Community-acquired pneumonia and positive urinary antigen tests: Factors associated with targeted antibiotic therapy

The use of rapid microbiological tests is supported by antimicrobial stewardship policies. Targeted antibiotic therapy (TAT) for community-acquired pneumonia (CAP) with positive urinary antigen test (UAT) has been associated with a favorable impact on outcome. We aimed to determine the factors associated with TAT prescription. We conducted a retrospective multicenter study including all patients presenting with CAP and positive UAT for Streptococcus pneumoniae or Legionella pneumophila from January 2010 to December 2013. Patients presenting with aspiration pneumonia, coinfection, and neutropenia were excluded. CAP severity was assessed using the Pneumonia Severity Index (PSI). TAT was defined as the administration of amoxicillin for pneumococcal infection and either macrolides or fluoroquinolones (inactive against S. pneumoniae) for Legionella infection. A total of 861 patients were included, including 687 pneumococcal infections and 174 legionellosis from eight facilities and 37 medical departments. TAT was prescribed to 273 patients (32%). Four factors were found independently associated with a lower rate of TAT: a PSI score≥4 (OR 0.37), Hospital A (OR 0.41), hospitalization in the intensive care unit (OR 0.44), and cardiac comorbidities (OR 0.60). Four other factors were associated with a high rate of TAT: positive blood culture for S. pneumoniae (OR 2.32), Hospitals B (OR 2.34), E (OR 2.68), and H (OR 9.32). TAT in CAP with positive UAT was related to the hospitals as well as to patient characteristics. L’utilisation des tests rapides est recommandée pour le bon usage des antibiotiques. L’antibiothérapie ciblée des pneumonies aiguës communautaires (PAC) avec antigènes urinaires positifs (AUP) est associée à un meilleur pronostic. Notre objectif était de déterminer les facteurs associés à la prescription d’une antibiothérapie ciblée. Étude multicentrique rétrospective incluant tous les patients avec une PAC, avec AUP pour le pneumocoque ou Legionella pneumophila (janvier 2010–décembre 2013). Les patients présentant une pneumonie d’inhalation, une co-infection ou une neutropénie étaient exclus. La gravité de la PAC était estimée par score de Fine. L’antibiothérapie ciblée était définie par l’administration d’amoxicilline pour une infection à pneumocoque ou d’un macrolide ou d’une fluoroquinolone (inactive sur le pneumocoque) dans la légionellose. Au total, 861 patients inclus : 687 infections à pneumocoque et 174 légionelloses de huit établissements de s