Received 5 October 2014; returned 4 November 2014; revised 9 November 2014; accepted 8 December 2014

Transcription

1 J Antimicrob Chemother 2015; 70: doi: /jac/dku548 Advance Access publication 27 January 2015 Risk factors for ambulatory urinary tract infections caused by high-mic fluoroquinolone-susceptible Escherichia coli in women: results from a large case control study Pinyo Rattanaumpawan 1,2 *, Irving Nachamkin 3, Warren B. Bilker 4, Jason A. Roy 4, Joshua P. Metlay 5, Theoklis E. Zaoutis 1 and Ebbing Lautenbach 1,6 on behalf of the CDC Prevention Epicenters Program 1 Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, PA, USA; 2 Division of Infectious Diseases and Tropical Medicine, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand; 3 Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA, USA; 4 Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA, USA; 5 Division of General Internal Medicine, Massachusetts General Hospital, Boston, MA, USA; 6 Division of Infectious Diseases, Department of Medicine, University of Pennsylvania, Philadelphia, PA, USA *Corresponding author. Tel: ; Fax: ; pinyo@mail.med.upenn.edu Received 5 October 2014; returned 4 November 2014; revised 9 November 2014; accepted 8 December 2014 Objectives: The prevalence of high-mic fluoroquinolone-susceptible Escherichia coli (FQSEC) has been increasing. These isolates are one step closer to full fluoroquinolone (FQ) resistance and may lead to delayed response to FQ therapy. Our study aimed to investigate the epidemiology of high-mic FQSEC in ambulatory urinary tract infections (UTIs). Patients and methods: A case control study was conducted at outpatient services within the University of Pennsylvania Health System, Philadelphia. All female subjects with non-recurrent UTI caused by FQSEC (levofloxacin MIC, 4 mg/l) were enrolled. Cases were subjects with high-mic FQSEC UTI (levofloxacin MIC.0.12 but, 4 mg/l) and controls were subjects with low-mic FQSEC UTI (levofloxacin MIC 0.12 mg/l). Data on microbiology results and baseline characteristics were extracted from electronic medical records. Results: During the 3 year study period (May 2008 April 2011), episodes of E. coli bacteriuria were identified. The prevalence of FQSEC, FQ-intermediate susceptible E. coli and FQ-resistant E. coli was 75.0%, 0.4% and 24.6%, respectively. A total of 2001 female subjects with FQSEC UTI were enrolled into our study (165 cases and 1836 controls). Independent risk factors for high-mic FQ susceptibility included Asian race (OR¼2.92; 95% CI¼ ; P¼0.02), underlying renal disease (OR¼2.18; 95% CI¼ ; P¼0.02) and previous nitrofurantoin exposure (OR¼8.86; 95% CI¼ ; P¼0.005). Conclusions: Asian race, underlying renal disease and previous exposure to nitrofurantoin were identified as independent risk factors for high-mic FQSEC. There may be some factors that are more common in Asians, which may result in the selection of high-mic FQSEC. Further studies are necessary to explore these findings. Keywords: fluoroquinolone resistance, UTIs, Escherichia coli Introduction Fluoroquinolone (FQ) resistance generally emerges by point mutations in the coding regions of the DNA gyrase subunits (gyra and gyrb) and DNA topoisomerase IV (parc and pare) inastepwise pattern. 1 Data from several molecular studies have revealed that Escherichia coli isolates with reduced susceptibility to FQs typically harbour a single mutation in a gyra gene while E. coli isolates with full resistance to FQ usually have double mutations in the gyra and parc genes. 2 4 Enterobacteriaceae are considered FQ-susceptible (FQS) if the levofloxacin MIC is,4 mg/l, FQ-intermediate susceptible (FQI) if the levofloxacin MIC is 4 but,8 mg/l and FQ-resistant (FQR) if the levofloxacin MIC is 8 mg/l. 5 Some investigators have categorized the FQS E. coli (FQSEC) into two additional categories: (i) low-mic FQSEC (levofloxacin MIC 0.12 mg/l); and (ii) high-mic FQSEC (levofloxacin MIC.0.12 but,4 mg/l). These two categories are sometimes called fully susceptible strains and reduced susceptible strains, respectively. 3,6 A 1998 Taiwan survey reported that 11.3% and 21.7% of their 1203 E. coli urinary isolates were FQR and high-mic FQS, respectively. 3 Another surveillance study of faecal E. coli isolates at two US hospitals reported that the prevalences of FQR isolates and high-mic FQS isolates were 12.9% and 5.8%, respectively. 6 # The Author Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please journals.permissions@oup.com 1547

