Double-blind comparison of cefamandole and penicillin in pneumococcal pneumonia.

We conducted a prospective, randomized, double-blind comparison of intravenous penicillin and cefamandole in the therapy of pneumococcal pneumonia. Patients received either 1 g of cefamandole or 600,000 U of aqueous penicillin G every 6 h. Of the 100 patients entered into the study, 96 had clinical and radiographic evidence of pneumonia. Microbial etiology was determined from the results of sputum and blood cultures and/or sputum Gram stains. Streptococcus pneumoniae was pathogenic in 49 patients, of whom 24 received cefamandole and 25 received penicillin. There was no statistically significant difference in the response or cure rate. Of the 100 patients, 93 were treated for 3 days or more and were evaluated for adverse effects and toxicity. There was no significant difference between cefamandole-treated and pencillin-treated patients in the incidence of colonization, superinfection, phlebitis, thrombocytosis, decrease in hematocrit, or elevated liver function tests. Eosinophilia occurred more frequently in patients treated with penicillin (20 of 42) than in those treated with cefamandole (11 of 42 (chi square, P < 0.05). Only one patient receiving cefamandole developed a positive direct Coombs test. No patient in either group developed meningitis. We conclude that, with the doses and route of administration employed in this study, cefamandole is as effective as penicillin in the therapy of pneumococcal pneumonia without an increased incidence of colonization, superinfection, or adverse effects.