Few studies have been performed on urinary tract infections (UTIs) in men. In the present study, general practitioners (n = 42) from the Dutch Sentinel General Practice Network collected urinary samples from 560 male patients (≥18 years) suspected of UTI and recorded prescribed antibiotic treatment. In this way, the antibiotic susceptibility of Gram-negative uropathogens, including extended-spectrum beta-lactamase (ESBL-) producing Escherichia coli could be determined. In addition, E. coli susceptibility and antibiotic prescriptions were compared with data from a similar UTI study among women and with data collected 7 years earlier. Of 367 uropathogens (66%) identified (≥103 cfu/mL), most were Gram-negative (83%) and E. coli being isolated most frequently (51%). Antibiotic susceptibility to ciprofloxacin, norfloxacin and nitrofurantoin was 94%, 92% and 88%, respectively, whereas co-amoxiclav (76%) and co-trimoxazole (80%) showed lower susceptibilities. One ESBL (0.5%) was found. A significantly higher proportion of female UTIs was caused by E. coli compared with men (72% versus 51%, P<0.05). E. coli susceptibility tended to be lower in men compared with women, although not reaching statistical significance. No changes in E. coli susceptibility were observed over time (all P>0.05). Co-amoxiclav and nitrofurantoin prescriptions increased over time (11% versus 28% and 16% versus 23% respectively, both P<0.05), whereas co-trimoxazole prescriptions decreased (24% versus 14%, P<0.05). In conclusion, given the observed gender differences in uropathogen distribution and (tendency in) E. coli antibiotic susceptibility, empirical male UTI treatment options should be based on surveillance studies including men only. When awaiting the culture result is clinically not possible, fluoroquinolones are advised as first-choice antibiotics for male UTIs in Dutch general practices based on current antibiotic susceptibility data. The prevalence of ESBL-producers was low and no differences were observed in antibiotic susceptibility over a 7-year period. In addition, antibiotic prescriptions changed in accordance with national guidelines during this time period.
Citation: den Heijer CDJ, Penders J, Donker GA, Bruggeman CA, Stobberingh EE (2013) The Importance of Gender-Stratified Antibiotic Resistance Surveillance of Unselected Uropathogens: A Dutch Nationwide Extramural Surveillance Study. PLoS ONE 8(3): e60497. https://doi.org/10.1371/journal.pone.0060497
Editor: Praveen Thumbikat, Northwestern University, United States of America
Received: December 3, 2012; Accepted: February 27, 2013; Published: March 29, 2013
Copyright: © 2013 den Heijer et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Dutch Foundation of the Working Party on Antibiotic Policy (SWAB). The Dutch Sentinel GP network is supported by a grant from the Ministry of Health, Welfare and Sports (the Netherlands). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Most studies of urinary tract infections (UTIs) focus on female patients because of the higher incidence in women than in men. For this reason, most UTI guidelines are based on studies performed among women, despite the obvious genito-urinary differences . Even so, it was estimated that in 2010 128,000 episodes of UTI were experienced by men in the Dutch general practice setting, representing 16 per 1000 male patients .
Several authors have stressed the importance that gender-stratified UTI surveillance studies should be performed, because of observed differences in antimicrobial susceptibility , . However, these studies were biased by the inclusion of selected isolates, probably after initial treatment failure leading to an overestimation of resistance , .
In the Netherlands, regular surveillance studies are performed on unselected urinary isolates of both men and women to evaluate empirical treatment options in general practice, from which the female data have recently been reported . With this data, we were able to assess the actual difference in E. coli susceptibility between male and female patients.
Furthermore, the current treatment options for male UTIs could be evaluated as stated in the guidelines of the Dutch College of General Practitioners (NHG). These are based on the presence of symptoms of tissue invasion (i.e. high fever (>38°C), chills, malaise, flank or perineal pain) with nitrofurantoin as first-choice agent and trimethoprim as second option for men without these symptoms. Male UTI with symptoms of tissue invasion should preferably be treated with co-amoxiclav followed by fluoroquinolones and co-trimoxazole as shared second-choice .
Also, the prevalence of extended-spectrum beta-lactamase (ESBL-) producing E. coli was assessed, as male gender has been identified as a risk factor for ESBL-producing Enterobacteriaceae .
