Study objective: To evaluate the safety and efficacy of transthoracic aspiration with an ultrathin needle in the microbiologic diagnosis of pulmonary infections in HIV-infected patients.
Design: Retrospective review of cases.
Setting: A 500-bed teaching hospital in Lleida, Spain.
Patients: Forty-five HIV-infected patients admitted between March 1989 and March 1993 with clinical and roentgenographic evidence of pulmonary infection and without contraindications for transthoracic needle aspiration (TNA).
Interventions: Forty-seven TNAs were performed in the emergency room (20) or during hospitalization (27). The TNA procedures were done without premedication and without fluoroscopic guidance. Specimens were processed using routine microbiologic and cytologic techniques; in addition, polymerase chain reaction (PCR) for Pneumocystis carinii was carried out since March 1992. Development of adverse effects was carefully evaluated.
Results: The TNA was effective in 29 (62 percent) out of 47 procedures. The diagnosis was obtained for 14 of 15 patients with P carinii pneumonia, 8 out of 14 patients with bacterial pneumonia, and 4 out of 12 patients with tuberculosis. Other pathogens recovered were Nocardia asteroides, Cryptococcus neoformans, Rhodococcus equi, and Mycobacterium avium. No false-positive results were obtained. Pneumothorax developed in eight (17 percent) procedures, but only one procedure resulted in a pleural drainage; the incidence of other adverse effects was low and clinically irrelevant.
Conclusion: Our study suggests that TNA can be a useful technique in establishing the etiologic diagnosis of pulmonary infections in HIV-infected patients, with a good sensitivity, high specificity, and relatively low incidence of serious complications, with TNA appearing as a reliable alternative to more uncomfortable methods.