UBC Faculty Research and Publications

Antimicrobial co-resistance patterns of gram-negative bacilli isolated from bloodstream infections: a… Wong, Patrick H; von Krosigk, Marcus; Roscoe, Diane L; Lau, Tim T; Yousefi, Masoud; Bowie, William R Oct 12, 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12879_2014_Article_3852.pdf [ 729.85kB ]
Metadata
JSON: 52383-1.0223638.json
JSON-LD: 52383-1.0223638-ld.json
RDF/XML (Pretty): 52383-1.0223638-rdf.xml
RDF/JSON: 52383-1.0223638-rdf.json
Turtle: 52383-1.0223638-turtle.txt
N-Triples: 52383-1.0223638-rdf-ntriples.txt
Original Record: 52383-1.0223638-source.json
Full Text
52383-1.0223638-fulltext.txt
Citation
52383-1.0223638.ris

Full Text

RESEARCH ARTICLEAntimicrobial co-resistancaeKeywords: Co-resistance, Multi-drug resistance, Gram-negative bacilli, Bloodstream infections, AntibiogramsWong et al. BMC Infectious Diseases 2014, 14:393http://www.biomedcentral.com/1471-2334/14/393patient safety and national security, and has publishedEast, VGH, 2733 Heather Street, Vancouver, BC V5Z 3J5, CanadaFull list of author information is available at the end of the articleBackgroundGram-negative bacilli (GNB) are a significant cause of in-fection in community and nosocomial settings [1]. Besidesinherent and chromosomally mediated mechanisms of re-sistance, the development of multidrug resistance inGNB is further facilitated by the acquisition of plasmids,integrons and transposons carrying resistance genes, whichis typically a consequence of selective antimicrobial pres-sure exerted by prolonged antibiotic use [2]. It is alsobecoming increasingly common to find multiple re-sistance genes that are linked together, thus antibioticsfrom unrelated classes may contribute to the selectivepressures and maintain the expression of these multidrugresistant genes [3,4].The Infectious Diseases Society of America views anti-microbial resistance as a serious threat to public health,* Correspondence: phpwong@mail.ubc.ca; william.bowie@ubc.ca1Division of Infectious Diseases, Department of Medicine, Faculty ofMedicine, The University of British Columbia (UBC), 452D, Heather Pavilionantimicrobial therapy.Patrick HP Wong1*, Marcus von Krosigk2, Diane L Roscoe3,4, Tim TY Lau1,2, Masoud Yousefi5 and William R Bowie1*AbstractBackground: Increasing multidrug resistance in gram-negative bacilli (GNB) infections poses a serious threat topublic health. Few studies have analyzed co-resistance rates, defined as an antimicrobial susceptibility profile in asubset already resistant to one specific antibiotic. The epidemiologic and clinical utility of determining co-resistancerates are analyzed and discussed.Methods: A 10-year retrospective study from 2002–2011 of bloodstream infections with GNB were analyzed fromthree hospitals in Greater Vancouver, BC, Canada. Descriptive statistics were calculated for antimicrobial resistanceand co-resistance. Statistical analysis further described temporal trends of antimicrobial resistance, correlations ofresistance between combinations of antimicrobials, and temporal trends in co-resistance patterns.Results: The total number of unique blood stream isolates of GNB was 3280. Increasing resistance to individualantimicrobials was observed for E. coli, K. pneumoniae, K. oxytoca, E. cloacae, and P. aeruginosa. Ciprofloxacinresistance in E. coli peaked in 2006 at 40% and subsequently stabilized at 29% in 2011, corresponding to decreasingciprofloxacin usage after 2007, as assessed by defined daily dose utilization data. High co-resistance rates wereobserved for ceftriaxone-resistant E. coli with ciprofloxacin (73%), ceftriaxone-resistant K. pneumoniae withtrimethoprim-sulfamethoxazole (83%), ciprofloxacin-resistant E. cloacae with ticarcillin-clavulanate (91%), andpiperacillin-tazobactam-resistant P. aeruginosa with ceftazidime (83%).Conclusions: Increasing antimicrobial resistance was demonstrated over the study period, which may partially beassociated with antimicrobial consumption. The study of co-resistance rates in multidrug resistant GNB providesinsight into the epidemiology of resistance acquisition, and may be used as a clinical tool to aid prescribing empiricgram-negative bacilli isolinfections: a longitudinalfrom 2002–2011© 2014 Wong et al.; licensee BioMed Central LCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.Open Accesse patterns ofted from bloodstreampidemiological studytd. