Last update:
17-Mar-2007

Rational clinical decisions are divided inτο decisions with certainty, decisions under risk and decisions under uncertainty. In decisions with certainty the choice has to be made between acts that lead with certainty to defined outcomes. In decisions under risk, each action may lead to different possible outcomes, each of which has a subjective probability to happen. The criterion of these decisions is the criterion of Bayes. In decisions under uncertainty, outcomes of possible choices have not been given subjective probabilities. In decisions under risk, in order to decide which action to take, the clinician need only estimate the disease probability and compare it with the treatment threshold. Treatment threshold is the disease probability at which there should be no difference between giving treatment and withholding treatment. Treatment threshold is the disease treatment below which treatment is not given and above which treatment is given. Treatment threshold divides patients into those who are treated and those who are not treated. Treatment threshold is equal to C/C+B, where B is the net benefit of treating diseased patients and is therefore the difference in utility between treating and not treating a diseased person [B=U(D⁺T⁺)-U(D⁺T^-), U: utility, D: disease, T: treatment). C is the net cost of treating well persons and is therefore the difference in utility between treating and not treating a well person [C=U(D^-T^-)-U(D^-T⁺)].

Key words: Benefit, Cost, Expected utility, Treatment threshold, Utility.