Sputum induction compared with bronchoalveolar lavage by Ballard catheter to diagnose Pneumocystis carinii pneumonia.

Specimens from sputum induction (SI) are often of insufficient volume or have poor diagnostic yield for the diagnosis of Pneumocystis carinii pneumonia (PCP). The Ballard catheter for bronchoalveolar lavage (BAL) is a disposable BAL device which is passed transnasally into the airway and wedged by sensation. Thirty consecutive episodes (21 inpatient and 9 outpatient) in 28 patients positive for HIV (15 with AIDS) were studied with SI and BAL at a tertiary-care university hospital. Six SIs yielded no specimen, and six were judged inadequate for investigation by the laboratory. The BAL return averaged 53 ml (range, 10 to 77 ml), and all specimens were excellent quality based on microscopy. Of the 10 patients (33 percent) who were PCP-positive on BAL, only 2 (6.9 percent) were detected by SI (McNemar p = 0.0078). Of 12 patients in whom SI was unobtainable or inadequate, 6 were positive for PCP on BAL. Two adequate specimens from SI were negative, but BAL specimens were positive for PCP. No patients had specimens that were positive for PCP on SI who had negative BAL specimens. The adequacy of the specimen and the PCP diagnosis with BAL were statistically superior to SI (McNemar p = 0.007). The Ballard BAL catheter allows easy transnasal access to the airway and safe BAL, with a statistically significant superior yield of specimens when compared to SI.