Finally, differences in antibiotic susceptibility results over time could be determined for men, because a similar UTI surveillance study was performed between January 2003 and December 2004 .
All data in this study were analyzed anonymously, only diagnostic specimens sent for diagnostic investigations and clinical data were used. Therefore, no consent was required from the patient and the ethics committee did not have to be approached. This is in agreement with the Medical Research Involving Human Subjects Act, the code for proper use of human tissue as formulated by the Dutch Federation of Medical Scientific Societies and the policy of the Medical Ethics Committee of the Maastricht University Medical Centre.
Patients and antibiotic prescriptions
General practices (n = 42) from the Dutch NIVEL Sentinel General Practice Network were used for the recruitment of patients. The patient population of this network is nationally representative by age, gender, region and population density .
From January 2009 to June 2011, GPs included male general practice patients (≥18 years of age) with symptoms of UTI, i.e. dysuria, urinary frequency and/or urgency. Catheterized patients and patients suspected of having a sexually transmitted disease were excluded. With the exception of benign prostate hypertrophy, patients were also excluded when having urological or nephrological co-morbidity, diabetes mellitus or other immunocompromising diseases.
For all eligible patients, GPs filled in a patient form including age, symptoms of tissue invasion (fever (>38°C), flank pain) (yes/no) and empirical antibiotic treatment prescribed during the patient's visit (yes/no and if yes the prescribed agent was specified).
Urine collection and processing
Patients provided a fresh-voided (midstream) urine sample, which was used to prepare a dipslide (Uriline, 56508, Biomérieux, Plainview, NY, USA) according to the manufacturer's instructions. For incubation and further microbiological analysis, dipslides were sent to the microbiological laboratory of Maastricht University Medical Centre, The Netherlands.
Bacterial growth was determined on the day of arrival and was considered positive at ≥103 cfu/mL . If no growth was observed, dipslides were incubated overnight at 37°C and assessed again. Standard microbiological methods were used for the identification of the uropathogens . When more than one bacterium was cultured, only the predominant one was included in the final analysis. After isolation and identification, all bacterial strains were kept at −20°C in peptone/glycerol (30% w/v) for further testing.
Antibiotic susceptibility testing
The susceptibility of Gram-negative isolates was determined using the microdilution method for the following antibiotics: amoxicillin, co-amoxiclav, trimethoprim, co-trimoxazole, norfloxacin, ciprofloxacin and nitrofurantoin. E. coli ATCC 35218 and ATCC 25922 were used as control strains. Methods and susceptibility breakpoints were in accordance with the EUCAST guidelines .
Putative ESBL-producing E. coli strains were selected on the basis of resistance to co-amoxiclav , and confirmed by means of a combination disc diffusion test (Neo Sensitabs, Rosco Diagnostica, Denmark) with ceftazidime, cefepime and cefotaxime with and without clavulanic acid, following the guidelines of the Dutch Society of Medical Microbiology (NVMM) .
Comparison of E. coli susceptibility between men and women
From January 2009 to July 2009, a similar surveillance study among women suspected of UTI was performed . This study was conducted by the same research group as the present study and patients were also recruited by GPs from the Dutch NIVEL Sentinel General Practice Network.
Comparison between 2004 and 2011 data
The chosen categories (uropathogens and antibiotics) and all methods, including the network of the participating GPs, were the same in both studies, except for the guidelines used for the E. coli susceptibility breakpoints, i.e. CLSI (2004) and EUCAST (2011). Since both studies were performed by the same research group, the original MIC values from the 2004 study were available and 2004 susceptibility data could be recalculated according to the EUCAST guidelines.
For the comparison of two groups, a Pearson's Chi-square test was used for categorical variables and an independent-samples t-test for continuous variables when parametric assumptions were met. The Mantel-Haenszel extension test for trend was used to determine a trend between more than two categorical variables. The programme PSAW 18.0 for Windows was used for statistical analyses and P<0.05 was considered statistically significant.
Patients and culture results
In total, 603 men were included with a median age of 65 years (range 18–97 years). Patients were considered eligible for analyses when data from the patient form were complete (560/603, 93%). The median age of population with complete data did not differ from the full patient population. Of these 560 patients, 137 (24%) had symptoms of tissue invasion. The median age of the populations with and without signs of tissue invasion was similar (66 years versus 64 years, respectively, P>0.05).
