Improvements for multipurpose bacteriological identification tables to suit the diagnosis of Vibrio cholerae

The aim of the study was to reduce to key tests the 4 extensive polyvalent diagnostic biochemical tables most widely used in Croatia and to adapt them for the demonstration of Vibrio cholerae and its differentiation from the 3 Vibrios (V. alvinolyticus, V. mentschikovii, V. fluvialis) important in differential diagnosis. The fourth table has now been adapted to differentiate among all 12 Vibrio species known to be human pathogens (V. mimicus, V. cincinatiensis, V. holisae, V. damsela, V. furnisi, V. parahaemolyticus, V. vulnificus, V. carchariae). Using the inductive Learning by Logic Minimization Method (ILLM), we analyzed 2 tables (i.e. identification matrices) that were a part of bioMérieux's commercial packaged polyvalent identification systems widely used in Croatia (API 20E and ATB 32E), as well as 2 compilation tables by M. T. Kelly et al. The tables contained 27, 32, 59 and 8 tests, respectively. Cutting these solely to the key tests involved rationalizing them from 59 to the 5 necessary to differentiate Vibrio cholerae from 3 related Vibrios. Further rationalizations were from 32 to 2 and from 27 to the 3 necessary to differentiate Vibrio cholerae from 2 related Vibrios. By reducing the table of 8 tests to 7, and adding 4 new ones to these we achieved an optimization permitting mutual differentiation of all 12 known human Vibrio pathogens. Use of the selective TCBS plating medium was the only precondition for making these tables effective.