Cause of death: where there is no physician
The InterVA (Interpret VA) algorithm was developed and revised over a number of years, starting in 2003. Based on Bayes’ rule for conditional probabilities, for each death InterVA produces values for the likelihoods of each cause, given the indicators reported as present in a VA interview and a set of evidence. It makes use of physician-derived conditional probabilities that give the likelihoods of various indicators being associated with various causes.
For a set of deaths, InterVA sums across the likelihoods of the causes to yield the population-level fractions of deaths resulting from the causes. For each death, InterVA reports single value point estimates for the likelihoods of the three causes with the largest likelihoods, if they fall above a set threshold; otherwise, the cause is ruled ‘indeterminate’. Full details, source code, and compiled executables that implement InterVA-4 (version 4.04) are available.