Traditionally, the assessment of new therapeutic strategies for cancer has relied almost solely on safety and clinical efficacy. Recent concerns over escalating US health-care expenditures and with optimizing patient outcomes have led to increased interest in the economic assessment of cancer therapies as a secondary objective in the evaluation of such treatments. In particular, attention has focused on the efficiency of resource allocation in health-care delivery. New economic measures are increasingly being integrated into cancer clinical trials to assess the cost-effectiveness of therapy and patients' quality of life after treatment. These measures will improve the assessment of both the costs and benefits of cancer treatment strategies.
Given the growing recognition by clinicians, patients, and health-care financing organizations of resource constraints on the provision of health-care services, economic evaluation of new drugs and new technologies has taken on greater importance [1-3]. Economic evaluation is being used to help guide clinical practice by informing clinicians and patients about specific screening and treatment strategies. These assessments are often performed by hospital or health maintenance organization formulary committees as part of the selection process for preferred treatments within these organizations [4-6]. They are also being used increasingly by health policy decision-makers at the national level in Canada and Australia to help make funding decisions for their national health insurance systems [4,5], as well as in the United States through consensus treatment and screening recommendations [7-15].
Economic evaluation includes an assessment of both the costs and benefits of cancer therapies. The costs of care may fall into four categories:
Direct medical costs, which are the costs of medical services
Direct nonmedical costs, which are the costs incurred in receiving medical care (such as transportation to and from a bone marrow transplant center)
Indirect costs, which are the costs of morbidity and mortality due to illness
Intangible costs, which are the costs of pain and suffering related to illness.
An economic assessment of cancer treatment may include any or all of these categories of costs.
Costs may be calculated from different perspectives. Costs of medical care can be computed from the viewpoint of the patient, caregiver, physician, payor, or society. The calculation of societal costs represents the total costs of the transaction, whereas the calculation of costs from the other perspectives may include only those costs relevant to that specific party. For example, a patient whose insurance requires a 20% copayment incurs a $1,000 hospitalization cost. Thus, the insurer would pay $800 and the patient would pay $200. The costs of that hospitalization episode viewed from the perspectives of the insurer, patient, and society are $800, $200, and $1,000, respectively. An economic evaluation can measure costs from a single perspective or multiple perspectives.
Increasingly, investigators are interested in assessing quality of life or patients' perceptions of their health status as primary or secondary end points for cancer clinical trials [16-59]. This type of assessment is a relatively new area for cancer clinical studies. Inclusion of quality-of-life end points in clinical studies reflects the growing recognition that traditional clinical trial end points may not adequately capture the impact of cancer and cancer treatment on patients' lives. In broad terms, quality of life can be measured within clinical trials using two different measurement constructs: functional status and patient preference.
Functional status instruments are used to quantify the ability or limitation of patients across specific areas of functioning related to health. These areas, called "domains," generally include physical, emotional, and social functioning. Functional status instruments compare patients to standardized levels of functioning, the results of which are then quantified into a quality-of-life score. Functional status measures are further divided into global and disease-specific functional status measures.
Global assessment measures, such as the Sickness Impact Profile (SIP) , the Medical Outcomes Study Short Form (MOS-SF-36) , and the Nottingham Health Profile (NHP) , have been widely used, and may be generalizable across populations of patients with different clinical conditions. However, global measures may not be sensitive to small changes in health status within therapeutic categories.
Disease-specific measures typically include some of the same elements measured by global assessments but are designed to be more responsive to the illness being evaluated.Various cancer-specific functional status instruments have been developed to focus on conditions experienced by cancer patients and on the adverse side effects of cancer therapies [16-59].
In contrast to functional status instruments, which evaluate patient health status against standardized levels of functioning, patient preference assessments measure patients' interpretation of their health states [63-65]. Patient preference measures include developed instruments as well as specific techniques to elicit these data.
The most common instruments used to assess patient preference are the EuroQol© , the Quality of Well Being (QWB) scale , and the Mark Health Utilities Index (Mark HUI-III) [30,31,65]. Both the QWB and Mark HUI-III evaluate functional status across specific domains and then combine these measures into a single preference or utility score based on a predetermined population preference assessment for each component of the different health states described by the instrument. The EuroQol© asks patients to evaluate their current health state using a 0 to 100 scale, where 0 is the worst imaginable health state and 100 is the best imaginable state.
Preference elicitation techniques are interview methods that use risk or uncertainty explicitly to elicit patient preferences for their health states. The standard gamble[64,65] and the time tradeoff [68,69], for example, offer the patient the choice between a health state of defined impairment and an option requiring patients to take a risk on their health outcomes offering a probability of perfect health and the reciprocal probability of death. Patient preference for a given health state is calculated from the point of indifference between the two options. The economic outcome measure, quality-adjusted life-years (QALYS), is calculated using both survival and patient preference information .
Different outcome measures can be used to assess the clinical benefits of therapy for an economic evaluation. These outcome measures are intermediate clinical end points, such as complete or partial response to treatment, and final end points, such as a change in survival. The final end points are more generalizable across all the study populations within an economic evaluation.
Interpretation of economic studies in the assessment of cancer screening or treatment strategies is based on the relationship between the costs and benefits of the strategy being assessed. As depicted in Figure 1, an economic study can have four potential results: An improvement in clinical outcome at a reduced cost of medical services (+/+, upper left box) would be considered a dominant strategy and would be adopted. A strategy that resulted in an increase in the cost of medical services while yielding a reduction in clinical outcomes (-/-, bottom right box) would be considered to be dominated and would be rejected.
The other two potential results (the shaded boxes, +/- and -/+) are ambiguous, and their interpretation depends on both the magnitude of the change in cost and the clinical outcome, as compared with the base-case strategy. To make a treatment decision in these ambiguous cases, cost-effectiveness analysis would be used to compare the costs of the medical services provided relative to the clinical outcome.