Abstract
A series of 391 unselected expectorated sputum specimens was examined microscopically, and six different published criteria for judging the acceptability of the specimens were applied. Of the 391 specimens, 234 were found to be acceptable or unacceptable by all six criteria; 157 specimens were discrepant. By the criteria of Murray and Washington and of Barry, 25 and 23% of the specimens, respectively, were rejected; only 19 of 143 specimens which contained potential pathogens as part of their predominant microbial flora were rejected by both sets of criteria. The criteria described by Geckler et al. and Bartlett missed fewer potential pathogens; only 9 or 17% of the specimens, respectively, were found unacceptable. The criteria of Heineman and Radano and of Van Scoy resulted in the greatest percentages of specimens judged unacceptable (28 and 29%, respectively), including 19 and 24% of specimens containing potential pathogens. The reproducibility of sputum screening results was also assessed, comparing the method of Murray and Washington with that of Barry. Six separate slides were prepared from each of 45 different specimens: three samples with purulent or bloody flecks and three samples in which the specimens had been mixed with an applicator swab. Satisfactory reproducibility was observed with both criteria and both sampling methods; no significant differences in reproducibility could be documented in this limited series.