be monitored to assure accuracy, and should be automated to improve the timeliness of quality data. Data gathered into national databases, in particular, should be made available quickly for analysis by investigators and evaluators.
RECOMMENDATION 8: Public and private sponsors of cancer care research should support national studies of recently diagnosed individuals with cancer, using information sources with sufficient detail to assess patterns of cancer care and factors associated with the receipt of good care. Research sponsors should also support training for cancer care providers interested in health services research.
Grants to support the analysis of data that focus on pressing health policy questions, especially about how the organization and financing of cancer care affect the processes and outcomes of care, should be a high priority. Methodologic research is also needed to improve the quality of cancer-related health services research, for example, to develop tools for "case-mix" adjustments to reduce the potential for bias inherent in observational cancer research.
An annual report that provides a description of the status of cancer-related quality-of-care research, and summarizes relevant published literature in the area would serve as a valuable resource for health services researchers and those involved in quality assessment. Such a report would also help organizations set priorities for research, ensure that their research portfolios address important quality-of-care questions, and ensure that their research programs are complementary and coordinated.
RECOMMENDATION 9: Services for the un-and underinsured should be enhanced to ensure entry to, and equitable treatment within, the cancer care system.
Cancer is among the most expensive conditions to treat, and individuals with cancer and their families invariably bear some of the financial burden. Most individuals diagnosed with cancer are elderly and have Medicare coverage, but an estimated 7 percent of individuals facing a new diagnosis of cancer lack any health insurance at all. For these individuals, cancer can be catastrophic to their finances as well as their health. The problem that affects far more individuals, however, is underinsurance—health plans and insurance coverage offer some, but often incomplete, protection against the high costs of cancer care. High deductibles, copayments or coinsurance, and coverage caps can all contribute to high out-of-pocket expenditures. Medicare, for example, was estimated to cover only 83 percent of typical total charges for lung cancer and 65 percent of charges for breast cancer in 1986. Some individuals have additional protection through other insurers (e.g., Medigap policies or Medicaid), but despite this, the financial burden of cancer can be substantial even among those covered by a health plan. Limits on prescription drag coverage, an expensive and widely used benefit (e.g., outpatient pain medications), are a particular problem for