No abstract

With the dramatic increase in the frequency of Pneumocystis carinii pneumonia associated with human immunodeficiency virus infection, there has been a need for more rapid and less invasive diagnostic techniques. Recent studies have shown that examination of induced sputum can establish the diagnosis of P. carinii pneumonia in about 55 percent of cases. To assess whether a recently developed indirect immunofluorescent stain using monoclonal antibodies was more sensitive than Giemsa or toluidine blue O stains in detecting P. carinii in sputum, we undertook two prospective studies. Of 63 patients at one institution from whom sputum specimens were obtained, 49 were ultimately given a diagnosis of P. carinii pneumonia, 46 of them by staining of sputum. The sensitivity of the three stains in detecting P. carinii was 45 of 49 (92 percent) for immunofluorescence; 37 of 49 (76 percent) for Diff-Quik (a Giemsa-type stain); and 39 of 49 (80 percent) for toluidine blue O. There were no false positive immunofluorescent stains. In a similar study of a series of 25 patients at another institution, a diagnosis of P. carinii pneumonia was made in 23 of 25 patients by staining of induced sputum. We conclude that examination of induced sputum is a rapid, sensitive, and inexpensive method for diagnosing P. carinii pneumonia and that indirect immunofluorescence is a practical and highly sensitive staining technique for establishing this diagnosis.

Human immunodeficiency virus (HIV) is an important risk factor for invasive pneumococcal disease, but information on clinical course and infecting serotypes is limited. To help develop strategies to reduce the morbidity due to invasive pneumococcal disease, episodes of pneumococcal bacteremia were identified by retrospective review of microbiology records (November 1983-November 1987) at 10 San Francisco hospitals and, for patients 20-55 years old living in San Francisco, HIV antibody status was determined by review of medical records. Pneumococcal isolates from one hospital were serotyped. Of 294 patients with pneumococcal bacteremia identified, 32 (11%) had AIDS at the time pneumococcal bacteremia was diagnosed and another 43 (15%) were HIV-infected but did not have AIDS; 12 HIV-infected patients developed AIDS after the episode of pneumococcal bacteremia. The rate of pneumococcal bacteremia in AIDS patients was estimated to be 9.4/1000 patient-years. Serotypes of 27 (82%) of 33 pneumococcal isolates from HIV-infected patients and 107 (90%) from 119 patients without known HIV infection were among the 23 serotypes included in the currently available polysaccharide vaccine. The rate of pneumococcal bacteremia is approximately 100-fold greater in AIDS patients in San Francisco than rates reported before the AIDS epidemic, but more than half the episodes of pneumococcal bacteremia in HIV-infected patients occurred in patients without AIDS. Data on pneumococcal serotypes causing invasive disease in HIV-infected patients suggest that the current pneumococcal vaccine, if effective in this population, could provide significant protection against pneumococcal disease.