Journal of Hospital Infection

Journal of Hospital Infection xxx (2011) 1e5 Available online at www.sciencedirect.com Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin Imipenem and ciprofloxacin consumption as factors associated with high incidence rates of resistant Pseudomonas aeruginosa in hospitals in northern France K. Miliani a,f.l Hériteau a, L. Lacavé a, A. Carbonne a, P. Astagneau a,b, * on behalf of the Antimicrobial Surveillance Network Study Group 1 a Regional Coordinating Centre for Nosocomial Infection Control (C-CLIN Paris-Nord), Paris, France b Department of Public Health, Pierre et Marie Curie University School of Medicine, Paris, France article info summary Article history: Received 11 June 2010 Accepted 26 November 2010 Available online xxx Keywords: Antimicrobial resistance Ciprofloxacin consumption France Hospital antimicrobial consumption Imipenem consumption Resistant P. aeruginosa In France, Pseudomonas aeruginosa is the third most common isolate from nosocomial infections. To determine whether high consumption rates of some antibiotics are risk factors for resistance to ceftazidime, imipenem, ciprofloxacin and amikacin in P. aeruginosa, we conducted a study based on data from the Antimicrobial Surveillance Network in northern France and the French public reporting system of infection control indicators. These data were related to hospital characteristics (size, type, proportion of non-acute care beds), antibiotic consumption, incidence rates of some key resistances and quality indicators of healthcare-associated infection (HAI) control. In univariate analysis, high total and specific antibiotic consumption (except amoxicillin/clavulanate and imidazoles) were associated with high P. aeruginosa resistance rates. In multivariate analysis, high resistance rates were related to high imipenem and ciprofloxacin consumption [odds ratio (95% confidence interval): 7.9 (2.24e28.09), P < 0.05 for both], but were not significantly related to quality indicators of HAI control. These findings suggest that imipenem and ciprofloxacin use could play a major role in driving P. aeruginosa resistance, independent of other infection control performance. Ó 2011 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. Introduction In France, Pseudomonas aeruginosa is the third most frequently isolated micro-organism from nosocomial infections. 1 This pathogen, naturally resistant to several antimicrobials, has a remarkable capacity for acquiring new resistance mechanisms under selective antibiotic pressure. 2,3 Resistance can result in treatment failure, increased morbidity and mortality (especially in critically ill patients), prolonged hospitalisation and higher healthcare costs. 4,5 Antimicrobial use is a risk factor for the development of antimicrobial resistance, but the relationship is complex. 2,3,6 From an ecological perspective, aggregate hospital data on antibiotic use * Corresponding author. Address: C-CLIN Paris-Nord, Site Broussais, Pavillon Leriche, 3e ETG, 96 rue Didot, 75014 Paris, France. Tel.: þ33 1 40 27 42 00; fax: þ33 140274217. E-mail address: pascal.astagneau@sap.aphp.fr (P. Astagneau). 1 Members: S. Alfandari, K. Blanckaert, C. Bonenfant, E. Bouvet, A. Chalfine, Y. Costa, E. Delière, F. Espinasse, N. Fortineau, Z. Kadi, J-L. Schmit, P. Votte. and bacterial resistance rates could be useful for researchers who study this relationship, and such data may be obtained as part of local or regional surveillance efforts. 6e9 Hence, as part of a multicentre hospital-based surveillance system in northern France, we conducted a study to determine whether high consumption rates of some antibiotics could be risk factors for high incidence rates of resistance by P. aeruginosa to ceftazidime, imipenem, ciprofloxacin and amikacin. Methods The study was based on data from the Antimicrobial Surveillance Network in northern France (described elsewhere 9 ) and the French public reporting system of infection control indicators (available online at URL: http://www.icalin.sante.gouv.fr) in 2007. The antimicrobial surveillance network, coordinated by the regional centre for nosocomial infection control in northern France (C-CLIN Paris Nord), was set up in 2002 with voluntary participation of healthcare facilities (HCFs). The main objective of the 0195-6701/$ e see front matter Ó 2011 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jhin.2010.11.024

