Pneumocystis carinii organisms cluster in alveolar casts. A method for concentrating alveolar casts in sputum specimens might improve the diagnostic usefulness of sputum studies in possible cases of pulmonary infection with Pneumocystis carinii. The use of sputum induction to detect Pneumocystis carinii was studied in 40 consecutive patients with known or suspected human immunodeficiency virus infection having bronchoscopy for evaluation of pulmonary disease. Sputum produced by deep coughing induced by 3% saline via a jet nebulizer was liquefied with dithiothreitol, and cells were sedimented and stained. Liquefaction of mucus allowed concentration of alveolar casts where P. carinii cysts were readily identified. A total of 28 patients were found to have P. carinii, which was detected in concentrated sputum in 21 of 38 patients, unconcentrated sputum in 10 of 38 patients, and by bronchoscopy in 25 of 37 patients. Sensitivity of concentrated sputum compared with bronchoscopy was 78% (95% CI, 58 to 90), with a negative predictive value of 71% (95% CI, 47 to 87). Values for unconcentrated sputum were 43% (95% CI, 25 to 63) and 48% (95% CI, 30 to 67), respectively. Liquefaction of adequately induced sputum facilitates processing and interpretation and increases yield for P. carinii, eliminating the need for invasive procedures in most patients.

No abstract

1. Sputum sols, obtained from patients with chronic obstructive pulmonary diseases, were applied to frogs' isolated palates and the effect on ciliary beat frequency was examined. 2. During clinical exacerbations, a significant number of asthmatic patients produced a distinct, slurry type of sputum, which induced ciliostasis. This ciliary inhibition largely disappeared with clinical improvement, at which time there was also a change of the type of sputum from slurry to mucoid. 3. Samples obtained from normal volunteers and from the great majority of the patients, excluding those with asthma, had no effect on ciliary beating.

_The natural variation in viscosity and elasticity in a single 0.5-ml sample of sputum, as well as among several expectorates of a patient during a single day and during a three-day period, has been quantitated. Other properties of sputum, such as gross appearance, cellular density, percentage of solid content, types of cells present, and pH were also studied, and possible correlations were investigated. The viscosity variation (at a shear rate of 1 sec-l) averaged 11 ± 8 per cent (mean ± SD) for a single sample; 26 ± 18 per cent for three expectorates of a single patient during one day; 49 ± 33 per cent for one patient during three days. Variation in elasticity averaged 6 per cent for one sample; 25 ± 15 per cent for one day; 33 ± 9 per cent for three days.Samples classified by eye as mucoid or purulent according to pre-established color and opacity criteria showed a close correlation between this classification and the variables described above. Samples called purulent had a high density of cells with 85 per cent of the samples filling twothirds or more of a low power (40X) field on wet mount whereas mucoid samples all filled cine half or less of the same field. All samples called purulent had a cellular distribution showing more than 70 per cent neutrophils, whereas 80 per cent of mucoid samples had a neutrophil percentage of less than 60 per cent. The concentration of solids for all purulent samples was between 6 and 10 per cent whereas 83 per cent of mucoid samples were between 2 and 5 per cent. Purulent samples had a slightly higher average pH of 7.83 than the 7.59 of mucoid samples. The elastic recoil of purulent samples, measured at 100 dyne per cm. 2 , averaged 4.41 units, which was less than the 6.98 units for mucoid samples.The large variability in viscosity among patients precluded the definition of quantitative differences ,between purulent and mucoid sputum. Nevertheless, whenever there was a change in the type of sputum of a given patient, purulent sputum was always more viscous.