Treating respiratory tract infections in ambulatory care in Belgium: fluoroquinolone consumption and resistance development

International Journal of Antimicrobial Agents 26 (2005) 62 68 Treating respiratory tract infections in ambulatory care in Belgium: fluoroquinolone consumption and resistance development S. Simoens a,, J. Verhaegen b, G. Laekeman a, W.E. Peetermans c a Centre for Drug and Patient Information, Faculty of Pharmaceutical Sciences, K.U. Leuven, Edward van Evenstraat 4, 3000 Leuven, Belgium b Belgian Pneumococcal Reference Laboratory, Department of Microbiology, Faculty of Medicine, K.U. Leuven, Belgium c Department of Internal Medicine, Faculty of Medicine, K.U. Leuven, Belgium Received 23 December 2004; accepted 18 March 2005 Abstract This study analyses consumption patterns of fluoroquinolones in treating respiratory tract infections in ambulatory care in Belgium and describes susceptibility of Streptococcus pneumoniae isolates to fluoroquinolones. Consumption data were obtained from IMS Health. Pneumococcal resistance was investigated in 600 blood isolates collected from 1998 to 2003. Although consumption of fluoroquinolones has increased rapidly over the last decade, this trend does not seem to persist more recently. Fluoroquinolones were mainly used to treat urinary and lower respiratory tract infections, but rarely in the management of upper respiratory tract infections. The use of new fluoroquinolones (levofloxacin, moxifloxacin) and the ongoing use of older fluoroquinolones have not led to increased pneumococcal resistance, which remained below 1% for levofloxacin and was 0% for moxifloxacin. 2005 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Fluoroquinolones; Respiratory tract infections; Streptococcus pneumoniae; Consumption; Resistance 1. Introduction Respiratory tract infections are associated with significant morbidity and mortality. In its World Health Report 2004, the World Health Organization estimated that respiratory tract infections generated 94.6 disability-adjusted life-years lost worldwide and were the fourth major cause of mortality, responsible for 4.0 million deaths or 6.9% of the global number of deaths in 2002 [1]. Respiratory tract infections are also associated with substantial societal costs, in terms of direct costs (costs of medication, physician office visits, emergency room visits and hospitalisation) and indirect costs (costs of productivity losses) [2]. Antibiotics are commonly used in the treatment of respiratory tract infections in ambulatory care, even though spontaneous resolution is common and the use of antibiotics is not always warranted or associated with any significant benefit. There is evidence suggesting that two-third of patients Corresponding author. Tel.: +32 16 323 465; fax: +32 16 323 468. E-mail address: Steven.Simoens@pharm.kuleuven.ac.be (S. Simoens). presenting with acute sinusitis improve without any antibiotics within 2 weeks [3]. Systematic reviews published by the Cochrane Collaboration have found no benefit of antibiotic therapy in upper respiratory tract infections and only modest benefits in acute bronchitis [4 6]. This is because the cause of upper respiratory tract infections is generally viral and thus the use of antibiotics is not indicated. On the other hand, antibiotic therapy is recommended in patients with severe acute exacerbations of chronic obstructive pulmonary disease and community-acquired pneumonia [7]. Moreover, antibacterials appear effective in improving cure rates and decreasing duration of acute sinusitis, but only in a subset of patients who have a microbiological diagnosis of bacterial infection or severe disease [8]. In recent years, Belgian policy-makers have taken a special interest in the use of antibiotics for two main reasons. First, in an era of constrained budgets, rising consumption of antibiotics is contributing to increasing pharmaceutical expenditure. For instance, IMS Health data show that annual antibiotic consumption in Belgium valued at public prices has risen from D 210 million in 1993 to D 233 million in 1998. 0924-8579/$ see front matter 2005 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2005.03.010

