A causal probabilistic network for optimal treatment of bacterial infections

The fatality rate associated with severe bacterial infections is about 30 p ercent and appropriate antibiotic treatment reduces it by half. Unfortunate ly, about a third of antibiotic treatments prescribed by physicians are ina ppropriate. We have built a causal probabilistic network (CPN) for treatmen t of severe bacterial infections. The net is based on modules, each module representing a site of infection. The general configuration of a module is as follows: Major distribution factors define groups of patients, each of t hem with a definite prevalence of infection caused by a given pathogen. Min or distribution factors multiply the likelihood of one pathogen, without ch anging much of the prevalence of infection. Infection caused by a pathogen causes local and generalized signs and symptoms. Antibiotic treatment is ap propriate if it matches the susceptibility of the pathogens in vitro and ap propriate treatment is associated with a gain in life expectancy. This is b alanced against the cost of the drug, side effects, and ecological damage, to reach the most cost effective treatment. The net was constructed in such a way that the data for the conditional probability tables will be availab le, even ii it meant sometimes giving up on fine modeling details. For data , we used large databases collected by us in the last 10 years and data fro m the literature. The CPN was a convenient way to combine data from databas es collected at different locations and times with published information. A lthough the net is based on detailed and large databases, its calibration t o new sites requires data that is available in most modern hospitals.