2 Rattanaumpawan et al. While many studies focused on risk factors for infections caused by FQR pathogens, risk factors for infections caused by high-mic FQSEC have never been successfully identified. Therefore, we conducted a case control study aiming to investigate risk factors for ambulatory urinary tract infections (UTIs) caused by high-mic FQSEC in women. This study is the largest specifically designed to explore this issue. Patients and methods Study settings and design We conducted a case control study of female subjects with ambulatory FQSEC UTIs at outpatient practices within the University of Pennsylvania Health System from May 2008 to April The study was approved by the University of Pennsylvania Institutional Review Board. Only practices that sent their clinical specimens to the Hospital of the University of Pennsylvania Microbiology Laboratory (HUP-MicroLab) were eligible for the study. These included two university hospitals, two study centres, three community practices and eight outpatient clinical practices within the Clinical Care Associates and the Clinical Practices of the University of Pennsylvania. Study population and study definitions All females (age 18 years) for whom a urine culture grew FQSEC were identified through the HUP-MicroLab database. Only those who met our study definition of UTI (Table S1, available as Supplementary data at JAC Online) within 7 days before or after the index date were considered as having FQSEC UTI. The index date was the date on which a urine specimen with FQSEC was collected. Only the first UTI episode during the study period was included. Subjects who had had a UTI episode within 30 days prior to the beginning of the study were also excluded. The study definition for ambulatory UTIs was specifically created by combining the CDC/National Healthcare Safety Network surveillance definition of healthcare-associated UTIs and the ICD-9 coding algorithm to detect UTI in an electronic database based on our pilot study. 7,8 To validate our study definition, we randomly sampled 100 subjects with significant FQSEC bacteriuria. Our gold standard is the written diagnosis of UTI on the medical record. Our study definition provided 87.8% sensitivity and 85.7% specificity to identify FQSEC UTI within our source population. Microbiological tests were routinely processed using the Vitek-2 system, according to the performance standards for antimicrobial susceptibility testing established by the CLSI. 5 An E. coli isolate was considered FQSEC, FQI E. coli (FQIEC) and FQR E. coli (FQREC) if the levofloxacin MIC was,4, 4 but,8 and 8 mg/l, respectively. In addition, FQSEC isolates were further categorized into two additional categories: (i) low-mic FQS, levofloxacin MIC 0.12; and (ii) high-mic FQS, levofloxacin MIC.0.12 but,4 mg/l. Cases were subjects with high-mic FQSEC UTI and controls were subjects with low-mic FQSEC UTI. Data collection We obtained all necessary data from the HUP-MicroLab database and the Electronic Medical Record. The presence of comorbidities was evaluated by applying the Enhanced ICD-9-CM coding algorithms to ICD-9 codes. 9,10 We excluded any antibiotics prescribed within 7 days prior to the index date to avoid overestimation of the effect of antibiotic exposure. Urine culture results within 90 days prior and after the index date were also obtained. Statistical analysis Categorical variables were summarized by frequency while continuous variables were summarized by the mean, median, standard deviation and range. Bivariable analysis was performed to determine the unadjusted association between high-mic FQSEC and potential risk factors. Mantel Haenszel stratified analysis was performed to evaluate the possible effect of confounders and interactions. Finally, we performed multiple logistic regression analysis by adding any variables that were associated with high-mic FQSEC on bivariable analysis (P 0.20). All calculations were performed using STATA version 12.0 (Stata Corp., College Station TX, USA). Atwo-tailedP value of,0.05 was considered statistically significant. Results During the study period, there were urine specimens that grew E. coli 10 5 cfu/ml. The distributions of FQSEC, FQIEC and FQREC were 75.0%, 0.4% and 24.6% respectively. Among females with non-recurrent episodes of E. coli UTI, 74.9% were in the FQSEC group, 0.4% in the FQIEC group and 24.7% in the FQREC group. Of these 2001 females with FQSEC UTI, 91.8% (1836/2001) had low-mic FQSEC UTI and 8.2% (165/2001) had high-mic FQSEC UTI. Details of the susceptibility distribution are shown in Table S2. Demographics, comorbidities and previous drug exposure of cases and controls are shown in Table 1. Baseline characteristics were mostly comparable between cases and controls. Asian race was significantly associated with high-mic FQSEC in bivariable analysis (OR¼2.82; 95% CI¼ ; P¼0.02). Cases were