S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 63 Second, treatment of respiratory tract infections is inhibited by development of resistance to common antibiotics. Data from the Belgian Pneumococcal Reference Laboratory indicate that resistance of Streptococcus pneumoniae to macrolides has increased from 5.2% in 1986 to 36.1% in 2003 in Belgium. Similarly, resistance has grown from 16.4% to 30.2% in the case of tetracyclines, and from 2% to 13% in the case of penicillin. Although international studies report a low level of resistance to fluoroquinolones, which are a new class of antibiotics with increased activity against Gram-positive bacteria including S. pneumoniae, increasing consumption of fluoroquinolones may raise resistance [9]. In response to these concerns, Belgian policy-makers have taken a number of measures to promote a more rational use of antibiotics. These included the creation in 1996 of Pharmanet, a database that brings together data regarding the prescribing behaviour of general practitioners (GPs). From late 1998 onwards, each individual GP received periodic updates of his/her individual drug prescription profile from Pharmanet, and this information was discussed in local peerreview groups. Additionally, three successive information campaigns were launched between 2000 and 2003 in which both physicians and patients were encouraged to use antibiotics responsibly. Practice guidelines for the appropriate use of antibiotics in ambulatory and hospital care have been developed and approved by CEBAM, the Belgian branch of the Cochrane Collaboration. More recently, in 2003 an agreement was struck between representatives of physicians and health insurance funds to make an increase in physician fees conditional on curbing the growth in the consumption of antibiotics, with particular emphasis on fluoroquinolones and broad-spectrum penicillins, as well as of angiotensin converting enzyme II inhibitors [10]. The aim of this article is to analyse the consumption of fluoroquinolones in the management of respiratory tract infections in ambulatory care in Belgium. To investigate whether the policy goals of containing consumption and resistance development have been met, this article analyses consumption patterns and susceptibility of blood isolates of S. pneumoniae to fluoroquinolones over time in Belgium. Such information is valuable as it can inform the prescription decision of physicians and can aid decision-makers in drafting policy measures that safeguard the value of fluoroquinolones and contain pharmaceutical expenditure. 2. Methods To explore consumption patterns, data on fluoroquinolones sold by wholesalers in Belgium were obtained from IMS Health. Both annual and monthly consumption data were available. Annual data ran from 1993 until 2003, and monthly data pertained to April 1999 to April 2004 (last month for which data were available at time of study). Annual figures were expressed as a moving annual total 1, meaning that, for instance, data for 1993 related to the period from February 1993 to January 1994. The analysis also made use of data from 2003 on consumption by type of infection for which fluoroquinolones were used (gastrointestinal tract infection, genital tract infection, urinary tract infection, upper and lower respiratory tract infection, and other infections). To collect such data, IMS Health invited a panel of 500 physicians to report the number of prescriptions that they issued for each type of infection. This concerns fluoroquinolone prescriptions to patients seen by physicians of all specialties offering general medical services, except for hospital inpatients. The composition of the panel of physicians corresponds to the distribution of the Belgian population of physicians by specialty and by region. When extrapolating the number of fluoroquinolone prescriptions by type of infection issued by the panel of 500 physicians to consumption of fluoroquinolones in the Belgian population, care was taken that the frequency of results adhered to the Gauss curve as closely as possible. As a breakdown by type of infection was not available for a number of generic drugs, the assumption was made that the breakdown by type of infection for the original drug also applied to the generic drug. All data were processed and analysed using Microsoft Excel. To study the issue of resistance development, the susceptibility of blood isolates of S. pneumoniae to ciprofloxacin, levofloxacin and moxifloxacin was tested. The S. pneumoniae strains were isolated in Belgian laboratories participating in the National Surveillance Programme. The strains were sent by post to the Belgian Pneumococcal Reference Laboratory on freshly inoculated blood agar without prior incubation. In the Reference Laboratory, the identity of isolates was first confirmed by examining the appearance of their