Quantitative assessment of chest radiographs to evaluate the efficacy of ceftazidime versus ceftriaxone in a randomized, double-blind study in adults with community-acquired pneumonia

Abstract

Background: The therapeutic strategy for antibiotic treatment of community-acquired pneumonia (CAP) and assessments of the efficacy of such treatments are usually based on microbiologic, clinical, and chest radiographic findings. However, all these techniques have several limitations, particularly evaluation of pulmonary lesions on chest radiographs (CXRs), which is still based only on semiquantitative criteria. Objective: The aim of the present study was to use a novel computerized mathematical-statistical method of digital imaging for the quantitative evaluation of standard CXRs as an additional tool to assess and compare the efficacy of different antimicrobial treatments in patients with CAP. Methods: In this prospective, multicenter (10 centers), randomized, double-blind trial, hospitalized patients with CAP received intramuscularly either ceftazidime 1 g BID for 7 days or ceftriaxone 1 g QD (plus an injection of an inert solution every 12 hours to maintain the double-blind provision) for 7 days. CXRs were obtained before, during, and at the end of treatment in both groups. The computerized digital imaging of CXRs and their consequent mathematical-statistical analyses were performed at a central institute. The clinical outcome was also evaluated by the treating physicians using a semiquantitative, 4-level scale at baseline, in the interim, and 1 to 3 days after the end of treatment. Results: Sixty-eight patients were enrolled; 33 patients received ceftazidime and 35 received ceftriaxone. The standardized analysis of CXR digital imaging showed a significantly faster regression of the original pulmonary lesions ( P = 0.024) in the ceftazidime group, although no differences were observed in clinical outcome between the 2 treatment groups. Conclusions: Some quantitative and reproducible indices that enable a comparison of the radiologic outcome in 2 treatment arms of a clinical study can be obtained with this new method. The findings indicate that this method can be used as a tool to decrease the variability in radiographic findings during antibiotic treatment for CAP.