The Prognostic Significance of Small Numbers of Tubercle Bacilli in the Sputum, Detected by the Concentration Smear Method *

The detection of tubercle bacilli in the sputum is regarded by many as of unfavorable prognostic significance, regardless of the number detected and the frequency with which they are found. This study is an attempt to evaluate the prognostic significance of small numbers of tubercle bacilli detected in the sputum by concentration methods. Sputum with modified Gaffky counts of i, ii, or iii were considered. These contain, roughly, from 1 to 4 tubercle bacilli in two glass smears, 1 tubercle bacillus per microscopic field, and from 2 to 3 bacilli per field, respectively. Each half of the smears was examined for ten minutes. The sputum had been collected for three days, treated with normal sodium hydroxide (4 per cent), shaken and centrifuged, and the sediment examined after staining with carbol-fuchsin (Cooper modification). That tubercle bacilli may evade detection by this method is easily understood. As computed by William Steenken, Jr., an ordinary glass smear usually covers the slide for the entire width and a linear distance of from 1 Y4 to 1X2 inches. In this area there are about 51,906 oil immersion fields. In an examination lasting ten minutes, about 600 fields are examined, leaving about 51,000 unexamined. Thus, the ratio of examined to unexamined fields is approximately 1:85 and it would seem relatively easy to have small numbers of tubercle bacilli present without their being detected. That this does occur is shown by the fact that if the sediment is planted on six tubes of media at the time the smears are stained, fully one-fourth of the sputum specimens will show colonies of tubercle bacilli on culture while the stained smears are negative (Steenken et al.2). Inasmuch as comparisons will be made of patients with negative sputum smears and those with positive smears,