Cost-Effectiveness Plane

Stochastic cost-effectiveness analysis

· Cost-effectiveness planes

· Cost-effectiveness ratio confidence intervals

· Cost-effectiveness acceptability

· Net benefit approach

· The relevance of inference?

Cost-effectiveness plane

CER = Cost-effectiveness Ratio

CER confidence intervals

Problems

· Independent cost and effect confidence intervals may give too wide an overall confidence interval

· The probability distribution of a ratio of two random variables does not have a known distribution

· The ratio has discontinuities at the boundaries between different quadrants

Some solutions

· The “box” method – separate distributions

· The “ellipse” method – joint distributions

· Fieller’s theorem

· Bootstrapping

Cost-effectiveness acceptability

Cost-effectiveness decision rule

ICER= DC/DE

Compare with “ceiling ratio” = Rc

If Rc > DC/DE, treatment is cost-effective

Rc can also be interpreted as the cost-effectiveness threshold or the willingness to pay for a unit of health effect

Cost-effectiveness acceptability

Different thresholds for gains and losses in health

Cost-effectiveness acceptability

Calculate the proportion of the distribution (P) which lies below the line defining a particular ceiling ratio (Rc)

Cost-effectiveness acceptability

Calculate Pi for each level of Rci

Cost-effectiveness acceptability

Cost-Effectiveness Acceptability Curve (CEAC)

The Net Benefit Approach

ICER= DC/DE

where DC is in £ and DE is not

Net benefit = DE - DC

where DC and DE are in the same units

Rc can be used to convert costs to the same units as effects, or effects to the same units as costs:

Monetary net benefit (MNB):

MNB = Rc *DE - DC

Health net benefit (HNB):

HNB = DE - DC/Rc

The Net Benefit Approach

Cost-effectiveness decision rule

If Rc > DC/DE, treatment is cost-effective

Net benefits decision rule

Monetary net benefit:

If Rc *DE - DC > 0, treatment is cost-effective

Health net benefit:

If DE - DC/Rc > 0, treatment is cost-effective

Net Monetary Benefit Curve

The relevance of inference?

· Inference about CERs may lead to different conclusions than inference about costs and effects

· Are decisions based on statistical significance perverse?

· Is quantifying uncertainty useful to decision makers?

· Expected value of information

Reading

There are many relevant publications. However, this lecture only gives an overview of the area. A more detailed treatment will be given in the Clinical Trials and Survival Analysis course, where a full reading list will be given. It is suggested that you read the following easily available article:

Glick H, Briggs AH, Polsky D. Quantifying Stochastic Uncertainty and Presenting Results of Cost-effectiveness Analyses, Expert Review of Pharmacoeconomics and Outcomes Research 2001; 1(1): 25-36.

http://www.ihs.ox.ac.uk/herc/publications/glick.pdf