Background: Important hospital-acquired infections include pneumonia, mainly because of the increasing resistance of bacterial pathogens to antimicrobials and the associated potential failure of antibiotic therapy. The present study aimed at determining the most frequent etiological agents of hospital-acquired pneumonia (HAP) and assessing the relationship between 30-day mortality and adequacy of antibiotic therapy. Based on the obtained information, optimal patterns of antibiotic therapy were to be defined, including a pharmacoeconomic perspective.
Methods: In patients with clinically confirmed HAP, bacterial etiological agents were identified, their susceptibility to antimicrobials was determined and statistical methods were used to assess the relationship between adequacy of antibiotic therapy and 30-day mortality.
Results: The study comprised 68 patients with clinically confirmed HAP. The most common etiological agents were strains of Pseudomonas aeruginosa (30.8 %), Klebsiella pneumoniae (23.1 %) and Burkholderia cepacia complex (15.4 %). Gram-negative bacteria accounted for 86.5 % of all bacterial pathogens. The overall mortality reached 42.5 %. In the subgroup of patients with inadequate antibiotic therapy, 30-day mortality was significantly higher (83.3 %) than in the subgroup with adequate therapy (30.0 %; p = 0.002). The risk for 30-day mortality was 2.78 times higher in case of inadequate antibiotic therapy (95%CI: 1.52-5.07). The proportion of Pseudomonas aeruginosa strains was significantly higher in the subgroup of patients with inadequate antibiotic therapy than in those with adequate therapy (67 % vs. 27 %; p = 0.032).
Conclusion: Results of the present study suggest a significant relationship between mortality of patients with HAP and ineffective antibiotic therapy due to resistance of the bacterial pathogen. Thus, it is clear that initial antibiotic therapy must be based on qualified assumption of sufficient activity against the most common bacterial pathogens and results of surveillance of bacterial resistance in the relevant epidemiological unit. At the same time, however, it must be stressed that it is impossible to cover all potential variants of the etiological agents and their resistance phenotypes.