Determinants of antibiotic use in surgical wards and in intensive care units in southwestern France

Abstract number: P1267

Dumartin C., Rogues A.M., Amadeo B., Fourrier-Reglat A., Gachie J.P., Parneix P.

Objective: To study factors associated with antibiotic (AB) consumption in surgical wards and in intensive care units (ICU) and their relevance for benchmarking, in the context of recent French guidelines promoting reporting of AB use to health authorities.

Method: Data were retrospectively recorded for the whole years 2002 and 2005 from a convenience sample (51 and 55 surgical wards with 1,105,232 and 991,506 PD, 35 and 31 ICU with 139,102 and 139,976 PD): consumption of AB for systemic use (defined daily doses per 1,000 patient-days: DDD/1,000 PD and per 100 admissions: DDD/100 AD), number of central-line per 1,000 PD (2005) and hospital characteristics (administrative type, length of stay (LOS), number of beds: total and for ICU and surgical wards). Additional data were obtained from nosocomial infection surveys for 13 ICU and 23 surgical wards. AB were classified according WHO recommendations (ATC-DDD system, 2005). Multiple linear regression was performed in a backward stepwise approach.

Results: Median AB consumption, expressed in DDD/1,000 PD and DDD/100 AD was 1186 and 621 in 2002 and 1416 and 932 in 2005 in ICU; 477 and 268 in 2002 and 505 and 237 in 2005 in surgical units. Administrative type of hospital was associated with total AB use in both areas with highest consumption in public hospitals. Pattern of AB use was associated with LOS in ICU: use of penicillins with betalactamase inhibitors (J01CR) was lower in ICU with higher LOS, use of glycopeptides (J01XA) was higher. Total AB use in ICU was linked with number of central-line per 1,000 PD in 2005. Surgical units in public and private hospitals had different patterns of consumption, with more first and second generation cephalosporins in private and more J01CR in public hospitals in 2005. No relationship was found between patient characteristics and AB use for the 23 surgical units providing data on patient case-mix.

Conclusion: Specific data on risk factors in ICU or surgery were difficult to collect and did not seem to be more linked with AB use than readily available characteristics such as type of hospital, LOS, number of beds and of admissions for each area, number of central-line. Therefore, these factors seem relevant as surrogate to reflect patients case-mix, and should be used for stratification when comparing AB consumption from various hospitals.

Session Details

Date: 19/04/2008
Time: 00:00-00:00
Session name: 18th European Congress of Clinical Microbiology and Infectious Diseases
Location: Barcelona, Spain
Presentation type:
Back to top