Background Antibacterial consumption has been associated with an increase in the growth of resistant bacterial strains. Hospital Pharmacists play an important role in antimicrobial stewardship strategies, by improving compliance with anti-infective prescribing recommendations.

Purpose To identify real-world situations in which pharmacists can intervene to improve antibacterial patterns of use in Hospital da Luz, Lisbon, Portugal.

Materials and methods Retrospective audit study. All courses of antibacterials for systemic use (therapeutic class J01), prescribed in patients over 18 years-old admitted to Hospital da Luz during January 2011, were extracted from the electronic medical records and analysed. Descriptive statistics were performed.

Results A total of 913 patients were admitted to hospital during the study period, being 81.1% (n=740) prescribed 961 antibacterial courses. Surgical prophylaxis represented 63.7% (n=612) of the courses. The following potential improvement areas were identified:

• ▶ In 4.9% (n=47) cases the reason for prescription was not identified.

• ▶ Prophylaxis duration was longer than 48 h in 2.1% and between 24 and 48 h in 10.8% courses. A clear distinction between antibacterials prescribed for prophylaxis and therapy was found, except for second-generation cephalosporins (78.3% vs 21.7%), quinolones (24.6% vs 75.4%), and imidazole derivatives (57.1% vs 42.9%).

• ▶ Only 3.8% (n=12) of the 349 non-surgical prophylactic and therapeutic antibacterial prescriptions were discontinued after microbiological identification and/or antibiotic susceptibility test results. Parenteral administration represented 73.9% (n=258) of these 349 courses, whereas only 8.4% (n=20) were discontinued due to intravenous-to-oral switch.

Conclusions After an indepth audit process, the following opportunities for pharmacists' interventions to improve the antibacterial pattern of use in our hospital were identified: unclear prescription indication, inappropriate extension of surgical prophylaxis duration, inappropriate selection of the prophylactic antibacterial agent, insufficient microbiological identification follow-up, and scarce intravenous-to-oral switch.