A positive culture result was found in 367 (66%) men, of whom 100 (27%) had symptoms of tissue invasion. The probability of a culture being positive increased with age, with men aged 70+ having a 3-times higher chance of a positive culture than men below the age of 50 years (Table 1).
Of all isolated uropathogens, 83% was Gram-negative with E. coli being most commonly found (51% of all uropathogens). Differences in uropathogen distribution per age category were not observed (Table 2).
Prevalence of antibiotic susceptibility
With respect to the preferred agents for men without signs of tissue invasion according to NHG, Gram-negatives showed a high overall susceptibility to nitrofurantoin (88%), whereas a lower susceptibility was found to trimethoprim (75%). Co-amoxiclav, the first-choice agent for men with signs of tissue invasion according to NHG, showed a relatively low antibiotic susceptibility (76%), as well as co-trimoxazole (80%). On the other hand, the fluoroquinolones (i.e. norfloxacin and ciprofloxacin) had a high susceptibility to the Gram negatives (92% and 94% respectively) (Table 3). When these Gram-negative susceptibility data were stratified according to the age categories mentioned earlier, i.e. 18–50, 51–70 and 70+ years, no differences were observed per antibiotic with age (all P>0.05, data not shown).
Empirical antibiotic treatment was prescribed to 359 patients (64%), of whom 284 (79%) were later confirmed as being culture positive. 70+ year old men had a 2-times higher chance to receive antibiotic treatment empirically compared with 18-50 year old men (OR: 2.0, 95% confidence interval: 1.2–3.2, P for trend <0.001). Fluoroquinolones showed the highest prevalence of prescription (29%), followed by co-amoxiclav (28%), nitrofurantoin (23%) and co-trimoxazole (14%).
In Figure 1 a flow chart is given based on the presence or absence of signs of tissue invasion, empirical therapy and a positive culture result. In the group with signs of tissue invasion (n = 137), co-amoxiclav was prescribed more frequently and nitrofurantoin less frequently compared with the prescriptions of the population without signs of tissue invasion (39% versus 24%, P = 0.004, and 11% versus 27%, P = 0.002, respectively).
Comparison of E. coli susceptibility between men and women
In women, E. coli accounted for 72% of the total isolates, compared with 51% in men (P<0.001). Except for nitrofurantoin, E. coli susceptibilities tended to be lower in men compared with women: amoxicillin (65% versus 66%), co-amoxiclav (84% versus 87%), trimethoprim (77% versus 81%), co-trimoxazole (78% versus 84%), fluoroquinolones (94% versus 97%) and nitrofurantoin (both 100%), without being significant for the antibiotics tested (all P>0.05).
Comparison of 2004 and 2011 data
No differences were observed between E. coli antibiotic susceptibilities in the 2004 and 2011 studies (Table 4). This also applied to the prevalence of extended-spectrum beta-lactamases (ESBLs), as no ESBL-producing E. coli were found in 2004 and one (0.5%) in 2011 (P = 0.44).
No significant differences in E. coli susceptibilities (EUCAST) were observed between 2004 and 2011 study (all P>0.05)When comparing the antibiotic prescriptions between 2004 and 2011, GPs prescribed more co-amoxiclav and nitrofurantoin (11% versus 28%, P<0.001, and 16% versus 23%, P = 0.023, respectively) and less co-trimoxazole (24% versus 14%, P = 0.002) in 2011. For fluoroquinolones, amoxicillin and trimethoprim the prevalence of prescription was similar in the two studies (33% versus 29%, 2% versus 3% and 2% versus 1% respectively).
In this study of UTIs in Dutch general practice, differences were observed between men and women regarding the distribution of uropathogens, and uropathogenic E. coli susceptibilities tended to be lower among men compared with women. For male UTIs, fluoroquinolones showed the highest prevalence of susceptibility to isolated Gram-negative uropathogens, followed by nitrofurantoin. ESBL-producing E. coli prevalence was low and no differences in E. coli susceptibility were observed over a 7-year period.
