Economic Aspects of Evaluation - Evaluation Methods in Biomedical Informatics - page 304

TABLE 11.1. Types of economic evaluation in health care.
Measurement Measurement of

Type of analysis Description of cost outcome

Cost minimizing Assumes outcomes are Dollars None

equivalent

Cost consequence Lists the costs and outcomes Dollars Variable, possibly

(even if multiple endpoints multiple

are used) of each option

Cost-effectiveness Measures outcomes in clinical Dollars Clinical outcome

terms (lives, life expectancy,
number of infections averted)

Cost utility Measures outcomes in utilities: Dollars Quality-adjusted life

measures of the preferences years (QALYs)

for the outcome state

Cost-benefit Requires that outcomes (lives, Dollars Dollars

quality of life) be given a
monetary value

efits. 4 In the simplest case, an analysis of the benefits goes no further than
to assume the clinical effect of the possible strategies is the same; the only
difference arises from difference in costs. This type of analysis is termed a
cost-minimizing study, because with outcomes assumed equal, the only rel-
evant goal would be to minimize costs. The most common clinical example
of a cost-minimizing study would be an evaluation of a therapeutic drug
substitution program in which the efficacy of the two drugs was equivalent
and the analysis looks only at the costs of the drug, its administration,
and monitoring. Cost-consequence studies are only slightly more compli-
cated, as they list the costs in dollars and outcome in whatever units are
appropriate for the particular situation being evaluated. For example,
suppose new upgrades are being proposed for network hardware to try to
improve both network performance (perhaps measured in loading time for
an electronic record of a standard size) and reliability (measured in
expected number of minutes of unscheduled downtime per month). A cost-
consequence study would simply generate lists, such as option A will cost
$238/workstation, will require 6.8 milliseconds to load a chart, and will have
35 minutes of unscheduled downtime per month; whereas option B will cost
$312/workstation, will require 3.2 milliseconds to load a chart, and will have
41 minutes of unscheduled downtime each month. No attempt is made to
equate the value of downtime and speed; the results are presented to allow
the decision makers to place their own value on the relative worth of the
various outcomes and the trade-off between them.
Cost-effectiveness analysis (CEA) is defined by the use of a clinically
relevant outcome to quantify the benefits, and is consequently the most
common type of economic analysis used in health care. The clinical effec-
tiveness measure may span a wide range, from global outcomes such as lives