Impact of ureido/carboxypenicillin resistance on the prognosis of ventilator-associated pneumonia due to Pseudomonas aeruginosa
1 Surgical ICU, Hôpital Edouard Herriot, 5 place d'Arsonval, Lyon 69437, France
2 Medical Polyvalent ICU, University hospital A Michallon, Bd de la Chantourne BP 217, Grenoble 39043 Cedex 9, France
3 Integrated Research Center U823, University Grenoble 1 - Albert Bonniot Institute, Rond point de la Chantourne, La Tronche Cedex 38706, France
4 Microbiology and Infection Control Unit, Necker Teaching Hospital, 149 rue de Sevres, Paris 75743, France
5 Medical-Surgical ICU, Saint Joseph Hospital Network, 185 rue Raymond Losserand, Paris 75014, France
6 Medical ICU, Saint Louis Teaching Hospital, 1 avenue Claude Vellefaux, Paris, 75010, France
7 Outcomerea, 7 rue Louise Thuliez, Paris 75019, France
8 Surgical ICU, Antoine Béclère Teaching Hospital, 157 rue de la porte de Trivaux, Clamart 92141, France
9 Medical-Surgical ICU, Delafontaine Hospital, 2 rue du Dr Delafontaine, Saint-Denis 93200, France
10 Medical-Surgical ICU, Avicenne Teaching Hospital, 125 rue de Stalingrad, Bobigny 93009, France
Critical Care 2011, 15:R112 doi:10.1186/cc10136Published: 11 April 2011
Although Pseudomonas aeruginosa is a leading pathogen responsible for ventilator-associated pneumonia (VAP), the excess in mortality associated with multi-resistance in patients with P. aeruginosa VAP (PA-VAP), taking into account confounders such as treatment adequacy and prior length of stay in the ICU, has not yet been adequately estimated.
A total of 223 episodes of PA-VAP recorded into the Outcomerea database were evaluated. Patients with ureido/carboxy-resistant P. aeruginosa (PRPA) were compared with those with ureido/carboxy-sensitive P. aeruginosa (PSPA) after matching on duration of ICU stay at VAP onset and adjustment for confounders.
Factors associated with onset of PRPA-VAP were as follows: admission to the ICU with septic shock, broad-spectrum antimicrobials at admission, prior use of ureido/carboxypenicillin, and colonization with PRPA before infection. Adequate antimicrobial therapy was more often delayed in the PRPA group. The crude ICU mortality rate and the hospital mortality rate were not different between the PRPA and the PSPA groups. In multivariate analysis, after controlling for time in the ICU before VAP diagnosis, neither ICU death (odds ratio (OR) = 0.73; 95% confidence interval (CI): 0.32 to 1.69; P = 0.46) nor hospital death (OR = 0.87; 95% CI: 0.38 to 1.99; P = 0.74) were increased in the presence of PRPA infection. This result remained unchanged in the subgroup of 87 patients who received adequate antimicrobial treatment on the day of VAP diagnosis.
After adjustment, and despite the more frequent delay in the initiation of an adequate antimicrobial therapy in these patients, resistance to ureido/carboxypenicillin was not associated with ICU or hospital death in patients with PA-VAP.