ABSTRACT
Figure 3.1 Comparisons of inflammatory indices in bronchoalveolar lavage (BAL) and sputum. The differences (log values) between induced sputum (IS) and BAL eosinophil cationic protein (ECP) and fibrinogen concentrations are plotted as a function of the mean of the two values. Intraclass correlation coefficients (R) are shown. Solid lines represent the mean difference; dashed lines represent ± 2 standard deviations (SD) of mean difference. In a Bland-Altman graph, if the variable is reproducible, it is expected that 95% of the differences between measures are less than 2 SD. Since all the data points are present above the line depicting the mean difference, this indicates that the levels in IS are consistently higher than those in BAL. Adapted with permission from Pizzichini E, Pizzichini MM, Kidney JC, et al. Induced sputum, bronchoalveolar lavage and blood from mild asthmatics: inflammatory cells, lymphocyte subsets and soluble markers compared. Eur Respir 1998; I 1:828-34
Table 3.1 Factors that influence interpretation of mediator measurements in sputum
1. Method of induction:
a. spontaneously expectorated sputum or induced by saline
b. use of normal (isotonic) or hypertonic saline
c. duration of induction
d. volume of nebulized saline delivered
e. interval between inductions
2. Preparation and processing of sputum
a. sampling of selected or entire (unselected) sputum
b. processing with reducing agents
FACTORS RELATED TO THE METHOD OF SPUTUM INDUCTION OF RELEVANCE TO MEDIATOR MEASUREMENT
Few individuals, except for those with purulent pulmonary conditions such as cystic fibrosis or COPD, can expectorate spontaneously. Nevertheless, in some cases, especially when safety is the issue, a sample has to suffice. This can have a major effect on mediator measurement, which must be considered when interpreting the data. The concentration of fibrinogen has been found to be higher in spontaneously expectorated sputum than in an induced sample (Figure 3.2). There is also a trend towards increased ECP in the spontaneous sample.
