Criteria for probabilistic dominance of one strategy over another - both of the following (page 214):
(1) for each possible outcome the probability of achieving this outcome or any worse outcome is no greater for the first strategy than for the second
(2) for at least one possible outcome the probability of achieving this outcome or any worse outcome is lower for the first strategy than for the second
where:
• To simplify analysis, the outcomes should be entered by severity rank.
• If the outcomes are the same (none lower for one) then the 2 strategies would seem to be probababilistic equivalence.
• The probabilities need to be rounded to a clinically relevant number.
If there are 4 outcomes (A, B. C. D) being evaluated:
probability for outcome 1 for strategy A <= probability for outcome 1 for strategy B
probability for outcomes 1+2 for strategy A <= probability for outcomes 1+2 for strategy B
probability for outcome 1+2+3 for strategy A <= probability for outcome 1+2+3 for strategy B
probability for outcome 1+2+3+4 for strategy A <= probability for outcome 1+2+3+4 for strategy B
probability for outcome 2 for strategy A <= probability for outcome 2 for strategy B
probability for outcomes 2+3 for strategy A <= probability for outcomes 2+3 for strategy B
probability for outcome 2+3+4 for strategy A <= probability for outcome 2+3+4 for strategy B
probability for outcome 3 for strategy A <= probability for outcome 3 for strategy B
probability for outcomes 3+4 for strategy A <= probability for outcomes 3+4 for strategy B
probability for outcome 4 for strategy A <= probability for outcome 4 for strategy B
NOTE: The above listing is according to the definition. However, it could be simplified to the equations comparing single outcomes (probability of X for strategy A <= probability of X for strategy B) for all possible outcomes. If this is true, then the sum of probabilities for all subset outcomes will be <= also.
