Guideline-Based Clinical Assessment Versus Procalcitonin-Guided Antibiotic Use in Pneumonia: A Pragmatic Randomized Trial

Study objective Efforts to reduce unnecessary and unnecessarily long antibiotic treatment for community-acquired pneumonia have been attempted through use of procalcitonin and through guidelines based on serial clinical assessment. Our aim is to compare guideline-based clinical assessment– and procalcitonin algorithm–guided antibiotic use among patients with community-acquired pneumonia. Methods We performed a pragmatic, randomized, multicenter trial from November 2012 to April 2015 at 12 French hospitals. We included emergency department (ED) patients older than 18 years with community-acquired pneumonia. Patients were randomly assigned to either the procalcitonin-guided or clinical assessment group. In accordance with past studies, we hypothesized that serial clinical assessment would be superior to procalcitonin-guided care. The primary outcome was antibiotic duration, and secondary outcomes included rates of antibiotic duration less than or equal to 5 days, and clinical success and combined serious adverse outcomes at 30 days in the intention-to-treat population. Results Of 370 eligible patients, 285 (77%) were randomly assigned to either clinical assessment– (n=143) or procalcitonin-guided care (n=142). Median age was 67 years (range 18 to 93 years) and 40% of patients were deemed to have Pneumonia Severity Index class IV or V. Procalcitonin algorithm adherence was 76%. Antibiotic duration was not significantly different between clinical assessment– and procalcitonin-guided groups (median 9 versus 10 days, respectively). Clinical success rate was 92% in each group and serious adverse outcome rates were similar (15% versus 20%, respectively). Conclusion Guideline-based serial clinical assessment did not reduce antibiotic exposure compared with procalcitonin-guided care among ED patients with community-acquired pneumonia. The strategies were similar in terms of duration of antibiotic use and clinical outcomes.