colonies on blood agar and their susceptibility to optochin. Subsequently, the strains were typed by phase-contrast microscopy using Neufeld s reaction with 46 type or group sera obtained from the Statens Seruminstitut. For the purposes of this study, 100 blood isolates per year were selected at random from the 8525 pneumococci that the Reference Laboratory received from 1998 to 2003. Minimum inhibitory concentrations (MICs) were determined by the Etest (AB Biodisk). The same batch of Mueller Hinton agar (Oxoid, Basingstoke, UK) enriched with 5% horse blood was used for all strains. Inoculum of 0.5 McFarland standard was obtained by suspension of colonies in 0.5 ml Tryptic Soy Broth and checked by a densitometer. Inoculation of agar plates was done by Retro C80 and strips applied with the Simplex C76. Plates were incubated for 24 h at 36 Cin5%CO 2. MIC was the concentration of antibiotic that inhibited completely the growth of the pneumococcus. Streptococcus pneumoniae ATCC 49619 was used for quality control of the tests. The breakpoints used for levofloxacin and moxifloxacin were those defined by the National Committee for Clinical Laboratory Standards (NCCLS) for S. pneumoniae. The

64 S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 exception was ciprofloxacin, for which the breakpoints for Staphylococcus were used. 3. Results 3.1. Consumption of fluoroquinolones A first policy issue regarding the use of fluoroquinolones in Belgium concerns the trend in consumption levels over time. Is there evidence of a slowdown in the growth or even a decrease in the consumption of fluoroquinolones? Using annual data, Fig. 1 indicates that the volume of consumption of fluoroquinolones in ambulatory care in Belgium nearly doubled, from 1.69 defined daily doses per 1000 inhabitants daily (DID) in 1993 to 2.98 DID in 2003. The growth in consumption accelerated with the introduction of levofloxacin in 2000 and moxifloxacin in 2002. In the 1990s, consumption was dominated by norfloxacin, although this fluoroquinolone was increasingly replaced by ciprofloxacin and ofloxacin in the second half of the 1990s. Consumption of pefloxacin amounted to 0.02 DID in 1993, but this fluoroquinolone has no longer been commercialised in Belgium in recent years. In the early 2000s, consumption of norfloxacin and ciprofloxacin continued to decrease and rise, respectively, while the consumption of ofloxacin dropped. Following a rapid increase in consumption of levofloxacin to a peak of 0.66 DID in 2001, the volume of consumption subsequently fell to 0.31 DID in 2003. Looking at annual consumption of fluoroquinolones valued at public prices, this increased from D 24.1 million in 1993 to a maximum of D 44.4 million in 2002, and then decreased to D 43.1 million in 2003. Despite this substantial increase in consumption levels of fluoroquinolones in the last decade, more recent evidence suggests that this trend may be coming to an end. Monthly data indicate that the volume of consumption of fluoroquinolones in ambulatory care in Belgium has stagnated: it amounted to an average of 0.25 DID during the period May 2002 to April 2003, and to an average of 0.24 DID during May 2003 to April 2004. Additionally, for each of the first 4 months of 2004, the volume of consumption of fluoroquinolones was lower than the consumption in the same month of 2003: 0.29 DID versus 0.24 DID in January; 0.27 DID versus 0.21 DID in February; 0.26 DID versus 0.25 DID in March; and 0.25 DID versus 0.24 DID in April. To gain a better insight into the market of fluoroquinolones, the impact of the introduction of new fluoroquinolones in Belgium on consumption patterns is examined. Fig. 2 presents monthly data on the volume of consumption of levofloxacin (Tavanic ; introduced in July/August 2000) and moxifloxacin (Avelox and Proflox ; both introduced in April 2002). Focusing on the first 2 years since its introduction, the volume of consumption of levofloxacin has remained constant at an average of 0.05 DID during August 2000 to July 2001 and during August 2001 to July 2002. Since April 2002, there was a trend towards not only substitution of moxifloxacin for levofloxacin, but also growth in the combined volume of consumption of these two fluoroquinolones as a result of the introduction of moxifloxacin. There was no change in the volume of consumption of moxifloxacin since Fig. 1. Volume of consumption of fluoroquinolones in ambulatory care in Belgium (in defined daily doses per 1000 inhabitants daily (DID)). Source: IMS Health.

