Risk-adjusted antibiotic consumption in 34 public acute hospitals in Ireland, 2006 to 2014

• Published in: Euro surveillance : bulletin européen sur les maladies transmissibles 2016-08, Vol.21 (32), p.1
• Main Authors: Oza, Ajay, Donohue, Fionnuala, Johnson, Howard, Cunney, Robert
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Anti-Bacterial Agents - therapeutic use; Antibiotics; Child; Child, Preschool; Databases as Topic; Drug dosages; Drug Utilization - statistics & numerical data; Female; Hospitals, Public; Humans; Inpatients - statistics & numerical data; Ireland; Length of Stay; Male; Middle Aged; Regression Analysis; Retrospective Studies; Risk Adjustment; Risk assessment; Surveillance and Outbreak Report; Tertiary Care Centers; Young Adult
• ISSN: 1560-7917; 1025-496X; 1560-7917
• DOI: 10.2807/1560-7917.ES.2016.21.32.30312

As antibiotic consumption rates between hospitals can vary depending on the characteristics of the patients treated, risk-adjustment that compensates for the patient-based variation is required to assess the impact of any stewardship measures. The aim of this study was to investigate the usefulness of patient-based administrative data variables for adjusting aggregate hospital antibiotic consumption rates. Data on total inpatient antibiotics and six broad subclasses were sourced from 34 acute hospitals from 2006 to 2014. Aggregate annual patient administration data were divided into explanatory variables, including major diagnostic categories, for each hospital. Multivariable regression models were used to identify factors affecting antibiotic consumption. Coefficient of variation of the root mean squared errors (CV-RMSE) for the total antibiotic usage model was very good (11%), however, the value for two of the models was poor (> 30%). The overall inpatient antibiotic consumption increased from 82.5 defined daily doses (DDD)/100 bed-days used in 2006 to 89.2 DDD/100 bed-days used in 2014; the increase was not significant after risk-adjustment. During the same period, consumption of carbapenems increased significantly, while usage of fluoroquinolones decreased. In conclusion, patient-based administrative data variables are useful for adjusting hospital antibiotic consumption rates, although additional variables should also be employed.