The important feature of this study is the participation of general practices from the NIVEL Sentinel General Practice Network. This network has a nationally representative patient population, making it possible to generalize our conclusions to the whole Dutch outpatient population. Moreover, to our knowledge, no other study has been able to include over 600 unselected urinary samples from men suspected of UTI in an outpatient setting.
The fact that no susceptibilities were determined for Gram-positive bacteria could be seen as a limitation. However, empirical treatment for uncomplicated female UTIs is often primarily based on E. coli susceptibilities, covering 75–95% of the total spectrum of uropathogens in female UTIs . The proportion of Gram-negatives in the present study is within this range, as also described by Lipski , thereby supporting an evidence-based empirical antibiotic choice for UTI in men.
Also, our study population could be regarded as heterogeneous with the inclusion of men aged 18+, because pathophysiology of male UTI differs with age . We have tried to circumvent this problem by using strict in- and exclusion criteria, in order to obtain a homogeneous sample of men suspected of UTI. Moreover, stratification of antibiotic susceptibility results showed no differences by age.
UTI in men has also been studied by Hummers-Pradier et al. and they used 102 and 105 cfu/mL as cut-off values to determine a UTI . However, their prevalence of UTI (60%), using 102 cfu/mL as cut-off, was still lower than our prevalence. The fact that only 36% of the patients included were prescribed an antibiotic in that study and that GPs suspected a UTI in just over 50% of the included patients, suggests that their patient population was at lower risk of UTI than our population with a prescription rate of 64%. In addition, their results were derived from a relatively small patient population (n = 79), resulting in estimates with large confidence intervals. Our choice to use 103 cfu/mL as cut-off value for the diagnosis of male UTI was based on the European urinalysis guidelines and is supported by Lipsky et al. , .
The prevalence of positive culture (66%) we found was also slightly higher than the one reported in the 2004 study (56%) . The higher median age in the present study (65 versus 58 years) could explain this difference, because it is known that UTI incidence increases with age, and especially in men, due to the higher frequency of prostatic hypertrophy . In the present study we could confirm this age-dependent trend for positive cultures.
Only 51% of the uropathogens causing male UTI was E. coli versus 72% of the UTIs in women. Moreover, uropathogenic E. coli susceptibilities tended to be lower in men in women, although no statistical significance was reached. The difference in these important characteristics for determining optimal empirical treatment, show that men with UTI should be considered as a specific patient category.
Given the more heterogeneous population of uropathogens that causes a UTI in men compared with women, empirical treatment should best be avoided for male UTI. When immediate treatment is necessary based on clinical judgement, our data can be used for the optimal choice of empirical treatment. Our recommendations can be applied to all men aged 18+ suspected of UTI in the Dutch GP setting, because no differences in uropathogen distribution and antibiotic susceptibility per age category were observed.
We found that fluoroquinolones showed a higher prevalence of susceptibility than co-amoxiclav, the current first-choice for men with symptoms of tissue invasion , and the former agents possess specific pharmacokinetic properties leading to high prostate tissue concentrations, which co-amoxiclav lacks –.
The choice for EUCAST instead of CLSI guidelines had an impact on the E. coli susceptibility to co-amoxiclav. Using EUCAST breakpoints, the E. coli susceptibility to co-amoxiclav in 2004 decreased from 100% to 89%. The current prevalence of susceptibility of Gram-negatives to co-amoxiclav (77%) is below the proposed threshold (80%) for antibiotics to still be effective as empirical treatment for UTI .
Based on these findings, fluoroquinolones may be regarded as the current preferred treatment in men with tissue invasion. However, in addition to antibiotic resistance data, side effects of the drugs related to age, financial consequences and patient characteristics should be included in the choice of the preferred antibiotics.
Except for Proteus species and non-fermenters, high antibiotic susceptibilities (≥87%) to Gram-negatives were also found for nitrofurantoin. Similar results were described by Guay . Despite this high prevalence of susceptibility, clinical trials on the effectiveness of nitrofurantoin for male UTIs are currently lacking. Since nitrofurantoin does not reach therapeutic concentrations in prostatic tissue , differentiation between cystitis with and without prostatitis is important. So far, no symptoms have been found that allow a conclusive differentiation between these two conditions . The Dutch GP guidelines have based their treatment choices on the presence or absence of so-called ‘symptoms of tissue invasion’, although no reference was given. Results from future trials would support evidence-based treatment of UTI in men and could help to restrict the use of fluoroquinolones, thereby limiting the chance of the development of antibiotic resistance to this antibiotic group. In this perspective, also the value of fosfomycin for male UTIs needs exploration .