2 K. Miliani et al. / Journal of Hospital Infection xxx (2011) 1e5 network is to enable HCFs to monitor both antimicrobial consumption and antimicrobial resistance rates. It serves as a comparison tool allowing HCFs to recognise problems and improve the way in which antimicrobials are used. Furthermore, in 2005 the French Ministry of Health implemented an infection control indicators public reporting system for all HCFs in France as part of the French National Program for Prevention of Healthcare- Associated Infections (HAI) and Antimicrobial Resistance. 10 Five indicators were defined by a task group in 2003 and were progressively released through a multistep procedure of expert evaluation and validation. These indicators refer to structures, procedures and results in the field of HAI control. 10 Data collection Data regarding hospital characteristics and administrative data, antimicrobial consumption and incidence rates of resistant P. aeruginosa (ceftazidime-, imipenem-, ciprofloxacin- and amikacin-resistant) were obtained from the Antimicrobial surveillance Network for the 2007 surveillance period. These resistances were chosen in order to include one member of each of the four main classes of antipseudomonal agent. Data were retrospectively collected by the network with a standard protocol for all volunteer HCFs. In addition, for each participating HCF, the score attributed to three quality indicators of HAI control were extracted from the French public reporting system of infection control indicators for the 2007 year. Hospital characteristics and administrative data were captured regarding hospital type, hospital ownership and teaching status, type and number of beds and number of patient-days. Hospital-wide antimicrobial consumption of amoxicillin, amoxicillin/clavulanate, piperacillin/tazobactam, ceftazidime (CAZ), other third-generation cephalosporins, imipenem (IPM), ciprofloxacin (CIP), other fluoroquinolones, amikacin (AMK), other aminoglycosides, glycopeptides, macrolides, imidazoles and total consumption were collected and expressed in defined daily doses (DDDs) per 1000 patient-days according to the 2007 version of the Anatomical Therapeutic ChemicaleDDD classification from the World Health Organization. Data on bacterial resistance were provided by the laboratories serving the HCFs participating in the survey. Data included information on the number of isolates that were non-susceptible (intermediately susceptible and resistant) and the number of isolates tested for antimicrobial susceptibility from various consecutive clinical isolates (excluding those from stool cultures) of nosocomial P. aeruginosa. Antimicrobial susceptibility was interpreted in accordance with criteria recommended by the Antibiogram Committee update 2007 of the French Society for Microbiology (SFM). 11 We focused on P. aeruginosa resistant to CAZ, IPM, CIP, or AMK. The incidence rates of these resistances were expressed as the number of non-susceptible (intermediately susceptible and resistant) isolates per 1000 patient-days. Duplicate isolates from the same person were excluded. The quality indicators of HAI control used in our study were: (i) the composite score (scale: 0e100) measuring the development of the infection control structures and activities against nosocomial infections (ICALIN); (ii) the composite score (scale: 0e20) related to organisation of antibiotic use policy, the dedicated resources and activities (ICATB); and (iii) the percentage of target volume of hydro-alcoholic hand-rub products used in 2007 (in litres) per 1000 patient-days (ICSHA). 