S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 65 Fig. 2. Volume of consumption of levofloxacin and moxifloxacin in ambulatory care in Belgium (in defined daily doses per 1000 inhabitants daily (DID)). Source: IMS Health. its introduction, amounting to an average of 0.06 DID during April 2002 to March 2003 and during April 2003 to March 2004. The pattern of seasonal variation observed for levofloxacin prior to the introduction of moxifloxacin almost disappeared after its introduction. On the other hand, consumption of moxifloxacin exhibited a typical pattern of seasonal variation. This may reflect the fact that levofloxacin was used for treating respiratory tract infections prior to the introduction of moxifloxacin, and for treating gastrointestinal and urinary tract infections after its introduction. Moxifloxacin, a fluoroquinolone typically used in the treatment of respiratory tract infections, followed the pattern of seasonal variation that would be expected. Annual data collected by the European Surveillance of Antimicrobial Consumption Project show that the Belgian trend of growing use of fluoroquinolones over time is also seen in other European countries [11]. Moreover, among a sample of 26 European countries, Belgium had the third highest consumption of fluoroquinolones, after Italy and Portugal. The volume of consumption of fluoroquinolones ranged from 0.17 DID in Denmark to 3.76 DID in Italy in 2002. Focusing on new fluoroquinolones, the consumption of levofloxacin and moxifloxacin of 1.3 DID in Belgium was only surpassed by that in Italy in 2002. 3.2. Breakdown of consumption by type of infection A breakdown of the volume of consumption of fluoroquinolones by type of infection in Belgium in 2003 is portrayed in Fig. 3. Fig. 3 shows that the principal use of fluoroquinolones was in the treatment of urinary tract infections (35% of consumption; volume of 1.06 DID) and lower respiratory tract infections (24% of consumption; volume of 0.71 DID). Management of upper respiratory tract infections with fluoroquinolones amounted to 9% of consumption (volume Fig. 3. Consumption of fluoroquinolones by type of infection in Belgium in 2003 (in defined daily doses per 1000 inhabitants daily (DID)). Source: IMS Health. Note: Fluoroquinolones marketed in Belgium are ciprofloxacin, levofloxacin, moxifloxacin, norfloxacin, ofloxacin and pefloxacin.

66 S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 of 0.26 DID) in 2003. This suggests that fluoroquinolones are primarily used in those indications where they have been shown to yield clinical benefit [12,13]. Although the use of antibiotics in treating upper respiratory tract infections is more controversial than their use in managing lower respiratory tract infections [3 8], the data indicate that the volume of antibiotic consumption in the treatment of upper respiratory tract infections (9.71 DID) exceeded the volume of antibiotics used in managing lower respiratory tract infections (7.51 DID) in Belgium in 2003. If policy-makers wish to discourage the use of antibiotics in the treatment of upper respiratory tract infections, which antibiotics do they need to target? In Belgium, upper respiratory tract infections were mainly managed by broadspectrum penicillins (55% of consumption in 2003; volume of 5.34 DID), macrolides and cephalosporins (17% of consumption; volume of 1.70 DID each). Fluoroquinolones were used marginally in managing upper respiratory tract infections, with a share of consumption of 3% (volume of 0.26 DID). This suggests that policy needs to focus on encouraging a more responsible use of broad-spectrum penicillins, macrolides and cephalosporins in the treatment of upper respiratory tract infections. 3.3. Susceptibility of S. pneumoniae to fluoroquinolones To quantify the level of resistance of S. pneumoniae and to investigate whether there is any evolution in resistance levels in Belgium over time, a sample of 100 blood isolates in each year from 1998 to 2003 was tested. Table 1 shows the cumulative MICs of the different antibiotics for the six different years. Using breakpoint concentrations for ciprofloxacin of staphylococci, the percentage of susceptible pneumococci fluctuated between 64% and 91%. However, a tendency of increase or decrease in susceptibility was not observed. Between 1998 and 2003, no evolution in the distribution of the MICs was detected