The relationship between the results of quantitative cultures of TPC and BAL specimens for the microorganisms isolated in 30 patients (r = 0.71; n = 32; p<0.0001) is shown in Figure 1.
In 28 (70 percent) of the 40 cases studied, quantitative TPC cultures yielded ^lO cfu/ml. In all but one of these cases, BAL cultures yielded ^lO cfu/ml and in five of them S pneumoniae was isolated from blood cultures. In four additional cases, TPC cultures were negative but BAL cultures yielded ^lO cfu/ml. In three cases, intermediate counts were obtained with both techniques and the etiology was judged as doubtful. In the remaining five cases, both techniques gave negative results.
Using the TPC cultures, the etiology was correctly established in 28 (70 percent) of the cases. If the BAL cutoff point had been set at 103 cfu/ml, all but one of the cases confirmed by TPC and four additional cases would have been correctly diagnosed, raising the diagnostic value to 77 percent.
BAL cultures in the control group, composed by patients clearly not having CAP, were either negative (seven cases) or yielded normal flora (three cases). In the remaining cases, 0-hemolytic Streptococcus, 1,200 cfu/ml, Staphylococcus epidermidis, 2,000 cfu/ml, and Staphylococcus aureus, 200 cfu/ml, were cultured. These microorganisms were thought to be clinically not relevant because they are components of normal flora or because low colony counts were obtained.
Figure 1. Relation between TPC and BAL quantitative cultures. Bacterial titers are expressed as log cfu/ml.