10 Data analysis Analysis focused only on HCFs that had antimicrobial susceptibility data for more than ten P. aeruginosa isolates. All variables were considered as categorical. The dependent variable has been dichotomised as: (1) HCFs with no resistance rates that exceed the 75th percentile (p75); and (2) HCFs with at least one resistance rate exceeding p75. Antibiotic consumption variables were divided into two classes according to 75th percentile (p75) in p75 or >p75 of overall distribution. The HCFs were classed according to type (public non-teaching, public teaching, private non-profit, private for-profit), size (<300 or 300 beds) and the proportion of nonacute care beds (i.e. the proportion of beds belonging to long term care, rehabilitation units or psychiatric wards in two classes <25% or 25%). The quality indicators were categorised according to the 25th percentile (p25) in p25 or >p25 of overall distribution. Univariate analysis was performed using Pearson c 2 -test and univariate logistic regression to obtain the crude odds ratios (ORs). Variables for which P 0.10 in the univariate analysis were included in a multivariate logistic regression model. A manual backward-stepwise variable-selection procedure was used. Variables remained in the multivariate model if the likelihood ratio test (e2ll) was significant (P 0.05). Akaike s information criterion (AIC) and the area under the receiver operating characteristic curve (ROC) were determined as measure of fit and discriminative accuracy, respectively. Statistical analysis was carried out using Stata statistical software, release 10.1 (Stata Corp., College Station, TX, USA). Results From 102 HCFs that participated in both surveys (antimicrobial consumption and antimicrobial resistance) in 2007, 84 tested more than 10 P. aeruginosa isolates and were included in the analysis. More than half (59.5%) of these HCFs were public, and among them 11 (13.1%) had teaching status, whereas 34 (40.5%) were private HCFs. Forty-three (51.2%) had <300 beds and 57 (67.9%) had 25% non-acute care beds. Most HCFs with <300 beds were private (69.8%), whereas almost all (90.2%) HCFs with 300 beds were public (teaching and non-teaching). Likewise, most HCFs with <25% of non-acute care beds were private (70.4%), whereas a similar percentage (73.7%) of HCFs with 25% non-acute care beds were public. A total of 13 993 P. aeruginosa isolates were tested for antimicrobial susceptibility in the 84 HCFs included in analysis (median: 107 isolates per HCF; maximum: 2190). The numbers of isolates resistant to CAZ, IPM, CIP and AMK were 2462, 2967, 4253 and 2691, respectively. The median and interquartile ranges (IQR) of resistance rates are shown in Table I. Because aggregate data were obtained from each HCF, no information was available to determine whether isolates were multidrug resistant. Thirty (35.7%) HCFs had at least one key resistance rate >p75: seven (8.3%) HCFs had only one key resistance rate >p75; five (6.0%) had two key resistance rates >p75; five (6.0%) had three key resistance rates >p75 and 13 (15.4%) had all four key resistance rates >p75. The median overall antibiotic consumption was 428.8 DDDs/ 1000 patient-days (IQR: 291.7e531.0). The highest levels were observed for penicillins and fluoroquinolones, with median consumptions of 229.9 (150.8e305.6) and 53.5 (39.6e75.3) respectively. Amoxicillin and amoxicillin/clavulanate were the predominant penicillins, with median consumptions of 75.2 (42.9e102.8) and 138.3 (85.4e191.3) respectively. Table I shows the median and IQR of antibiotic used for the comparisons as well as the median and IQR of the three quality indicators of HAI. In the univariate analysis (Table II), an incidence rate >p75 for at least one key P. aeruginosa resistance was associated (P < 0.10) with the type (public teaching hospitals and private non-profit) and structure (<25% of non-acute care beds) of HCFs, and with total and specific antibiotic consumption >p75 (except for amoxicillin/

K. Miliani et al. / Journal of Hospital Infection xxx (2011) 1e5 3 Table I Characteristic of study population (N ¼ 84 hospitals) Hospital characteristics Public non-teaching hospitals 39 (46.4%) Public teaching hospitals 11 (13.1%) Private non-profit 17 (20.25%) Private for-profit 17 (20.25%) Hospital size <300 beds 43 (51.2%) 300 beds 41 (48.8%) Proportion of non-acute care beds <25% 27 (32.1%) 25% 57 (67.9%) Quality indicators of HAI control a 2007 ICALIN score (range: 0e100) 98.0 (93.3e99.8) 2007 ICATB score (range: 0e20) 15.4 (12.5e17.4) 2007 ICSHA score (range: 0e100) 70.7 (49.6e93.5) Antimicrobial consumption (DDDs/1000 patient-days) a