more likely to have congestive heart failure (OR ¼2.44; 95% CI¼ ; P ¼ 0.001) and renal disease (OR¼2.05; 95% CI¼ ; P ¼ 0.02). Furthermore, the high-mic group tended to have previous exposure to any antibiotics (OR¼1.80; 95% CI¼ ; P ¼ 0.15) and nitrofurantoin specifically (OR¼5.65; 95% CI¼ ; P¼0.08). The variables that remained independent risk factors for high-mic FQSEC UTI in multivariable analysis are shown in Table 2. These factors included Asian race (OR¼2.92; 95% CI¼ ; P¼0.02), renal disease (OR¼2.18; 95% CI¼ ; P¼0.02) and previous nitrofurantoin exposure (OR¼8.86; 95% CI¼ ; P¼0.005). Discussion Among females with non-recurrent episodes of E. coli UTI, 74.9% were in the FQSEC group, 0.4% in the FQIEC group and 24.7% in the FQREC group. Among those with FQSEC UTI, only 8.2% were in the high-mic group while 91.8% were in the low-mic group. Compared with the Taiwan study, our prevalence of high-mic FQSEC was relatively low, but our prevalence of FQREC was relatively high. 3 Different from our study, the Taiwan study also included colonization and recurrent cases as well as the isolates from inpatient settings. Racial disparity has been described in various antibioticresistant infections However, the association between Asians and antibiotic resistance has never been documented. This may be the result of differences in the amount of antibiotic consumption, pharmacokinetics and pharmacodynamics of antibiotics, or dietary patterns. To clearly explain this association, further study exploring the racial differences is needed. Our study also identified renal disease as an independent risk factor. The correlation between chronic renal disease and antibiotic resistance has already been documented It is well known that patients with renal disease tend to receive antibiotics. 1548

3 High-MIC fluoroquinolone-susceptible Escherichia coli UTI JAC Table 1. Demographics, comorbidity and previous drug exposure among the cases and controls High MIC (N¼165) Low MIC (N¼1836) Variable n % n % Unadjusted OR (95% CI) P Demographics age (years), mean+sd ( ) 0.81 race white reference black ( ) Asian ( ) 0.02 other/unknown ( ) Comorbidity Charlson index, mean+sd having 1 Charlson conditions ( ) 0.67 acute myocardial infarction ( ) 0.16 congestive heart failure ( ) peripheral vascular disease ( ) 0.57 a cerebrovascular disease ( ) 0.93 dementia ( ) 0.54 a COPD ( ) 0.55 rheumatoid disease ( ) 0.32 a peptic ulcer ( ) 0.26 a mild liver disease ( ) 0.99 a moderate/severe liver disease a diabetes ( ) 0.38 hemiplegia or paraplegia ( ) 0.35 a renal disease ( ) 0.02 cancer ( ) 0.25 metastatic cancer ( ) 0.57 a AIDS Previous drug exposure all antibiotics (D8 D90) b ( ) 0.15 cotrimoxazole ( ) 0.40 a clindamycin a linezolid a metronidazole a nitrofurantoin ( ) 0.08 a vancomycin a cephalosporins a fluoroquinolone a aminoglycoside penicillin ( ) 0.26 a b-lactams ( ) 0.64 macrolide a non-antibiotic drugs (D0 D90) c proton pump inhibitor ( ) 0.48 H 2 blocker ( ) 0.20 a P value from the non-parametric test. b D8 D90: days 8 90 prior to the index date. c D0 D90: days 0 90 prior to the index date. Nitrofurantoin was also found to be an independent risk factor for high-mic FQSEC UTI. Based on data from chart-review, these six subjects received nitrofurantoin because they had UTI symptoms, but they did not meet our study definition of UTIs. Therefore, previous nitrofurantoin exposure may be the proxy of recurrent UTI. Surprisingly, we could not detect any significant association between high-mic FQSEC UTI and other antibiotics except 1549

4 Rattanaumpawan et al. Table 2. Independent risk factors for high-mic FQSEC UTI (multivariable analysis) Risk factor Unadjusted OR (95% CI) Adjusted OR (95% CI) P Race white reference reference black 1.00 ( ) 0.99 ( ) Asian 2.82 ( ) 2.92 ( ) 0.02 others 1.71 ( ) 1.72 ( ) Renal disease 2.05 ( ) 2.18 ( ) 0.02 Previous exposure to 5.65 ( ) 8.86 ( ) nitrofurantoin nitrofurantoin. Given the very low prevalence of previous antibiotic exposure among our study subjects, a small association may be missed. Our study has numerous strengths. This was the first study specifically designed to investigate risk factors for ambulatory UTIs caused by high-mic FQSEC. Unlike the previous study that failed to distinguish true infection from colonization, 3 our study enrolled only subjects with documented FQSEC UTI. Furthermore, we carefully performed a pilot study to select the most accurate diagnosis criteria for ambulatory UTIs because the standard definition for UTIs in the ambulatory setting has never been established. Our study also has some potential limitations. First, we did not perform genetic sequencing to detect any clonal dissemination. Second, we