No abstract

Samples of sputum were examined microscopically to determine their suitability for routine culture. When the number of squamous epithelial cells per field was less than 10, the number of bacterial species generally fell within the range of one to four. Squamous epithelial cells were not always a true indication because some unmarked transtracheal specimens showing more than 10 squamous epithelial cells also gave a range of isolation falling between one and four. When the presence of 25 or more polymorphs was used as the parameter, the number of bacterial isolates generally fell within the range of one to three, but this resulted in positive overbiasing with consequent rejection of valid specimens. Later it was found that when a differential system using both polymorphonuclear cells and squamous epithelial cells was applied, a significant number of specimens could be salvaged which would otherwise have been discarded.

Four media, Lowenstein-Jensen, Middlebrook, Petragnani, and ribonucleic acid, were tested for comparative ability to detect Mycobacterium tuberculosis. Specimens used included sputum, urine, tissue, and gastric washings. Three types of comparison were used: (i) comparison derived from randomized specimens; (ii) comparison of cultures from newly diagnosed cases that had received no prior therapy; and (iii) comparison of cultures from specimens whose initial direct smears were negative. Overall, ribonucleic acid medium performed best, but the differences among the four media were small.