Amoxicillin 75.2 (42.9e102.8) Amoxicillin/clavulanate 138.3 (85.4e191.3) Piperacillin/tazobactam 1.4 (0.1e3.6) Ceftazidime 1.7 (0.4e3.6) Other third generation cephalosporin 15.5 (8.7e25.9) Imipenem 1.7 (0.5e3.8) Ciprofloxacin 10.6 (5.9e21.6) Other fluoroquinolones 40.8 (22.5e53.0) Amikacin 3.0 (0.9e5.9) Other aminoglycosides 7.0 (2.9e14.4) Glycopeptides 4.5 (1.6e8.3) Macrolides 8.5 (4.3e15.4) Imidazoles 12.6 (5.5e21.5) Total consumption 428.8 (291.7e531.0) Bacterial resistance rates (no. of non-susceptible isolates/1000 patient-days) a Ceftazidime resistance 0.15 (0.09e0.27) Imipenem resistance 0.11 (0.06e0.33) Ciprofloxacin resistance 0.26 (0.17e0.50) Amikacin resistance 0.15 (0.09e0.33) HAI, healthcare-associated infection; ICALIN, composite score measuring the development of the infection control structures and activities against nosocomial infections; ICATB, composite score related to organisation of antibiotic use policy; ICSHA, indicator of overall consumption of hydro-alcoholic hand-rub products; DDDs, defined daily doses. a Median (interquartile range). clavulanate and imidazoles, P ¼ 0.19 for each). Conversely, HCFs with an ICATB or ICSHA score <p25 were less likely to have high (>p75) resistance rates (P < 0.10). There were no associations with hospital size or ICALIN score <p25 (P ¼ 0.87 and P ¼ 0.79 respectively). In the multivariate analysis (Tables III and IV), after adjusting for the effect of the other variables included in the model (P < 0.10), high incidence rates (>p75) for at least one of the target resistances were related to high (>p75) IPM and CIP consumption (OR: 7.9; 95% confidence interval: 2.24e28.09; P < 0.05 for both). By contrast, P. aeruginosa resistance was not significantly related to low (<p25) alcohol hand-rub consumption ICSHA score (P ¼ 0.30) nor to a low (<p25) score for the quality indicator related to organisation of antibiotic use policy ICATB score (P ¼ 0.32). Discussion Our study found that high CAZ, CIP, IPM, or AMK resistance rates in P. aeruginosa were independently related to high IPM and CIP consumption. The emergence of antibiotic-resistant P. aeruginosa can be related to the overuse of several antimicrobial agents, although the risk appears to differ with different agents. 13e15 However, prior use of fluoroquinolones or IPM has been more Table II Univariate analysis: antibiotic agents and hospital characteristics related to high rates (>75th percentile) of at least one key resistance (ceftazidime, imipenem, ciprofloxacin or amikacin) among Pseudomonas aeruginosa isolates (N ¼ 84 hospitals) Variables At least one key resistance a with incidence rate >p75 (N ¼ 30) b Crude OR (95% CI) Antibiotic consumptions Piperacillin/tazobactam p75 (n ¼ 64) d 16 (25.0) >p75 (n ¼ 20) 14 (70.0) 7.0 (2.30e21.27) <0.001 Ceftazidime p75 (n ¼ 63) d 15 (23.8) >p75 (n ¼ 21) 15 (71.4) 8.0 (2.64e24.28) <0.001 Other 3rd generation Cephalosporins p75 (n ¼ 63) d 18 (28.6) >p75 (n ¼ 21) 12 (57.1) 3.3 (1.20e9.27) 0.02 Imipenem p75 (n ¼ 63) d 14 (22.2) >p75 (n ¼ 21) 16 (76.2) 11.2 (3.49e35.97) <0.001 Ciprofloxacin p75 (n ¼ 63) d 14 (22.2) >p75 (n ¼ 21) 16 (76.2) 11.2 (3.49e35.97) <0.001 Amikacin p75 (n ¼ 63) d 17 (27.0) >p75 (n ¼ 21) 13 (61.9) 4.4 (1.55e12.46) 0.004 Glycopeptides p75 (n ¼ 63) d 17 (27.0) p75 (n ¼ 21) 13 (61.9) 4.4 (1.55e12.46) 0.004 Total consumption p75 (n ¼ 63) d 16 (25.4) >p75 (n ¼ 21) 14 (66.7) 5.9 (2.02e17.13) 0.001 Hospital characteristics Public non-teaching (n ¼ 39) d 8 (20.5) Reference Public teaching (n ¼ 11) 6 (54.6) 4.7 (1.13e19.21) Private non-profit (n ¼ 17) 11 (64.7) 7.1 (2.01e25.10) 0.007 Private for-profit (n ¼ 17) 5 (29.4) 1.6 (0.44e5.93) Proportion of non-acute care beds <25% (n ¼ 27) 14 (51.9) 2.8 (1.071e7.14) 0.03 25% (n ¼ 57) d 16 (28.1) ICATB 2007 p25 (n ¼ 25) 5 (20.0) 0.3 (0.11e1.03) 0.05 p25 (n ¼ 59) d 25 (42.4) OR, odds ratio; CI, confidence interval; ICATB, composite score related to organisation of antibiotic use policy. Only variables significant at P 0.05 are shown. a Ceftazidime, imipenem, ciprofloxacin, or amikacin resistance among Pseudomonas aeruginosa isolates. b Data are expressed as number and percentage, n (%). c Pearson c 2 -test. d Reference class. frequently associated with P. aeruginosa resistance. In a cohort study, Carmeli et al. 12 found that IPM use was associated with emergence of resistance to IPM, CIP, CAZ and piperacillin/tazobactam in P. aeruginosa. In addition, piperacillin/tazobactam and CIP were related to the emergence of resistance to themselves, whereas CAZ was not. 12 Lepper et al., in a three-year survey of antimicrobial consumption and resistance in P. aeruginosa, found that IPM consumption was not only significantly associated with IPM resistance, but also with CAZ and piperacillin/tazobactam resistance. 