The incidence of Pneumocystis carinii pneumonia has increased primarily because of the increase in the number of persons with human immunodeficiency virus (HIV) infection. Consequently, there has been an impetus to develop faster and more economical diagnostic techniques. Earlier reports have indicated that pneumocystis pneumonia can be diagnosed from recovery of the pneumocystis organism from sputum. We sought to determine the sensitivity of sputum induction supervised by respiratory care practitioners using ultrasonic nebulization as a noninvasive diagnostic procedure in patients with suspected pneumocystis pneumonia. METHODS & MATERIALS: From January 1987 through December 1989, all patients at our institution suspected of having pneumocystis pneumonia underwent sputum induction with 3% sodium chloride administered via ultrasonic nebulizer. RESULTS: 429 sputum-induction procedures were performed on 272 patients (198 HIV-positive, 74 HIV-negative). In patients who had multiple inductions, each procedure was considered a separate patient episode. Eighty-eight sputum specimens were positive for P carinii, 328 specimens were negative, and sputum could not be obtained 13 times. One hundred and fifty-four of the 328 sputum-negative patient episodes were followed by bronchoscopy that yielded sputum positive for P carinii in 16 patients. The 174 sputum-negative patients who were not followed by bronchoscopy were followed clinically for a minimum of 3 months. Of those 174 patients, 161 (128 HIV-positive and 33 HIV-negative) improved without treatment and 13 were treated empirically and improved. The 13 patients who were unable to produce sputum were followed by bronchoscopy with 5 positive and 8 negative specimens resulting. If one considers the true positives to be comprised of those who were positive by sputum induction, those negative by sputum induction but positive by bronchoscopy, and those negative by sputum induction but positive by clinical presentation who improved with treatment, and if one considers the false negatives to be those who were negative by sputum induction but positive by bronchoscopy or negative by sputum induction but positive by clinical presentation who improved with treatment, then sensitivity was 83% in HIV-positive patients and 69% in HIV-negative patients. CONCLUSIONS: Sputum induction via ultrasonic nebulization is a sensitive, noninvasive technique that can be effectively administered by respiratory therapists. Although we did not directly study cost, time, and morbidity, the technique may contribute to a reduction in those factors in establishing the diagnosis of P carinii pneumonia.
Godwin et al. (Tue,) studied this question.