Assessment of antibiotic prescriptions in two French intensive care units

Introduction

In the setting of antimicrobial policy, the French authorities recommend assessing the quality of antibiotic therapy [1]. The aim of this study is to assess the adequacy of antibiotic therapy in the ICU, to good practice guidelines and unit protocols. The three steps of prescription are analysed: choice of treatment, modalities of prescription and treatment follow-up.

Methods

This preliminary observational trial was conducted in two ICUs (October 2005 to August 2006 and March to August 2009). All patients with curative antibiotherapy were prospectively included. For each patient, we collected individual data of treatment and we evaluated them according to the recommendations of good use published in the literature [1, 2].

Results

One hundred and thirteen patients were included. Infections were community-acquired infection in 30% of the cases. Respiratory tract infections were the most frequently encountered, followed by intra-abdominal infections. The three most frequently isolated pathogens were P. aeruginosa, E. coli and S. aureus.

Initial antibiotic therapy, n = 113: lack of sample 14%, inadequacy with unit protocol 13%, too broad spectrum 7%, not justified association 8%, wrong posology 20%, incorrect route of administration 1%, wrong interval 6%, lack of plasmatic dosage (n = 46) 7%.

Treatment follow-up, n = 108: wrong delay of reappraisal 14%, no adaptation to microbiological results 10%, lack of de-escalation 5%, no reassessment of posology 3%, wrong posology (if change of treatment) 2%, wrong interval (if change of treatment) 3%, unjustified duration 10%.

Conclusions

The rate of prescriptions that did not conform to at least one criterion in this study was over 70%. This high rate is partly explained by the high number of assessment criterions (n = 15). The most frequent criterion of inadequacy was a wrong initial dosage, followed by the initiation of an antibiotherapy without a microbiologic sample, a wrong delay of reassessment, and a nonconformity of treatment to unit protocols. That last criterion must be assessed in a more qualitative way, as some transgressions could be justified.