for levofloxacin and moxifloxacin. The overall in vitro activities for the three fluoroquinolones, the MIC distributions of all strains, and their MIC 50, MIC 90 and susceptibility rates are presented in Table 2. Moxifloxacin had a five-fold lower mode MIC value than ciprofloxacin and levofloxacin. No pneumococcus strains resistant to moxifloxacin were detected, but 5 (0.8%) had MICs of 4 mg/l and 151 (26%) had MICs >1 mg/l and <4 mg/l for ciprofloxacin, and 4 (0.6%) had intermediate susceptibility to levofloxacin. Table 3 shows the cumulative MICs of levofloxacin and moxifloxacin for the ciprofloxacin intermediate susceptible and resistant isolates. All pneumococci not susceptible to ciprofloxacin were inhibited by a concentration of 3 mg/l of levofloxacin and 0.5 mg/l of moxifloxacin. The 156 pneumococci with MICs 1 mg/l for ciprofloxacin belong to the following capsular types: 19 (14.7%), 4 (10.4%), 14 (10.4%), 7 (9.1%), 9 (9.1%), 3 (9.1%), 1 (7.1%), 6 (5.8%) and 23 (5.8%). The remaining 18.5% of isolates Table 1 Cumulative proportions of 600 pneumococci inhibited by the different antibiotics at various concentrations Ciprofloxacin (mg/ml) a 0.19 0.25 0.38 0.5 0.75 1 1.5 2 3 4 1998 0 2 9 37 71 83 94 98 100 1999 1 5 14 33 57 81 93 99 100 2000 1 1 5 25 44 70 93 99 99 100 2001 1 2 4 18 38 69 85 94 98 100 2002 3 10 18 40 73 91 98 100 2003 1 2 9 16 35 64 92 98 100 Levofloxacin (mg/ml) 0.38 0.5 0.75 1 1.5 2 3 4 1998 0 16 56 90 98 100 1999 1 13 49 87 97 99 100 2000 0 5 28 73 94 99 100 2001 0 4 26 71 96 99 100 2002 0 17 49 94 99 99 100 2003 0 1 20 74 99 100 100 Moxifloxacin (mg/ml) 0.064 0.094 0.12 0.19 0.25 0.38 0.5 1998 0 13 72 92 98 100 1999 2 20 59 92 99 100 2000 1 6 52 84 97 98 100 2001 1 4 46 89 100 2002 5 22 73 96 100 2003 0 8 52 92 100 Source: Belgian Pneumococcal Reference Laboratory. Note: The percentage of susceptible isolates at breakpoint concentrations are shown in italics. a Breakpoints for Staphylococcus. not susceptible to ciprofloxacin belong to 14 other capsular types; but each of these types was found in less than 5% of the isolates. It is important to notice that capsular types 4 and 7 are responsible for 10.4% and 9.1%, respectively, of the ciprofloxacin non-susceptible isolates, while both types represent only 6.3% and 6.2%, respectively, of the 600 examined isolates. 4. Discussion This study has analysed two policy issues relating to the use of fluoroquinolones in the treatment of respiratory tract infections in ambulatory care in Belgium. To investigate claims of a burgeoning use, consumption patterns of fluoroquinolones were analysed in Belgium over time. Also, as injudicious use of antibiotics may foster resistance, a sample of 600 blood isolates was investigated to quantify susceptibility of S. pneumoniae to fluoroquinolones. Such information is needed to inform national policies for optimising the use of fluoroquinolones and to safeguard their value. Although consumption of fluoroquinolones has increased rapidly over the last decade, this trend does not seem to persist in more recent times. The volume of consumption in the

S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 67 Table 2 Overall susceptibility of 600 pneumococci to ciprofloxacin, levofloxacin and moxifloxacin MIC (mg/l) 0.064 0.094 0.12 0.19 0.25 0.38 0.5 0.75 1 1.5 2 3 4 Ciprofloxacin 0 0 0 7 13 30 89 143 162 98 42 11 5 Levofloxacin 0 0 0 0 0 1 55 172 261 94 13 4 Moxifloxacin 9 64 281 191 49 4 2 MIC 50 MIC 90 Range %S %I %R Ciprofloxacin 1 1.5 0.19 4 73.2 26 0.8 Levofloxacin 1 1.5 0.38 3 99.4 0.6 Moxifloxacin 0.12 0.19 0.064 0.5 100 Source: Belgian Pneumococcal Reference Laboratory. MIC, minimum inhibitory concentration; %S, %I, %R, percentage of isolates susceptible, intermediate resistant or resistant following National Committee for Clinical Laboratory Standards (NCCLS) breakpoints. year leading up to April 2004 was at the same level as that of the preceding year. In the first 4 months of 2004, the volume of consumption was lower than in the first 4 months of 2003. Additionally, the level of consumption of fluoroquinolones reached a maximum in 2002 and has fallen in 2003. It is important to follow-up consumption patterns of fluoroquinolones in future years to analyse whether this new trend continues. A major issue concerns the appropriate use of antibiotics in treating respiratory tract infections. Clinical evidence suggests that