STUDY OBJECTIVE: To evaluate the accuracy of PLA with a thin needle in the bacteriologic diagnosis of patients with lung abscess and in demonstrating possible coexistence of an underlying lung carcinoma, and the influence of this technique in the treatment and outcome of these patients. DESIGN: Case series. SETTING: Tertiary university referral center. PATIENTS: Consecutive sample of 50 patients with clinical picture suggestive of pulmonary infection and single or multiple cavitation of at least 1 cm in diameter on chest x-ray films, and lack of clinical suspicion of active pulmonary tuberculosis. One patient was excluded from the study after demonstration of tuberculosis by PLA. INTERVENTIONS: Lung aspirates were obtained under fluoroscopic guidance by introduction of a 22-gauge disposable spinal needle within the abscess cavity and were immediately transported to the bacteriology laboratory and pathology department for processing. All patients were initially treated with clindamycin. Tobramycin was added in all those patients with hospital-acquired infection, lack of foulness of sputum, and those who were initially severely ill. Definite treatment was based on the results of bacteriologic cultures. MEASUREMENTS AND MAIN RESULTS: Cultures of LAs were positive in 82 percent (40/49) of patients. In 20 cases the isolates were monobacterial (13 aerobic bacteria and seven anaerobic). In the remaining 20 cases, cultures grew more than one kind of bacteria (four exclusively aerobic, five exclusively anaerobic, and 11 mixed), with an average of 3.25 types of bacteria per case. Anaerobes were found as a single bacteria or associated with other aerobic bacteria in only 58 percent (23/40). The results of LA cultures led to change in the initial antibiotic trial in 23 patients (47 percent). Of ten cases in which bronchogenic carcinoma was demonstrated, cytologic study of LA was done in nine, and eight had positive cytologic yield. Pneumothorax occurred in seven cases (14 percent) as the sole complication. CONCLUSIONS: (1) Percutaneous lung aspiration had a high diagnostic yield and accuracy in our series, with a relatively low incidence of complications. (2) Anaerobic bacteria were less frequently implicated in our cases than previously reported. This finding led to significant changes in the initial empiric antibiotic treatment.
STUDY OBJECTIVE: To evaluate the accuracy of PLA with a thin needle in the bacteriologic diagnosis of patients with lung abscess and in demonstrating possible coexistence of an underlying lung carcinoma, and the influence of this technique in the treatment and outcome of these patients. DESIGN: Case series. SETTING: Tertiary university referral center. PATIENTS: Consecutive sample of 50 patients with clinical picture suggestive of pulmonary infection and single or multiple cavitation of at least 1 cm in diameter on chest x-ray films, and lack of clinical suspicion of active pulmonary tuberculosis. One patient was excluded from the study after demonstration of tuberculosis by PLA. INTERVENTIONS: Lung aspirates were obtained under fluoroscopic guidance by introduction of a 22-gauge disposable spinal needle within the abscess cavity and were immediately transported to the bacteriology laboratory and pathology department for processing. All patients were initially treated with clindamycin. Tobramycin was added in all those patients with hospital-acquired infection, lack of foulness of sputum, and those who were initially severely ill. Definite treatment was based on the results of bacteriologic cultures. MEASUREMENTS AND MAIN RESULTS: Cultures of LAs were positive in 82 percent (40/49) of patients. In 20 cases the isolates were monobacterial (13 aerobic bacteria and seven anaerobic). In the remaining 20 cases, cultures grew more than one kind of bacteria (four exclusively aerobic, five exclusively anaerobic, and 11 mixed), with an average of 3.25 types of bacteria per case. Anaerobes were found as a single bacteria or associated with other aerobic bacteria in only 58 percent (23/40). The results of LA cultures led to change in the initial antibiotic trial in 23 patients (47 percent). Of ten cases in which bronchogenic carcinoma was demonstrated, cytologic study of LA was done in nine, and eight had positive cytologic yield. Pneumothorax occurred in seven cases (14 percent) as the sole complication. CONCLUSIONS: (1) Percutaneous lung aspiration had a high diagnostic yield and accuracy in our series, with a relatively low incidence of complications. (2) Anaerobic bacteria were less frequently implicated in our cases than previously reported. This finding led to significant changes in the initial empiric antibiotic treatment.