A comparative study of three methods to evaluate an intervention to improve empirical antibiotic therapy for acute bacterial infections in hospitalized patients.

Background: In order to limit the use of broad-spectrum antibiotics, standardized empirical therapy against acute bacterial infections has been advocated.

Methods: Guidelines for acute bacterial infections recommending increased usage of benzylpenicillin and restricted use of fluoroquinolones and cephalosporins have been implemented in Kalmar County, Sweden. We evaluated this strategy by recording therapy in patients with bacteraemia, antibiotic requisition, and point prevalence surveys prior to this intervention and at 6 and 12 months after.

Results: Comparing the methods simultaneously, there was good agreement between them and an overall significant change in antibiotic usage. There was a significant shift from cefuroxime to cefotaxime and a borderline significant increase in the use of benzylpenicillin (p = 0.057). Based on the defined daily dose (DDD), a highly significant decrease in total cefotaxime and cefuroxime usage was observed that was not detected when applying the prescribed daily dose (PDD), which is adapted to local treatment practices. No change was found in mortality in Staphylococcus aureus bacteraemia or the incidence of Clostridium difficile infection.

Conclusions: We conclude that the implementation of the new guidelines has resulted in a significant change in antibiotic usage, which could be conveniently monitored by antibiotic requisition if PDD is used in addition to DDD.