antibiotic therapy can yield clinical benefits in treating lower respiratory tract infections, but it is not recommended for treating upper respiratory tract infections, except for in particular subgroups of patients [3 8]. A breakdown of consumption by type of infection revealed that fluoroquinolones are marginally used in the treatment of upper respiratory tract infections and are mainly employed to treat urinary tract infections and lower respiratory tract infec- Table 3 Cumulative proportions of strains with minimum inhibitory concentrations (MICs) of levofloxacin and moxifloxacin for pneumococci not susceptible to ciprofloxacin N Levofloxacin 0.38 0.5 0.75 1 1.5 2 3 4 1998 27 0 0 15 81 93 100 1999 19 0 0 5 47 84 100 2000 30 0 0 0 40 87 97 100 2001 31 0 0 3 38 86 96 100 2002 9 0 0 0 89 100 2003 36 0 0 0 50 94 100 N Moxifloxacin 0.064 0.094 0.12 0.19 0.25 0.38 0.5 1998 27 0 0 30 85 98 100 1999 19 0 5 37 69 90 100 2000 30 0 0 30 67 94 97 100 2001 31 0 0 16 48 100 2002 9 0 0 33 100 2003 36 0 0 36 86 100 Source: Belgian Pneumococcal Reference Laboratory. tions. Policy needs to focus on achieving a more responsible use of penicillins, macrolides and cephalosporins, as these antibiotics are commonly used in the treatment of upper respiratory tract infections. Fluoroquinolones inhibit bacterial DNA synthesis by inhibiting DNA gyrase and topoisomerase IV [14]. Fluoroquinolone resistance is generally related to mutations in the quinolone resistance-determining region (QRDR). In S. pneumoniae, resistance to the new fluoroquinolones generally requires at least two mutation events, one in DNA gyrase and one in topoisomerase IV. This study has investigated fluoroquinolone resistance in blood isolates rather than in noninvasive isolates. Data on the activity of various antibiotics, including fluoroquinolones, against non-invasive isolates of S. pneumoniae pointed to similar resistance rates in noninvasive isolates as those observed in our blood isolates taken from the Belgian Pneumococcal Reference Laboratory [15]. The in vitro results of this and other studies [16] demonstrate that ciprofloxacin cannot be considered an anti-pneumococcal fluoroquinolone and has an activity in the range 0.38 4 mg/l. Levofloxacin maintains activity against ciprofloxacin intermediate or resistant isolates, with MIC 90 values of 1.5 mg/l. Moxifloxacin is the most potent fluoroquinolone available for treatment of lower respiratory tract infections in Belgium, with a MIC 90 of 0.19 mg/l. The use of new fluoroquinolones (levofloxacin, moxifloxacin) and the ongoing use of the older fluoroquinolones (mainly for urinary tract infections) has not led, to date, to an increase in the rate of pneumococcal resistance to fluoroquinolones. It remains below 1% for levofloxacin and is 0% for moxifloxacin. This mirrors the resistance rate of S. pneumoniae to levofloxacin of 0.8% observed in the US-based component of the PROTEKT study in 2000 2001 [17]. In our multi-year surveillance study, a rightward shift in the distributions of MICs of the fluoroquinolones was not observed. This finding also suggests that first-step mutants have not become more prevalent in Belgian strains. These findings do not offer clear guidance regarding the appropriate setting for use of fluoroquinolones. On the one hand, data from the Belgian Pneumococcal Reference Labo-

68 S. Simoens et al. / International Journal of Antimicrobial Agents 26 (2005) 62 68 ratory indicate that resistance of S. pneumoniae to macrolides and tetracyclines in Belgium is relatively high and is rising over time, and thus a case can be made for switching to anti-pneumococcal fluoroquinolones as an alternative to penicillins, with our findings indicating that fluoroquinolones continue to constitute a class of potent agents against S. pneumoniae. On the other hand, it could be argued that the use of fluoroquinolones needs to be restricted in order to contain resistance development and to safeguard the value of this class of antibiotics. Further research needs to be carried out to attain consensus on the optimal role of fluoroquinolones in the treatment of lower respiratory tract infections in ambulatory care. Finally, in the agreement struck between representatives of physicians and health insurance funds in 2003, particular emphasis was placed on curbing the rising consumption of fluoroquinolones for purposes of containing