Sputum Bacteriology and In-vitro Antibiotic Susceptibility in Hospitalized Patients with Community Acquired Pneumonia in a State Tertiary- Referral Hospital - a Retrospective Study

Introduction To review the sputum bacteriology and its in-vitro antibiotic susceptibility of patients hospitalized with community-acquired pneumonia (CAP) in a state tertiary-referral Hospital (Penang hospital, Malaysia) in order to determine the most appropriate empiric antibiotics. Patients and Methods From September 2006 to May 2007, 68 immunocompetent adult patients [mean age: 52 years (range 16-89); 69% male] admitted to respiratory wards for CAP with positive sputum isolates within 48 hours of admission were retrospectively identified and reviewed. Results 62 isolates were Gram(-) bacilli (91%) & 6 were Gram(+) cocci (9%). Two commonest isolated pathogens were Pseudomonas aeruginosa (n=20) and Klebsiella pneumoniae (n=19) which together constituted 57% of all positive isolates. Among the Pseudomonas isolates, 84.2% were fully sensitive to cefoperazone and cefoperazon/sulbactam; 95% to ceftazidime, cefepime, piperacillin/tazobactam, ciprofloxacin and amikacin, and 100% to gentamycin, netilmycin, imipenem and meropenem. Among the Klebsiella isolates, 5.3% were fully sensitive to ampicillin; 84.2% to amoxicillin, ampicllin/sulbactam, cefuroxime and ceftriazone; 89.5% to piperacillin/tazobactam; 93.3% to cefoperazon/sulbactam and 100% sensitive to ceftazidime, cefepime, ciprofloxacin, all aminoglycosides and carbopenems. Conclusion In view of the high prevalence in respiratory Pseudomonas aeruginosa, ampicillin/sulbactam, currently the most prescribed antibiotic to treat CAP in our respiratory wards, may not be the most appropriate empiric choice. Higher generation cephalosporins with or without beta-lactamase inhibitors, ciprofloxacin or carbapenem may be the more appropriate choices. The lack of information on patients’ premorbidities such as recent hospitalization and prior antibiotic exposure, limits the interpretation of our findings and may have biased our results towards higher rates of Gram negative organisms.