Figure 4. A decision tree can be used to calculate conditional probabilities after a positive test result in one parent, assuming that a child is known to be suspectible to malignant hyperthermia (MHS) and that one and only one parent can be MHS. Probabilities of a branch are indicated by italics below the line. (A) Using the “incomplete” testing strategy, the untested parent is assumed to be negative. The probability that the first parent's result is a true positive (upper branch) is the positive predictive value (PPV; italics). The probability that it is a false positive (lower branch) is calculated as 1 - PPV. Therefore, the probability of error in the automatic assumption that the second parent is negative is calculated as 1 - PPV. (B) With a “complete” testing strategy, the second parent is also tested. If the first parent had a true-positive result (first upper branch), the second parent must be healthy under our assumptions. Therefore, the probability that diagnostic testing is negative is the specificity. The only other possible result is a false-positive one, which will occur with a probability of 1 - specificity. Returning to the root of the tree, if the first parent had a false-positive result, he or she must be, in reality, healthy (first lower branch). Therefore, under our assumptions, the second parent must be MHS, so a positive result for the second parent will occur with a probability of (sensitivity), and a false negative result with a probability of 1 - sensitivity. The chance of observing a false-negative result is calculated by multiplying the probabilities that the first parent will have a false-positive result (1 - PPV) by the probability of a false-negative result in the second parent (1 - sensitivity). Table 4shows important calculations.