costs and inhibiting resistance development [10]. One of the instruments available to policy-makers to achieve this objective is to put more stringent reimbursement conditions on fluoroquinolones. However, such an approach may only produce a switch in physician prescribing behaviour away from fluoroquinolones to broad-spectrum penicillins and other antibiotics, rather than have an impact on overall consumption. Instead of focusing on individual classes of antibiotics, an emphasis on particular indications is advocated here. A more valuable approach may consist of identifying those indications, such as upper respiratory tract infections, where the use of antibiotics is not recommended and to introduce policy measures such as clinical guidelines, peer-review with feedback, educational campaigns or financial incentives to discourage the use of antibiotics in the treatment of those indications. Acknowledgments Financial support for this research was received from Bayer HealthCare Pharmaceuticals. The authors have no conflicts of interest that are directly relevant to the content of this manuscript. The authors would like to express their gratitude to Eric Mostrey (market research, Bayer) for his assistance. References [1] World Health Organization. The World Health Report 2004 changing history. Geneva, Austria: WHO; 2004. p. 120, 126. [2] Bishai W. Current issues on resistance, treatment guidelines, and the appropriate use of fluoroquinolones for respiratory tract infections. Clin Ther 2002;24:838 50. [3] Piccirillo JF, Mager DE, Frisse ME, Brophy RH, Goggin A. Impact of first-line vs second-line antibiotics for the treatment of acute uncomplicated sinusitis. JAMA 2001;286:1849 56. [4] Becker L, Glazier R, McIsaac W, Smucny J. Antibiotics for acute bronchitis. The Cochrane Library, Issue 3. Chichester, UK: John Wiley & Sons Ltd.; 1999. [5] Arroll B, Kenealy T. Antibiotics for the common cold and acute purulent rhinitis. The Cochrane Library, Issue 4. Chichester, UK: John Wiley & Sons Ltd.; 2003. [6] Smucny J, Fahey T, Becker L, Glazier R. Antibiotics for acute bronchitis. The Cochrane Library, Issue 4. Chichester, UK: John Wiley & Sons Ltd.; 2003. [7] Feldman C. Appropriate management of lower respiratory tract infections in primary care. Prim Care Respir J 2004;13: 159 66. [8] Wasserfallen JB, Livio F, Zanetti G. Acute rhinosinusitis: a pharmacoeconomic review of antibacterial use. Pharmacoeconomics 2004;22:829 37. [9] Pallares R, Fenoll A, Linares J. The Spanish Pneumococcal Infection Study Network. The epidemiology of antibiotic resistance in Streptococcus pneumoniae and the clinical relevance of resistance to cephalosporins, macrolides and quinolones. Int J Antimicrob Agents 2003;22:S15 24. [10] Rijksinstituut voor Ziekte- en Invaliditeitsverzekering [National Social Security Institute]. Ambulant voorschrijfgedrag antibiotica en antihypertensiva. Akkoord artsen-ziekenfondsen 2004 2005. [Prescribing behaviour of antibiotics and antihypertensive drugs in ambulatory care. Agreement physicians health insurance funds 2004 2005]. Brussels, Belgium: Rijksinstituut voor Ziekte- en Invaliditeitsverzekering; 2004. [11] Goossens H, Ferech M, Vander Stichele R, Elseviers M, for the ESAC Project Group. Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet 2005;365:579 87. [12] Ball P. Efficacy and safety of levofloxacin in the context of other contemporary fluoroquinolones: a review. Curr Ther Res 2003;64: 646 61. [13] Bailey RR. Quinolones in the treatment of uncomplicated urinary tract infections. Int J Antimicrob Agents 1992;2:19 28. [14] Hooper DC. Mechanisms of action of antimicrobials: focus on fluoroquinolones. Clin Infect Dis 2001;32(Suppl. 1):S9 15. [15] Vanhoof R, Van Eldere J, Verhaegen J, for the Streptococcus Pneumoniae Multicentre Study Group. Activity of various antimicrobials against non-invasive clinical isolates of Streptococcus pneumoniae collected during winter 2002 2003. Poster presented at the RICAI (Interdisciplinary Meetings on Chemotherapy and Infection) Conference, Paris, 2003. [16] Boswell FJ, Andrews JW, Jevans G, Wise R. Comparison of the in vitro activities of several new fluoroquinolones against respiratory pathogens and their abilities to select fluoroquinolone resistance. J Antimicrob Chemother 2002;50:495 502. [17] Rybak MJ. Increased bacterial resistance: PROTEKT US an update. Ann Pharmacother 2004;38:S8 13.