For male UTIs, second-choice agents (i.e. fluoroquinolones) were still prescribed most, although overall compliance with national guidelines improved in comparison with prescription data from 2004. This was shown by an increase in co-amoxiclav and nitrofurantoin prescriptions, the first-choice agents for men with and without symptoms of tissue invasion respectively.
In the surveillance study on female UTIs in Dutch general practices, E. coli susceptibilities from 2004 and 2009 were compared, and also no differences in time trend were observed . Several other studies, including urinary samples from men and women, have reported an increased resistance over time among isolates from an outpatient setting , . Our regular surveillance studies, the results of which are implemented in clinical practice by including them in the update of the national GP guidelines, appear to have contributed to the control of antibiotic resistance in Dutch general practices. The low prevalence of ESBL-producing E. coli is in accordance with these findings.
In conclusion, among unselected E. coli originating from Dutch male GP patients suspected of UTI, the prevalence of ESBL-producers was low and no differences were observed in antibiotic susceptibility over a 7-year period. In addition, antibiotic prescriptions changed in accordance with national guidelines during this time period. For male UTIs in Dutch general practice, fluoroquinolones may be regarded as the best treatment option on the basis of current susceptibility data. Differences in uropathogen distribution, together with the tendency of a lower E. coli antibiotic susceptibility in men versus women, make it necessary to base male UTI treatment recommendations on UTI studies performed among men only.
We thank all participating GPs and their patients. Without their help this study could not have been conducted. In addition, we would like to acknowledge Christel Driessen, Mayk Lucchesi, Jacqueline Maes and Jacqueline Peeters for their excellent help with the performance of the microbiological analyses.
Conceived and designed the experiments: GD ES. Performed the experiments: CH. Analyzed the data: CH JP. Wrote the paper: CH JP GD CB ES.
- 1. Koeijers JJ, Kessels AG, Nys S, Bartelds A, Donker G, et al. (2007) Evaluation of the nitrite and leukocyte esterase activity tests for the diagnosis of acute symptomatic urinary tract infection in men. Clin Infect Dis 45: 894–896.
- 2. Donker GA Continuous Morbidity Registration Sentinel General Practice Network. Annual Report 2010. NIVEL, Utrecht. Available: www.nivel.nl/peilstations.
- 3. Rocha JL, Tuon FF, Johnson JR (2012) Sex, drugs, bugs, and age: rational selection of empirical therapy for outpatient urinary tract infection in an era of extensive antimicrobial resistance. Braz J Infect Dis 16: 115–121.
- 4. Bean DC, Krahe D, Wareham DW (2008) Antimicrobial resistance in community and nosocomial Escherichia coli urinary tract isolates, London 2005-2006. Ann Clin Microbiol Antimicrob 7: 13.
- 5. Critchley IA, Karlowsky JA (2004) Optimal use of antibiotic resistance surveillance systems. Clin Microbiol Infect 10: 502–511.
- 6. den Heijer CD, Donker GA, Maes J, Stobberingh EE (2010) Antibiotic susceptibility of unselected uropathogenic Escherichia coli from female Dutch general practice patients: a comparison of two surveys with a 5 year interval. J Antimicrob Chemother 65: 2128–2133.
- 7. van Pinxteren B, van Vliet SM, Wiersma TJ, Goudswaard AN (2006) [Summary of the practice guideline 'Urinary-tract infections' (second revision) from the Dutch College of General Practitioners]. Ned Tijdschr Geneeskd 150: 718–722.
- 8. Ben-Ami R, Rodriguez-Bano J, Arslan H, Pitout JD, Quentin C, et al. (2009) A multinational survey of risk factors for infection with extended-spectrum beta-lactamase-producing enterobacteriaceae in nonhospitalized patients. Clin Infect Dis 49: 682–690.