13 No correlation was observed between CAZ or piperacillin/tazobactam consumption and resistance to themselves or to IPM. 13 Unlike previous studies, Mohr et al., at a single institution in an eight-year period, did not observe an association between IPM use and resistance in P. aeruginosa. 14 However, an increase in total fluoroquinolone use was associated with an increase in the P c

4 K. Miliani et al. / Journal of Hospital Infection xxx (2011) 1e5 Table III Factors associated with high incidence rates (>75th percentile) of at least one key resistance (ceftazidime, imipenem, ciprofloxacin or amikacin) among Pseudomonas aeruginosa isolates in the multivariate analysis Variables Full model Final model a Adjusted OR (95% CI) P Adjusted OR (95% CI) P Consumption Amoxicillin >p75 2.2 (0.34e13.94) 0.41 e e Piperacillin/tazobactam >p75 1.6 (0.15e16.80) 0.71 e e Ceftazidime >p75 2.4 (0.28e20.40) 0.43 e e Other 3rd generation 1.4 (0.18e10.46) 0.76 e e cephalosporin >p75 Imipenem >p75 7.5 (1.09e51.49) 0.04 7.9 (2.24e28.09) 0.001 Ciprofloxacin >p75 7.0 (1.41e34.27) 0.02 7.9 (2.24e28.09) 0.001 Other fluoroquinolone >p75 3.6 (0.68e19.12) 0.13 e e Amikacin consumption >p75 0.8 (0.12e5.52) 0.82 e e Other aminoglycoside >p75 1.8 (0.25e12.89) 0.55 e e Glycopeptide >p75 0.5 (0.04e5.18) 0.54 e e Macrolide >p75 0.5 (0.05e4.29) 0.51 e e Public non-teaching Reference e e e Public teaching 1.0 (0.10e10.26) 0.99 e e Private non-profit 3.2 (0.49e20.64) 0.23 e e Private for-profit 4.7 (0.59e37.00) 0.15 e e Proportion of non-acute 0.2 (0.03e2.00) 0.19 e e care beds <25% ICATB score 2007 p25 0.4 (0.07e2.38) 0.32 e e ICSHA score 2007 p25 0.4 (0.06e2.37) 0.30 e e OR, odds ratio; CI, confidence interval; ICATB, composite score related to organisation of antibiotic use policy; ICSHA, indicator of overall consumption of hydro-alcoholic hand-rub products. a Output model obtained by stepwise logistic regression analysis. Table IV Model validation results Variable Full model Final model No. of hospitals 84 84 No. of hospital with at least one key 30 30 resistance a rate >p75 e2ll b 69.50 78.63 AIC 105.49 84.63 Area under the ROC curve 0.87 0.80 AIC, Akaike s information criterion; ROC, receiver operating characteristics. a Ceftazidime, imipenem, ciprofloxacin, or amikacin resistance among Pseudomonas aeruginosa isolates. b e2 log likelihood ratio test: c 2 ¼ 9.1 (degrees of freedom ¼ 15, P ¼ 0.87). incidence of IPM-resistant P. aeruginosa; a significant association between levofloxacin use and CIP resistance was also observed. 14 Our results emphasise the major role of IPM and CIP in P. aeruginosa resistance. 12,13,15 It should be noted that our study compared with preceding studies was hospital-based from a multicentre surveillance programme, and conducted with aggregate data on antibiotic consumption and P. aeruginosa resistance rates. It was carried out to determine whether high consumption rates of some antibiotics could be risk factors for HCFs having high incidence rates (>p75) of P. aeruginosa resistances (CAZ, IPM, CIP, or AMK resistance). Retrospective studies from multiple-hospital networks have found a positive association between increasing use of fluoroquinolones and fluoroquinolone-resistant P. aeruginosa. 16,17 Bhavnani et al. found that increasing expenditures of levofloxacin and ofloxacin, but not CIP, were associated with increased resistance rates to ciprofloxacin in P. aeruginosa. 17 Zervos et al. found a significant association between changes in fluoroquinolone use and changes in resistance in P. aeruginosa in 10 teaching hospitals from 1991 to 2000. 