Types of economic evaluation

The term economic evaluation is sometimes used inter-changeably with the term cost-effectiveness. There are three main types of economic evaluation as described in Drummond et al. (2005). These are cost-effectiveness analysis comprising cost-effectiveness analysis, cost-consequence analysis and cost-minimisation analysis as well as cost-utility analysis and cost-benefit analysis. Each type of economic evaluation is summarised next.

1. Cost-effectiveness analysis
1. Cost-effectiveness analysis (CEA)

Cost-effectiveness analysis compares the cost of each intervention with a single, common outcome. The outcome chosen is measured in a natural unit of effect and the choice of unit used reflects the health care field being evaluated. For example, for life-saving therapies such as treatments for chronic renal failure, the most appropriate effectiveness measure would be 'years of life gained' or 'premature death deferred'. Hence, cost-effectiveness analysis may be used to compare the costs and consequences of hospital haemodialysis versus renal transplantation. In contrast, in programmes where symptom relief is important, for example in the asthma field, relevant natural units include asthma-free days or symptom-free days.

2. Cost-consequence analysis (CCA)

A type of CEA. Cost-consequence analysis compares the costs and consequences of two or more interventions and the consequences are presented in a disaggregated way and are not combined with costs. CCA is similar to CEA in that outcomes chosen are measured in a natural unit of effect. However, it differs from CEA as a number of outcomes may be presented in disaggregated form, rather than focusing on a single outcome. CCA is used to display all the key costs and consequences associated with the interventions for comparison and the consequences should be measured in the most appropriate units. For example, consider the impact of two interventions on diabetes. Outcomes of interest may include HbA1c levels, adverse events, quality of life, diabetes-related complications and blood pressure. A CCA would present the costs of the two interventions along with the impact on each of these markers. The decision-maker could then apply their own weighting system to value the impact of all of the consequences of the two interventions and compare these with costs to determine which represents the best value for money.

3. Cost-minimisation analysis (CMA)

A type of CEA. Cost-minimisation analysis compares the cost of two or more interventions and it assumes that the interventions achieve the given outcome to the same extent. It is similar to CEA in that the outcomes that are assumed equivalent are measured in a natural unit of effect. However, it differs from CEA in that it involves the comparison of costs alone and its aim is to find the least-cost alternative.

An example of CMA is the comparison of minor surgery, such as hernia repair, undertaken as day-case surgery versus surgery using a conventional inpatient stay. If it is assumed that both types of surgery deliver the same outcome, such as the number of operations successfully completed, it would be possible to compare the two interventions to establish the least-costly alternative. However, it is not self-evident that day-case surgery would be the least cost alternative, simply because it involved (by definition) a shorter hospital stay. If it was found that day-case surgery led to more post-operative complications, then arguably it would be too simplistic to focus on costs alone and it may be preferable to examine the costs and outcomes of each alternative.

2. Cost-utility analysis (CUA)

Cost-utility analysis compares the cost of each intervention with a single, common outcome known as a utility. Utility is defined here as the preferences individuals or society may have for a particular set of outcomes.

The use of utility, as opposed to any other outcome is underlined by the concept that the two measures are different. It implies that the utility of treatment is different between individuals even though the treatment outcome may be the same. To illustrate this, take the preferences of two individuals, a professional footballer and an author, both of whom break a leg in a car accident. Also assume that there are a number of features of the injury that are identical, such as the type and extent of fracture, the surgical treatment required and the prognosis in terms of healing time. Whilst the treatment and outcome may be exactly the same for both individuals, the (dis)utility placed on this is likely to differ considerably across these individuals. For the footballer there is the disappointment of missing perhaps a full season, together with some uncertainty as to whether he will be the same player again following the recovery period. Although there are unlikely to be lost earnings in the short-term, the injury may have considerable longer-term consequences in terms of transfer value. The author, however, while inconvenienced by the injury, may not consider this to be a major handicap (as opposed to fracturing his or her writing arm for example), and with suitable word processing facilities at home may be able to carry on with daily activities largely unaffected. It is unlikely that any current or future earning potential would be affected by the period of disability.

In the above example the individual's preferences are considered. However, if preferences were considered based on a societal perspective, the value placed on the different outcomes may be very different, potentially influenced by the preferences people have for watching football or reading books.

A well-known example of an outcome used in CUA is the cost per Quality-Adjusted Life Year (QALY), introduced by the late Professor Alan Williams in 1995. This allows outcomes to be expressed in a single index (the QALY) that combines aspects of quality of life, using the utility estimate, and life expectancy. Therefore the length of time spent in each health state is adjusted by the utility associated with each health state. In the above example, the impact on survival is likely to be negligible or even zero, however, the quality of life implications of the injury are likely to differ markedly between the two individuals.

3. Cost-benefit analysis (CBA)

Cost-benefit analysis compares the cost of each intervention with their outcomes, measured in monetary units. In principle all costs and consequences associated with the interventions may be compared and therefore it is possible to examine whether the total costs of an intervention are justified by its total benefits, including non-health benefits. Methods for valuing outcomes in monetary units comprise revealed preference techniques, based on consumer behaviour, and stated preference techniques based on willingness to pay (WTP) techniques such as contingent valuation and choice experiments. For example, Donaldson et al (1996) used WTP to monetise the reassurance value to the public associated with avoiding food-borne risk.

Donaldson C, Mapp T, Ryan M et al. Estimating the economic benefits of avoiding food-borne risk: is 'willingness to pay' feasible? Epidemiology and Infection. 1996; 116: 285-294.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=8666072&dopt=Abstract

Drummond MF, Sculpher MJ, Torrance GW et al., Methods for the economic evaluation of health care programmes. 3rd ed. 2005, Oxford: Oxford University Press.
http://www.oup.com/uk/catalogue/?ci=9780198529453