- 9. Koeijers JJ, Verbon A, Kessels AG, Bartelds A, Donkers G, et al. (2010) Urinary tract infection in male general practice patients: uropathogens and antibiotic susceptibility. Urology 76: 336–340.
- 10. Donker GA (2010) Continuous morbidity registration at Dutch sentinel stations, 2008. Utrecht: Netherlands Institute for Health Services Research (NIVEL).
- 11. Lipsky BA, Ireton RC, Fihn SD, Hackett R, Berger RE (1987) Diagnosis of bacteriuria in men: specimen collection and culture interpretation. J Infect Dis 155: 847–854.
- 12. Isenberg HD (1998) Essential procedures for clinical microbiology. Washington, D.C.: ASM Press. xv: , 838 p. p.
- 13. European Committee for Antimicrobial Susceptibility Testing (EUCAST) of the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) (2000) EUCAST Definitive Document E.DEF 3.1, June 2000: Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by agar dilution. Clin Microbiol Infect 6: 509–515.
- 14. NVMM NVMM-richtlijn voor screening en confirmatie van extended-spectrum beta-lactamases(ESBL's) in Enterobacteriaceae. Kwaliteitskoepel website. Available: http://www.kwaliteitskoepel.nl/assets/structured-files/NVMM/SCREENINGENCONFIRMATIEVANEXTENDEDSPECTRUMBETA-LACTAMASESESBLsINENTEROBACTERIACEAE.pdf Accessed: 2013 Jan 28.
- 15. Gupta K, Hooton TM, Naber KG, Wullt B, Colgan R, et al. (2011) International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: A 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clin Infect Dis 52: e103–120.
- 16. Lipsky BA (1999) Prostatitis and urinary tract infection in men: what's new; what's true? Am J Med 106: 327–334.
- 17. Hummers-Pradier E, Ohse AM, Koch M, Heizmann WR, Kochen MM (2004) Urinary tract infection in men. Int J Clin Pharmacol Ther 42: 360–366.
- 18. European Confederation of Laboratory Medicine (2000) European urinalysis guidelines. Scand J Clin Lab Invest Suppl 60: 231-296.
- 19. Wagenlehner FM, Weidner W, Naber KG (2007) Therapy for prostatitis, with emphasis on bacterial prostatitis. Expert Opin Pharmacother 8: 1667–1674.
- 20. Wagenlehner FM, Naber KG (2006) Current challenges in the treatment of complicated urinary tract infections and prostatitis. Clin Microbiol Infect 12 Suppl 367–80.
- 21. Charalabopoulos K, Karachalios G, Baltogiannis D, Charalabopoulos A, Giannakopoulos X, et al. (2003) Penetration of antimicrobial agents into the prostate. Chemotherapy 49: 269–279.
- 22. Gupta K, Hooton TM, Stamm WE (2001) Increasing antimicrobial resistance and the management of uncomplicated community-acquired urinary tract infections. Ann Intern Med 135: 41–50.
- 23. Guay DR (2001) An update on the role of nitrofurans in the management of urinary tract infections. Drugs 61: 353–364.
- 24. Etienne M, Chavanet P, Sibert L, Michel F, Levesque H, et al. (2008) Acute bacterial prostatitis: heterogeneity in diagnostic criteria and management. Retrospective multicentric analysis of 371 patients diagnosed with acute prostatitis. BMC Infect Dis 8: 12.
- 25. Falagas ME, Vouloumanou EK, Togias AG, Karadima M, Kapaskelis AM, et al. (2010) Fosfomycin versus other antibiotics for the treatment of cystitis: a meta-analysis of randomized controlled trials. J Antimicrob Chemother 65: 1862–1877.
- 26. Yilmaz N, Agus N, Yurtsever SG, Pullukcu H, Gulay Z, et al. (2009) Prevalence and antimicrobial susceptibility of Escherichia coli in outpatient urinary isolates in Izmir, Turkey. Med Sci Monit 15: PI61–65.
- 27. Zhanel GG, Hisanaga TL, Laing NM, DeCorby MR, Nichol KA, et al. (2005) Antibiotic resistance in outpatient urinary isolates: final results from the North American Urinary Tract Infection Collaborative Alliance (NAUTICA). Int J Antimicrob Agents 26: 380–388.