Cost-Comparison Analysis

A cost-comparison analysis estimates the total costs of two or more interventions, including downstream costs, and the numbers of individuals affected by each intervention but does not estimate cost-effectiveness ratios relative to health outcomes. This approach was developed in the early 1970s as a method of cost accounting with specific applications to ascertaining the lowest-cost methods of pharmacologic dosing and laboratory testing. An assumption that is usually either explicit or implicit in such analyses is that health outcomes are comparable across interventions. Otherwise, the lowest-cost strategy would not necessarily be desirable.

A cost-comparison analysis, which is also commonly referred to as a cost-consequences analysis, is less demanding to perform because it does not require clinical or epidemiologic data on health outcomes, such as long-term morbidity or mortality, although short-term clinical outcomes or healthcare use are typically reported. This approach is attractive in assessing interventions for which it is difficult to ascertain ultimate health outcomes or to calculate summary measures of health that integrate multiple outcomes. The cost-comparison approach is particularly well-suited to assessing screening and diagnostic-testing strategies. It is typical for such analyses to report summary cost ratios, such as cost per individual tested or cost per case detected, for each strategy, as well as incremental cost ratios for pairwise comparisons.

The time horizon, or the period during which healthcare utilization and costs are included in the analysis, is variable for cost-comparison (or cost-consequences) studies. For analyses of pharmacological or surgical interventions, the time horizon that is used is typically quite short, often 12 months to several years from the time of intervention. On the other hand, cost-comparison analyses of genetic testing strategies typically project the costs of monitoring tested individuals over their remaining lifetimes, which can be 40 years or more.

Most published cost-comparison analyses are conducted from the perspective of a healthcare system and only include direct medical costs. However, it is also valuable to calculate cost-comparison analyses from the societal perspective and to include costs occurring outside the healthcare system. Costs of time spent by patients and family members are important to include for interventions requiring substantial time by individuals and relatives. The exclusion of such costs can make such interventions appear more cost-effective than they are. In particular, if one is interested in comparing the actual costs of clinic-based and home-based therapeutic or rehabilitative strategies from a societal perspective, it is essential to include the costs of unpaid or informal caregiving services.

Prior to the mid-1990s, clear distinctions were generally made between cost-comparison, cost-minimization, and cost-consequence analyses. Since then, differences among these methods have become blurred, and articles using them frequently overlap one another. A given analysis that reports or assumes equivalent outcomes of different interventions might be labeled as a cost-comparison analysis, cost-consequence(s) analysis, cost-minimization analysis, or even cost-effectiveness analysis, depending on the preferences of the authors. Consequently, readers should not assume that the terminology used to describe such studies necessarily corresponds to differences in the analytic methods employed. Originally, cost-comparison analyses reported data only on costs, not on outcomes; cost-minimization analyses reported on costs only after ascertaining that health outcomes were equivalent for the interventions being compared; and cost-consequence analyses reported both costs and health outcomes but did not explicitly compare the two in terms of ratios (to let decision makers decide which information is needed to draw inferences).

...