16 During this period, fluoroquinolone usage increased by 97% in the participating hospitals. In our study we also collected information from the French public reporting system of infection control indicators regarding three quality indicators likely to measure the general performance of HAI control (ICALIN), antibiotic use policy (ICATB) and hydroalcoholic products consumption (ICSHA). No association was found between high resistance rates (>p75) among P. aeruginosa and the level of these quality indicators. In France, the implementation in 2005 of quality indicators that focus on the control of HAI was made on the assumption that an improvement of these indicators over time would be associated with a reduction in the burden of nosocomial infections. The development of these indicators was based on expert opinion and their public reporting mandated by the Ministry of Health. A recent population-based study in French hospitals found that a higher ICALIN score was significantly associated with a lower prevalence of meticillin-resistant Staphylococcus aureus infection, whereas ICSHA score was not 18 Further studies may be necessary to demonstrate and validate whether these indicators are the best measures to reduce rates of HAI. There are some limitations to our study. Selection bias cannot be excluded because HCFs participating in the antimicrobial network were not a random sample of HCFs in northern France and further studies will be required to determine whether the results are broadly applicable. Furthermore, no information was requested to determine whether isolates were resistant to more than one antibiotic or whether molecular investigation was performed to address the mechanisms by which resistance occurred. The causes of emergence and spread of antibiotic-resistant bacteria in HCFs are multifactorial including the high selective pressure that results from widespread use of antibiotics, particularly in intensive care units, cross-transmission from patient-to-patient, inappropriate or poor infection control measures, patient case-mix, inter-hospital transfer of resistance (clonal spreading of resistant bacteria or horizontal transfer of resistance genes), acquisition of organisms from the hospital environment, a community contribution of resistance, etc. 7,19e24 Nonetheless, because our data were extracted from a hospital-based multicentre surveillance system, it was not possible to obtain information on all potential confounders. For instance, the extracted data did not include data from specific wards for which local epidemiology and usage patterns for antibiotics could differ. Also, our data did not allow analysis of individual patient risk factors.

K. Miliani et al. / Journal of Hospital Infection xxx (2011) 1e5 5 Our study strongly suggests that use of IPM and CIP were related to high rates of resistance in P. aeruginosa, and therefore high consumption of IPM and CIP could play a major role in selective pressure exerted by antibiotics in P. aeruginosa strains. Acknowledgements We thank the members of the Antimicrobial Surveillance Network study group and all healthcare facilities for their participation in the Antimicrobial Surveillance Network. Conflict of interest statement None declared. Funding sources No specific funding has been received for this study. These data have been generated as part of the routine work of our Antimicrobial Surveillance Network. References 1. Thiolet J-M, Lacavé L, Jarno P, et al. Prévalence des infections nosocomiales, France, 2006. Bull Epidemiol Hebd 2007;51e52:429e432. 2. Strateva T, Yordanov D. Pseudomonas aeruginosa e a phenomenon of bacterial resistance. J Med Microbiol 2009;58:1133e1148. 3. Mesaros N, Nordmann P, Plesiat P, et al. Pseudomonas aeruginosa: resistance and therapeutic options at the turn of the new millennium. Clin Microbiol Infect 2007;13:560e578. 4. Aloush V, Navon-Venezia S, Seigman-Igra Y, Cabili S, Carmeli Y. Multidrugresistant Pseudomonas aeruginosa: risk factors and clinical impact. Antimicrob Agents Chemother 2006;50:43e48. 5. Lautenbach E, Synnestvedt M, Weiner MG, et al. Imipenem resistance in Pseudomonas aeruginosa: emergence, epidemiology, and impact on clinical and economic outcomes. Infect Control Hosp Epidemiol 2010;31:47e53. 