Answers are grounded in this paper and cite their sources.
Dear Sir: In a recent Perspective article in this journal (1), Roux and Donaldson criticize the usefulness of cost-of-illness (COI) (1) studies, which they suggest are fraught with methodological flaws. Furthermore, they contend that COI studies should not be used to set priorities for government investment in the prevention of disease and injury. We acknowledge the validity of some of their arguments and wholeheartedly agree that scientifically rigorous economic evaluation (such as cost-benefit analysis and cost-effectiveness analysis) should be used in decisions for funding interventions. However, as we recently noted (2), COI studies can be valuable tools for promoting attention to the economic burden imposed by specific illnesses and are a crucial first step in the economic evaluation of prevention interventions. As suggested by Roux and Donaldson, COI data may assist in mobilizing interest and resources to a particular public health problem. We argue that this is how many COI studies are being used in the government: not to explicitly set health policy but rather to spotlight burden of disease that moves beyond traditional epidemiological estimates of mortality and morbidity. For example, a recent COI study by Finkelstein et al. (3) showed that the government pays for >50% of the health care costs associated with persons who are obese or overweight. These findings clarify the debate about whether obesity is a personal or societal issue and provide a clear motivation for government to try to reduce the costs of obesity. Because economic evaluation is useful only after interventions have been developed and evaluated, COI studies are an important part of the process for priority-setting in research on prevention. Many of the methodological “flaws” that Roux and Donaldson contend to be inherent in COI studies, such as overestimating productivity losses and double-counting costs for chronic diseases, can be overcome. For example, concerns about using earnings to measure productivity have led many prevalence-based COI studies to focus on direct costs of care only. Furthermore, whereas the human capital approach has limitations to valuing lost productivity, this is not specific to COI studies. Indeed, cost-benefit analyses in the health care field often use the human capital approach to calculate the economic benefits from preventing death and disability. Many recent COI studies rely on individual level data that allow for directly modeling the relationship between a particular disease and expenditures (4) and, therefore, do not suffer from problems of double-counting expenditures that were most likely to occur in epidemiological COI studies. As stated in our own critique of COI studies, the potential for misuse of these, and many other, studies is ever-present, and, therefore, authors need to clearly state how the results should and should not be used. We also believe that COI studies will continue to benefit from better data and methods. Further debate on this subject should focus on how to advance the state of science of COI studies and how to transfer the results of these studies to policy makers required to allocate our scarce public health resources.
No takes yet. Share an insight, caveat, or question.
Corso et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: