Some prior research has suggested that health spending for many diseases has been driven more by increases in so-called treated prevalence-the number of people receiving treatment for a given condition-than by increases in cost per case. Our study reached a different conclusion. We examined treated prevalence, clinical prevalence-the number of people with a given disease, treated or notand cost per case across all medical conditions between 1996 and 2006. Over this period, three-fourths of the increase in real per capita health spending was attributable to growth in cost per case, while treated prevalence accounted for about one-fourth of spending growth. Our evidence suggests that most of the treated-prevalence effect is due to an increase in the share of eligible people being treated rather than an increase in clinical prevalence of diseases. We conclude that efforts to curb health spending should focus more on reining in cost per case. C oncern about rising US health spending has existed for many years and has spawned substantial research into the factors that have caused it. 1 A 2008 report by the nonpartisan Congressional Budget Office identified seven key factors driving the historical growth of health care spending: aging of the population; changes in third-party payment; personal income growth; health sector prices; administrative costs; defensive medicine and supplier-induced demand; and technology-related changes in medical practice.2 Researchers have identified the latter as a leading factor, responsible for anywhere from one-third to two-thirds of the growth in real per capita health care spending. 1,3 Not listed among the seven key factors above is the effect on spending of changes in the prevalence of medical conditions beyond that caused by population aging. A number of recent studies, however, have pointed to a rise in the prevalence of conditions associated with the increase in obesity as an additional driver of spending. [4][5][6]

Several initiatives to improve care at the end of life involve educational programs to influence clinicians’ attitudes about care for patients with terminal illnesses. The objective of this research was to develop and test a short and easily administered instrument for measuring physicians’ and nurses’ attitudes towards care at the end of life. The instrument was tested using a cross-sectional study of 50 clinicians (25 physicians and 25 nurses) from general medicine, cardiology, oncology, and geriatric medicine. Both reliability and validity were assessed, and the instrument was found to have acceptable test-retest reliability and construct validity. Such an assessment instrument may be useful in evaluating the impact of initiatives to modify attitudes towards terminal care and in improving the quality of care at the end of life.