6. Mutnick AH, Rhomberg PR, Sader HS, Jones RN. Antimicrobial usage and resistance trend relationships from the MYSTIC programme in North America (1999e2001). J Antimicrob Chemother 2004;53:290e296. 7. Fridkin SK, Hill HA, Volkova NV, et al. Temporal changes in prevalence of antimicrobial resistance in 23 U.S. hospitals. Emerg Infect Dis 2002;8:697e701. 8. Rogues AM, Dumartin C, Amadeo B, et al. Relationship between rates of antimicrobial consumption and the incidence of antimicrobial resistance in Staphylococcus aureus and Pseudomonas aeruginosa isolates from 47 French hospitals. Infect Control Hosp Epidemiol 2007;28:1389e1395. 9. Miliani K, L Heriteau F, Alfandari S, et al. Specific control measures for antibiotic prescription are related to lower consumption in hospitals: results from a French multicentre pilot study. J Antimicrob Chemother 2008;62:823e829. 10. Carlet J, Astagneau P, Brun-Buisson C, et al. French national program for prevention of healthcare-associated infections and antimicrobial resistance, 1992e2008: positive trends, but perseverance needed. Infect Control Hosp Epidemiol 2009;30:737e745. 11. Members of the SFM Antibiogram Committee. Comité de l antibiogramme de la société française de microbiologie report 2003. Int J Antimicrob Agents 2003;21: 364e391. 12. Carmeli Y, Troillet N, Eliopoulos GM, Samore MH. Emergence of antibiotic-resistant Pseudomonas aeruginosa: comparison of risks associated with different antipseudomonal agents. Antimicrob Agents Chemother 1999;43: 1379e1382. 13. Lepper PM, Grusa E, Reichl H, Hogel J, Trautmann M. Consumption of imipenem correlates with beta-lactam resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother 2002;46:2920e2925. 14. Mohr JF, Jones A, Ostrosky-Zeichner L, Wanger A, Tillotson G. Associations between antibiotic use and changes in susceptibility patterns of Pseudomonas aeruginosa in a private, university-affiliated teaching hospital: an 8-yearexperience: 1995e2002. Int J Antimicrob Agents 2004;24:346e351. 15. Kang CI, Kim SH, Park WB, et al. Risk factors for antimicrobial resistance and influence of resistance on mortality in patients with bloodstream infection caused by Pseudomonas aeruginosa. Microb Drug Resist 2005;11:68e74. 16. Zervos MJ, Hershberger E, Nicolau DP, et al. Relationship between fluoroquinolone use and changes in susceptibility to fluoroquinolones of selected pathogens in 10 United States teaching hospitals, 1991e2000. Clin Infect Dis 2003;37:1643e1648. 17. Bhavnani SM, Callen WA, Forrest A, et al. Effect of fluoroquinolone expenditures on susceptibility of Pseudomonas aeruginosa to ciprofloxacin in U.S. Hospitals. Am J Health Syst Pharm 2003;60:1962e1970. 18. Grammatico-Guillon L, Thiolet JM, Bernillon P, Coignard B, Khoshnood B, Desenclos JC. Relationship between the prevalence of meticillin-resistant Staphylococcus aureus infection and indicators of nosocomial infection control measures: a population-based study in French hospitals. Infect Control Hosp Epidemiol 2009;30:861e869. 19. Johnson JK, Smith G, Lee MS, et al. The role of patient-to-patient transmission in the acquisition of imipenem-resistant Pseudomonas aeruginosa colonization in the intensive care unit. J Infect Dis 2009;200:900e905. 20. Lepelletier D, Cady A, Caroff N, et al. Imipenem-resistant Pseudomonas aeruginosa gastrointestinal carriage among hospitalized patients: risk factors and resistance mechanisms. Diagn Microbiol Infect Dis 2010;66:1e6. 21. Steinke D, Davey P. Association between antibiotic resistance and community prescribing: a critical review of bias and confounding in published studies. Clin Infect Dis 2001;33(Suppl. 3):S193e205. 22. White RL, Friedrich LV, Mihm LB, Bosso JA. Assessment of the relationship between antimicrobial usage and susceptibility: differences between the hospital and specific patient-care areas. Clin Infect Dis 2000;31: 16e23. 23. Pitten FA, Panzig B, Schroder G, Tietze K, Kramer A. Transmission of a multiresistant Pseudomonas aeruginosa strain at a German university hospital. J Hosp Infect 2001;47:125e130. 24. Bert F, Maubec E, Bruneau B, Berry P, Lambert-Zechovsky N. Multi-resistant Pseudomonas aeruginosa outbreak associated with contaminated tap water in a neurosurgery intensive care unit. J Hosp Infect 1998;39:53e62.