Common elements of economic evaluations
| ELEMENT | DESCRIPTION |
|---|---|
| Target population | The group of patients to whom the results apply; can be broken down into subgroups to address variability in the target audience |
| Comparators | The set of interventions under study. All medically acceptable comparators that fall within the technological capacity of the target setting should be included, and if they are not included, reasons for their omission should be clearly stated. It is appropriate to include a “do nothing” comparator if some patients in routine practice are not receiving specific care for a problem |
| Perspective | The viewpoint from which the costs are measured, commonly that of the publicly funded health care system or society* |
| Time horizon | Time during which costs and benefits are measured, typically a patient’s lifetime |
| Outcomes | Clinical outcomes of interest: QALYs, deaths, or other end points (eg, stroke, myocardial infarction). In cost-effectiveness analysis, final outcomes are preferred, such as death or hard clinical end points like myocardial infarction, as opposed to intermediate outcomes like blood pressure reduction or cholesterol reduction. In the case of QALYs, the valuation technique and the source of utility weights should be described |
| Data inputs | Effectiveness data: the expected relative difference in outcomes of 1 intervention vs 1 or more others; RCTs are the preferred source, and if more than 1 RCT has been done, meta-analysis can be used. Adverse events that result from a treatment should be included when they are serious or resource-intensive Supplementary clinical data: used to model the natural history of illness, often derived from observational cohort studies or administrative data Cost data: direct medical costs, including the cost of the treatment or intervention, costs associated with the intervention (eg, laboratory monitoring or adverse events related to the intervention), and future health care costs. If the societal perspective is taken, indirect costs and their valuation methods should be included. Sources include literature, fee schedules, and administrative data |
| Discounting | Future health effects and costs are valued in present terms, accounting for the differential timing of when costs are incurred and medical benefits are accrued. Rates of 3%–6% are typical, with Canadian guidelines recommending a rate of 5%10 |
| Results | Disaggregated results detail the total costs and consequences of each comparator. The summary statistic is the ICER; ICERs represent the cost per unit of effect or QALY gained over the next best alternative |
| Sensitivity analysis | Used to address parameter uncertainty in a model or to identify gaps in evidence. Calculations are repeated substituting a range of plausible values for 1 or more input parameters. If the results remain consistent, then the analysis is said to be robust |
ICER—incremental cost-effectiveness ratio or cost-utility ratio, QALY—quality-adjusted life-year, RCT—randomized controlled trial.
↵* When the perspective of the publicly funded health care system is used, as recommended in Canadian guidelines,10 only direct medical costs, such as physician fees and hospital charges, and costs borne by patients are included. Studies undertaken from the societal perspective will include indirect medical costs such as lost productivity due to illness; however, for many interventions, indirect costs might not differ between comparators.
Data from Drummond et al7 and Hunink et al.11