mainly relied on data from the Penn Data Store and HUP-MicroLab; therefore some information (i.e. travel history) could not be retrieved. Third, we may underestimate previous antibiotic exposure. To explore this issue, we performed a chart-review of 100 subjects and found that only 1.5% had had an off-network visit, none of them related to antibiotic prescription. Lastly, our study population was drawn from a number of ambulatory practices within the University of Pennsylvania Health System; therefore generalizability may be the issue. In conclusion, our study is the largest specifically designed to identify risk factors for UTIs caused by high-mic FQSEC. Our study revealed three independent risk factors: Asian race, renal disease and previous exposure to nitrofurantoin. Based on data from chart-review, previous nitrofurantoin exposure seems to be a proxy of recurrent UTI episodes. Further studies are necessary to explore the association between Asian race and high-mic FQSEC UTIs. Acknowledgements These data were presented at the ID Week 2013 TM, San Francisco, CA, USA (Poster 1579). Funding This study was supported by the National Institutes of Health [R21-AI (E. L.); K24-AI (E. L.)]. It was also supported in part by a Commonwealth University Health Enhancement (CURE) grant from the Pennsylvania Department of Health and by the CDC Prevention Epicenters Program [U54-CK (E. L.)]. The study sponsors had no role in study design, data collection, analysis or interpretation, or writing of the report. Transparency declarations None to declare. Author contributions The corresponding author had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: all authors. Acquisition of data: P. R. and I. N. Analysis and interpretation of data: P. R., W. B. B., J. A. R., J. P. M., T. E. Z. and E. L. Writing the manuscript: all authors. Obtained funding and study supervision: E. L. Supplementary data Table S1 and Table S2 are available as Supplementary data at JAC Online ( References 1 Hawkey PM, Jones AM. The changing epidemiology of resistance. J Antimicrob Chemother 2009; 64 Suppl 1: i Gales AC, Gordon KA, Wilke WW et al. Occurrence of single-point gyra mutations among ciprofloxacin-susceptible Escherichia coli isolates causing urinary tract infections in Latin America. Diagn Microbiol Infect Dis 2000; 36: McDonald LC, Chen FJ, Lo HJ et al. Emergence of reduced susceptibility and resistance to fluoroquinolones in Escherichia coli in Taiwan and contributions of distinct selective pressures. Antimicrob Agents Chemother 2001; 45: Shaheen BW, Boothe DM, Oyarzabal OA et al. Evaluation of the contribution of gyra mutation and efflux pumps to fluoroquinolone and multidrug resistance in pathogenic Escherichia coli isolates from dogs and cats. Am J Vet Res 2011; 72: Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: Twentieth Informational Supplement M100-S20. CLSI, Wayne, PA, USA, Lautenbach E, Metlay JP, Mao X et al. The prevalence of fluoroquinolone resistance mechanisms in colonizing Escherichia coli isolates recovered from hospitalized patients. Clin Infect Dis 2010; 51: Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008; 36: Landers T, Apte M, Hyman S et al. A comparison of methods to detect urinary tract infections using electronic data. Jt Comm J Qual Patient Saf 2010; 36: Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45: Quan H, Sundararajan V, Halfon P et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005; 43: Skiest DJ, Brown K, Cooper TW et al. Prospective comparison of methicillin-susceptible and methicillin-resistant community-associated 1550

5 High-MIC fluoroquinolone-susceptible Escherichia coli UTI JAC Staphylococcus aureus infections in hospitalized patients. J Infect 2007; 54: Camins BC, Farley MM, Jernigan JJ et al. A population-based investigation of invasive vancomycin-resistant Enterococcus infection in metropolitan Atlanta, Georgia, and predictors of mortality. Infect Control Hosp Epidemiol 2007; 28: Rattanaumpawan P, Tolomeo P, Bilker WB et al. Risk factors for fluoroquinolone resistance in Gram-negative bacilli causing healthcare-acquired urinary tract infections. J Hosp Infect 2010; 76: Agarwal R, Gupta D, Ray P et al. Epidemiology, risk factors and outcome of nosocomial infections in a Respiratory Intensive Care Unit in North India. J Infect 2006; 53: Goyal A, Prasad KN, Prasad A et al. Extended spectrum b-lactamases in Escherichia coli & Klebsiella pneumoniae & associated risk factors. Indian J Med Res 2009; 129: Gasink LB, Fishman NO, Nachamkin I et al. Risk factors for and impact of infection or colonization with aztreonam-resistant Pseudomonas aeruginosa. Infect Control Hosp Epidemiol 2007; 28: