Infections due to single or predominant pathogens must be
presented separately from true polymicrobial infections.
Data must be presented for both clinical (cure, improved,
failure) and bacterial responses. When the category `improved'
is used, clinical response data must be broken down into sepa-
rate `cure', ``improved'' and `failure' categories. Bacteriologic
response data must be presented with and without `presumed'
outcomes included. In addition, data must be presented for
each MIC