Background According to Portuguese legislation, consumption of carbapenems and quinolones should be reduced by 50% by 2020. The aim of the Antibiotic Stewardship Programmes is surveillance of antibiotic consumption and a decrease in incidence rate of multidrug resistant microorganisms.

Purpose The objectives of GCL-PPCIRA (Local Coordination Group Prevention Programme and Infections Control and Antimicrobial Resistance) were to reduce the number of inappropriate prescriptions and duration of antibiotic therapy, and prevent the emergence of antimicrobial resistance.

Material and methods This was a prospective study. Antibiotics prescriptions between June 2015 and May 2016 were analysed by the pharmacist by reviewing patient medical records. The pharmacist classified the requirements as empirical, inappropriate and documented prescriptions. Statistics: Stata 10.1 (5% significance level).

Results 618 patients were identified, mean (SD) age 68 (16.6) years and 54.9% men, resulting in 734 prescriptions of carbapenems (71%) and quinolones (29%). The most common site of infection was the urinary tract (31.3%). According to the prescribed therapeutic intervention, 150 (20.4%) were empirical, 275 (37.5%) were inappropriate, 231 (31.5%) were documented and 77 (10.5) were according to the protocol approved by the institution.

Mean duration of treatment was 9.4 days for documented prescriptions, 8.8 days for empirical prescriptions, 7.1 days for prescriptions according to the protocol and 6 days for inappropriate prescriptions (p=0.0001). PPCIRA changed 118 (16.1%) prescriptions. The interventions reduced the mean duration of therapy: 4.7 days for prescriptions with interventions and 8.4 days for those without (p<0.0001). It was found that in 362 prescriptions with microbial isolates, 201 were multidrug resistant microorganisms (55.5%).

Prescriptions for patients who were discharged with an antibiotic (23.7%) had a lower mean duration of treatment and a lower proportion of prescriptions with multidrug resistant microorganisms than prescriptions for patients who were discharged without an antibiotic (61.7%) or for patients who died (14.6%): 6.4 days and 33.8% for multidrug resistant microorganisms, 8.2 and 62.8% and 8.1 and 49.1%, respectively (p=0.0001 and p<0.001). 14 (13.1%) deaths were directly attributed to infection.

Conclusion The PPCIRA work has resulted in a timely intervention during the prescription process. The investment in surveillance of therapeutic protocols has been reflected in a decrease in the duration of inappropriate prescriptions and the fulfilment of targeted therapy.

References and/or acknowledgements Despacho No 3844-A/2016 DR 2ª serie No 52–15 de Março de2016.

No conflict of interest