Will a Reduction in Community Prescribing of Quinolones Decrease the Prevalence of Quinolone Resistance?
- 1Division of Clinical and Population Sciences and Education, University of Dundee, Dundee, Scotland, United Kingdom
- 2Department of Infectious Diseases, Ninewells Hospital, Dundee, Scotland, United Kingdom
- Reprints or correspondence: Prof Peter Davey, Div of Community and Population Sciences and Education, Mackenzie Bldg, Kirsty Semple Way, Dundee DD24BF, Scotland (p.g.davey{at}cpse.dundee.ac.uk).
Worldwide use of fluroquinolones for 2 decades has resulted in increasing prevalence of resistance [1, 2] and the emergence of plasmid-mediated resistance mechanisms [3]. Resistance in bacteria that cause community-acquired infections arises because of selection pressure from antibiotic use in hospitals, primary care facilities, or animals combined with the spread of resistance mechanisms among bacteria [2]. Most of the evidence that links community antibiotic use with resistance is observational and, therefore, is particularly susceptible to bias and confounding [4]. Nonetheless, a strong statistical association exists, and the consistency of evidence supports a cause-effect relationship [4]. The strongest scientific evidence comes from studies that sample individual patients before and after treatment; these studies show that even transient exposure to antibiotics substantially increases the risk of isolation of antibiotic-resistant bacteria from the normal flora [5, 6]. Mathematical models of the relationship between antimicrobial consumption in human communities and the frequency of resistance suggest that the impact of a reduction in prescribing may be modest and is critically dependent on the fitness cost of resistance to bacteria [7–9]. A systematic review in 2005 identified only 4 studies that provided evidence about the effect that interventions to reduce community antibiotic prescribing have on resistance, all focused on penicillin or macrolide resistance in streptococci [10]. The one study that showed a sustained reduction in resistance after an intervention had 4 years of follow-up [11], whereas the negative studies all had only 1 year of follow-up [10]. This suggests that it takes years for a reduction in prescribing to have an effect on resistance in streptococci.
In this issue of Clinical Infectious Diseases , Gottesman et al [12] describe the effect that a nationwide intervention restricting the use of quinolones in Israel had on the prevalence of resistance in Escherichia coli isolated from community urine samples. They noted a highly significant and rapid inverse relationship between the level of quinolone use and the prevalence of resistance in E. coli . The prevalence of resistance ranged from 14% to 9% during the highest and lowest [13] periods of quinolone use, with an average 1.16% decrease in resistance for each decrease of 1000 DDDs in quinolone use.
The study design used—interrupted time series—is a robust method for evaluating the effect that prescribing interventions have on resistance [13]. This intervention was triggered by the need to conserve quinolones for postexposure prophylaxis for an anticipated anthrax bioterrorism attack in October 2001. With respect to quinolone resistance, it is therefore a planned intervention. This is a considerable strength of the study, in comparison with an unplanned intervention that is prompted by rising levels of resistance. A decline in resistance after an unplanned intervention may be just regression to the mean, because the intervention occurred at the peak of the epidemic curve [13]. Moreover, the study is well designed, uses appropriate statistical analysis, and has a low risk of bias according to the Cochrane Effective Practice and Organisation of Care Group's criteria for interrupted time series [14].
The study has 2 weaknesses. First, it is an ecological study, meaning that data on prescribing were at the population rather than the individual level. However, ecological bias is likely to have weakened the association between prescribing and resistance, because previous studies in hospital and community populations have shown much stronger associations at the individual versus the population level [15, 16]. The second weakness is the potential for sampling bias, which is acknowledged by the authors.
What does Gottesman and colleagues' study add to our knowledge of the issue? It provides evidence that a reduction in quinolone use in the community may be followed by rapid decline in E. coli resistance. This is welcome, given the rather gloomy predictions of mathematical models [7–9]. However, we must remember George Box's famous statement that “all models are wrong but some are useful” [17]. There are high biological costs to the multiple mutations required to confer high-level quinolone resistance in E. coli [18]. Consequently, the new data may be consistent with the previously published models, and it will be interesting to find out.
These new data provide important support to efforts to reduce quinolone use in the community. In Scotland, reduction in seasonal variation in quinolone use in the community is now a national performance measure for health boards [19]. The rationale is that if antibiotic policies do not recommend quinolones for respiratory infection, then there should be no seasonal variation in prescribing [20]. The new data will be useful in modeling the potential impact that this intervention may have on resistance.
The new data should also have an important educational impact on physicians in primary care. A qualitative analysis of general practitioners in 2007 sought to understand views on resistance, whether resistance was important, whether they felt primary care was important in managing this, and whether resistance influenced their prescribing. Many of the general practitioners felt that there was little evidence linking their prescribing to resistance patterns and that there was little that they could achieve in light of other issues, such as secondary care prescribing and infection-control issues. While some of the general practitioners felt that their prescribing patterns might change if they had more information on local resistance patterns, they also felt that knowing about resistance would make them more likely to prescribe “stronger” broader spectrum and second-line antimicrobials (eg, quinolones). In summary, the study by Gottesman and colleagues provides important new evidence that quinolone resistance in E. coli is linked to community prescribing and supports efforts to reduce the unnecessary use of quinolones for respiratory or skin and soft-tissue infections caused by gram-positive bacteria, for which they are relatively ineffective.
Acknowledgments
Financial support. The European Surveillance of Antimicrobial Consumption project was supported by a grant from the Directorate General for Health and Consumer Affairs of the European Union.
Potential conflicts of interest. During the past 2 years, P.D. has received funding for research studies from Janssen-Cilag and Pfizer. L.U.: no conflicts.
- Received June 11, 2009.
- Accepted June 13, 2009.
- © 2009 by the Infectious Diseases Society of America
References
This Article
- ExtractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Classifications
-
- Articles and Commentaries
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 8.186
5-Yr impact factor: 7.898
For The Media
For Authors
- Services for authors
- Self-archiving Policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
