[Pneumocystis carinii pneumonia: laboratory diagnosis]

We worked with 51 samples, 7 bronchoalveolar lavages (BAL) and 44 sputa (S) of 31 AIDS patients with clinical and radiographic symptoms compatible with Pneumocystis pneumonia. With the aim of finding a specific sensitive methodology for the diagnosis of Pneumocystis carinii, we evaluated 4 coloration techniques (silver methenamine, its modification without gold chloride, toluidine blue and Giemsa). 35% of the patients studied were positive. P. carinii were observed in 18% of the 44 sputa. We observed that the analysis of a single sputum sample (S) has a very low sensitivity and that the processing of two or more samples is necessary since only one of the 14 patients who had sent a single sample was found P. carinii positive, while in the remaining ten who had sent more than one (S) sample, the microorganism was detected in 50%. 4 of the 7 BAL were positive. 4 BAL were preceded by the analysis of an (S) sample: in two cases the results were negative while BAL allowed us to make the diagnosis, thus demonstrating its greater efficacy. To enhance sensitivity each sample was centrifuged until exhaustion and 10 slides were prepared for coloration with the final sediment. The four techniques employed were specific and all the Pneumocystis pneumonia patients responded to the treatment. Silver methenamine, its modification without gold chloride, and toluidine blue were very sensitive, in contrast to of Giemsa. The stain to be chosen is either silver methenamine, or its modification, because both achieve the best contrast, allowing optimum P. carinii identification. We suggest the implementation of some of these techniques in laboratory routine.