JAC Advance Access originally published online on August 17, 2006
Journal of Antimicrobial Chemotherapy 2006 58(4):900-901; doi:10.1093/jac/dkl344
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Correspondence |
Comment on: Hospital consumption of antibiotics in 15 European countries: results of the ESAC Retrospective Data Collection (1997--2002)
Center for Infections Diseases and Travel Medicine, University Hospital Hugstetter Strasse 55, D-79106 Freiburg, Germany
*Corresponding author. Tel: +49-761-270-1819; Fax: +49-761-270-1820; E-mail: kern{at}if-freiburg.de
Keywords: pharmacoepidemiology , public health , fluoroquinolones , macrolides , population-based study
Sir,
The assessment of antimicrobial use and the determination of need for control measures have become important priorities of public health research and policy. Recently, retrospective data on antimicrobial drug consumption were presented by the European Surveillance of Antimicrobial Consumption (ESAC) project, established in 2001 with support of the European Commission. The ESAC database for outpatient antibiotic use is large and comprehensive1 but more limited for hospital antibiotic consumption.2 For the year 2002, ‘fairly complete’ national consumption data of systemic antibiotics for hospital care from quite different sources (manufacturers, wholesalers and pharmacies) were available for only 15 countries, namely Belgium, Croatia, Denmark, Estonia, Greece, Finland, France, Hungary, Luxembourg, Malta, Norway, Poland, Slovakia, Slovenia and Sweden.2 Validity was considered a problem for the datasets from Finland (obviously including parts of outpatient consumption), Poland (coverage of 60% only) and Greece (other reasons).
How can coverage and quality of the desperately needed antibiotic use database for hospital care be improved? What is the recommendation to public health authorities and researchers of those (often very large) countries which were unable to deliver ‘fairly complete’ datasets? Are the hospital sample data reported from the Netherlands (covering 58% of all hospitals), for example, not valid, reliable and sufficiently complete?3 Is it worthwhile to collect national consumption data from 
2000 hospitals in a large country like Germany, with its complex, decentralized health system? Is it not more acceptable (and perhaps desirable) to provide sample data from pharmacies than sales data from manufacturers/wholesalers? What is the critical sample size? How do hospital samples need to be stratified? Are regional data from large countries such as Italy, Spain, the UK and Germany relevant to reduce the white spots on the map?
Data from a convenience sample of hospitals (40 regional acute-care general hospitals and 2 university hospitals) located in south-western Germany, for example, are available for the years 2000–2002.4 The hospitals, located in Baden-Württemberg, a federal state with a 10.6 million population—a population size similar to that of Belgium, Greece, Portugal or Sweden—had a distribution of hospital size similar to that of all general hospitals in that state. This sample data and corresponding hospital statistics can easily be used to estimate state-wide hospital antibiotic consumption and to compare it with state-wide outpatient antibiotic consumption. We performed such an analysis for the year 2002.
For this purpose, use density estimates [in DDD/100 occupied bed-days (DDD/100)] were obtained separately for the different services of university hospitals (such as general surgery, orthopaedic surgery, ENT and surgical ICUs) and separately for medicine, surgery and ICUs in non-university hospitals. We calculated the 95% confidence interval of mean inpatient antibiotic use density per service for regional hospitals and assumed an `error' of ±10% in the estimates for university hospital services (due to the small number of university hospitals included) to obtain the lower and upper limit of the estimate for total hospital antibiotic consumption. We added an estimate for children's hospitals based on the same regional hospital sample. The sample size was 3 439 000 out of a total of 15 551 000 state-wide acute-care occupied bed-days in 2002 (22%), excluding radiotherapy/nuclear medicine hospital departments, psychiatry/psychotherapy and inpatient rehabilitation services, for which we assumed a negligible use density. Calculations were weighted according to the proportion of occupied bed-days in 2002 by each service. Finally, hospital consumption was expressed as DDD per 1000 inhabitants and day (DID).
This calculation yielded an estimate of 1.99 DID for hospital antibiotic consumption in Baden-Württemberg, with a ‘plausible range’ of 1.83–2.17. Considering outpatient (including nursing homes) antibiotic use estimates for the same region and year, this represents 14% (13.2–14.2%) of the total antibiotic consumption—a slightly higher percentage than reported by ESAC for most other European countries included in their survey.2 When we repeated these measurements for selected drug classes, we estimated that 21% of fluoroquinolone DDDs were prescribed in the hospital; the corresponding proportion for co-trimoxazole was 7%, for macrolides/clindamycin it was 5%, and for tetracyclines it was 1%.
Hospital antibiotic consumption of 2 DID for this part of Germany is well within the order of magnitude reported in the ESAC work (Figure 1). If one accepts that Finland (probably due to healthcentre inclusion) and France (for unknown reason) are outliers in the ESAC database, there is still some variation in hospital antibiotic use in the DID data format between many European countries and regions. Interestingly, this variation is much less evident in a comparison of hospital antibiotic use density estimates in the DDD/100 data format from France,5 Denmark,6 the Netherlands3 and different regions of Germany.7
|
There may be fewer white spots than anticipated on the European map of hospital antibiotic consumption. It is certainly possible (albeit cumbersome, we agree) to reduce them further, but a careful definition of acceptable data sources, quality, type and format is required. In large countries that are not yet on the map, the provision of reliable data may be feasible from stratified hospital samples, and this may be the only acceptable way for cost-effective data acquisition and analysis.
Transparency declarations
Nothing to declare.
Acknowledgements
This study has been supported in part by grant BMBF 01KI9951 from the German Federal Ministry of Education and Research.
References
1 Goossens H, Ferech M, Vander Stichele RH, et al. (2005) Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet 365:579–87.[Web of Science][Medline]
2
Vander Stichele RH, Elseviers MM, Ferech M, et al. (2006) Hospital consumption of antibiotics in 15 European countries: results of the ESAC Retrospective Data Collection (1997–2002). J Antimicrob Chemother 58:159–67.
3 Liem TB, Filius FM, van der Linden PD, et al. (2005) Changes in antibiotic use in Dutch hospitals over a six-year period: 1997 to 2002. Neth J Med 63:354–60.[Web of Science][Medline]
4 de With K, Steib-Bauert M, Schröder H, et al. (2006) Estimating hospital versus ambulatory care consumption of antibiotics in southwestern Germany. Clin Microbiol Infect 12:Suppl 4, Abstract P1479.
5 Rogues AM, Placet-Thomazeau B, Parneix P, et al. (2004) Use of antibiotics in hospitals in south-western France. J Hosp Infect 58:187–92.[CrossRef][Web of Science][Medline]
6
Muller-Pebody B, Muscat M, Pelle B, et al. (2004) Increase and change in pattern of hospital antimicrobial use, Denmark, 1997–2001. J Antimicrob Chemother 54:1122–6.
7 Steib-Bauert M, de With K, Meyer E, et al. (2006) Antibiotic consumption in German acute care hospitals. Clin Microbiol Infect 12:Suppl 4, Abstract P1467.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Berrington Hospital antibiotic prescribing data require careful local interpretation J. Antimicrob. Chemother., January 1, 2007; 59(1): 162 - 163. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||