The most frequently posed micro questions relate to the cost-effectiveness of community-based care, the resource consequences of new drug treatments, and the costs and benefits of new forms of secure accommodation. Economics has developed a number of evaluative tools to address these questions, most notably cost-effectiveness, cost-benefit, and cost-utility analyses. These different tools or modes of economic evaluation have much in common, each being concerned with the relationship between costs and outcomes. On the resource side, each measures costs according to widely accepted procedures. (23) Where they differ is in respect to their measurement of outcomes (Table 1).
Table 1 Economic evaluations—measurement of costs and outcomes
Cost-offset and cost-minimization analyses are the simplest of evaluations and are concerned only with costs. They assume either that health and quality of life outcomes have been well established from other research, or that outcomes are (currently) not measurable because of conceptual difficulties or research funding limitations. A cost-offset analysis compares costs incurred with (other) costs saved. A cost-minimization approach often proceeds in the knowledge that previous research has shown outcomes to be identical in the two or more treatment or policy alternatives being evaluated. The aim of each is to look for the lowest-cost alternative.
Other modes of economic evaluation are more interesting and informative, but correspondingly more complex to use. The best known is cost-benefit analysis. This approach is unique in that it addresses the extent to which a treatment or policy is socially worthwhile in the broadest sense, with all costs and benefits valued in the same monetary units. If benefits exceed costs, the evaluation would recommend providing the treatment, and vice versa. With two or more alternatives, the evaluation would recommend the one with the greatest net benefit as the most efficient. Cost-benefit analyses are intrinsically attractive, but conducting them is problematic in mental health care because of the difficulties associated with measuring the value of all outcomes in monetary terms. However, methodologies are being developed which aim to obtain direct valuations of health outcomes by patients, relatives, or the general public, such as 'willingness-to-pay' techniques, where an individual states the amount they would be prepared to pay (hypothetically) to achieve a given health state or health gain (see Diener et al.(4) for a review of studies in health care generally, and O'Brien et al.'(5) for an application of the method to antidepressant treatment).
Cost-effectiveness analysis is concerned with ensuring that the resources allocated to the mental health-care sector are used to maximum effect. Cost-effectiveness analysis is usually employed to help decision makers choose between alternative interventions available to or aimed at specific patient groups: if two care options are of equal cost, which provides the greater effectiveness? Or if two options have been found to be equally effective in terms of reduced symptoms, improved functioning, or enhanced quality of life, which is less costly? In the strict sense, a cost-effectiveness analysis looks at a single effectiveness dimension and constructs a cost-effectiveness ratio. The treatment with the greatest effectiveness per pound or dollar spent is then deemed the most cost-effective. For example, Jonsson and Bebbington(6> compare the cost per successfully treated patient using alternative antidepressants.
A generalization of cost-effectiveness analysis to multiple outcomes is cost-consequences analysis, which cannot produce a simple ratio measure of effectiveness to costs, but has the ability to evaluate policies and practices in a way which comes much closer to everyday reality. As with the results from any evaluation, the findings from a cost-consequence analysis would need to be reviewed by decision makers, and the different outcomes weighed up and compared with costs. The decision calculus may be more complicated than when using cost-effectiveness ratios or monetary measures of cost-benefit differences, but decision makers in health-care systems—from strategic policy makers to individual professionals—make these kinds of decisions every day. An example of a cost-consequences analysis is given below.
Cost-utility analysis is similar to cost-effectiveness analysis but it measures and then values the impact of an intervention in terms of improvements in health-related quality of life. The value of the quality of life improvement is measured in units of 'utility', expressed by a combined index of the mortality and quality of life effects of an intervention (the quality-adjusted life year (QALY) is the best known index). This is in contrast to cost-benefit analysis which uses monetary values, and also in contrast to cost-effectiveness analysis and cost-consequences analysis which rely on the kinds of measures conventionally found in clinical evaluations (symptom scores, behaviour ratings, and so on). Cost-utility analyses avoid the potential ambiguities with multidimensional outcomes but are obviously more general than the single-outcome cost-effectiveness analysis. The result is a series of cost-utility ratios (such as the cost per QALY year gained) upon which health-care resource allocation priorities can be based. They can be applied to choices across a range of treatments or diagnoses: they can compare one health-care area with another. The transparency of the methods used to derive utility scores is a particular strength, but the currently available measures (which are intended for use across the widest diagnostic range) may not be sensitive enough to the kinds of changes usually found in mental health care to provide the sole indicator of impact. (7>
Many design issues must be addressed before the principles of economic evaluation can be turned into practical (and insightful) empirical studies. The next sections discuss the main issues involved and offer illustrations of completed evaluations.
Was this article helpful?