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/4.0), which permits unrestricted use, distribution, andiginal work is properly credited. The Creative Commons Public Domaing/publicdomain/zero/1.0/) applies to the data made available in this article,Wong et al. BMC Infectious Diseases 2014, 14:393 Page 2 of 10http://www.biomedcentral.com/1471-2334/14/393policy recommendations for the US Congress to addressthe rising rates of antibiotic resistance together with de-clining approvals of new antibiotics [5]. Pop-Vicas andD’Agata (2005) have emphasized the need to further ex-pand our understanding of the dynamics of transmis-sion of these multidrug resistant pathogens to determinewhether it is related to cross transmission from humansto humans or antimicrobial selective pressures [6]. Kallenand Srinivasan (2010) have also highlighted the import-ance for ongoing surveillance of the incidence and epi-demiology of multidrug resistant GNB [1].Several studies have shown an increasing incidence ofmultidrug resistant GNB [7-9]. However, to our know-ledge, few studies provide a comprehensive evaluation ofco-resistance patterns with antimicrobial agents com-monly used to treat GNB. Co-resistance is defined in thisstudy as the antimicrobial susceptibility profile in a subsetof isolates already resistant to a specific antibiotic, andprovides a different means for monitoring multidrug re-sistance and displaying observed trends, such as increasingmultidrug resistance in P. aeruginosa that are resistant tociprofloxacin [10-12]. The study of co-resistance can thusbe quite broad and not limited to isolates that are known toharbour multidrug resistance, such as those with ESBLs.From a clinical perspective, co-resistance should be takeninto consideration when prescribing empiric therapy for pa-tients being treated for specific GNB where local antimicro-bial resistance rates are significant, and in patients whohave been exposed to prior courses of antimicrobial agents.The aim of our study is to examine the antimicrobialco-resistance patterns of GNB bloodstream isolates over a10-year time period in order to document changes in sus-ceptibility patterns, and to identify any potential causeswhen significant changes occur. The utility of the informa-tion gathered would be applicable in clinical practice whentailoring empiric antibiotic treatment.MethodsWe conducted a 10-year retrospective study from Jan2002 to Dec 2011 to quantify the trends of resistance andco-resistance patterns in GNB isolated from bloodstreaminfections at Vancouver Coastal Health (VCH), which in-cludes Vancouver General Hospital (VGH) (a 950-bed ter-tiary care teaching hospital with 21,000 admissions peryear), Richmond Hospital (a 175-bed community hospital),and Lion's Gate Hospital (a 268-bed community hospital).The study was approved by the University of BritishColumbia Clinical Research Ethics Board and by the Van-couver Coastal Health Research Institute.Utilizing the VCH laboratory information system,Sunquest®, which stores laboratory data in the Sunset® data-base, a report was generated to identify all positive bloodcultures with GNB from 2002–2011. Further refinementwas performed through computer programming to includeonly one bacterial isolate of the same identification (genusand species) and susceptibility pattern per patient per cal-endar year in the analysis. The first unique patient isolatefrom the first admission during each study year was usedin cases where multiple isolates of the same genus, speciesand susceptibility pattern were identified. When multiplecultures taken from the same patient had the same GNBidentified but with any difference in susceptibility patterns,these were considered unique isolates and were includedin the analysis. The Sunquest® system started to incorpor-ate data for Richmond Hospital in April 2005 and Lion'sGate Hospital in August 2007. Annual review of the anti-microbial susceptibility patterns from the individual hospi-tals in this study do not differ significantly from each otherand inclusion of additional sites over the study period wasnot felt to influence results.Identification and susceptibility testing were performedusing commercially available automated systems [BDPhoenix™ (Sep 2009 to study end date) or Siemens Micro-Scan® (2002 to Sept 2009) systems]. Quality controlmeasures were implemented during the changeover ofautomated systems in 2009 for validation and verification toensure that the results on both systems were concordantwithin acceptable limits. Antibiotic susceptibility testing wasperformed and interpreted using guidelines and interpretivebreakpoints for susceptible, intermediate, and resistantcategories established by the Clinical and Laboratory Stan-dards Institute (CLSI). All antimicrobial susceptibility resultsthat fell into the intermediate category were presumed to beresistant for the purposes of this study. Guidelines for sus-ceptibility breakpoints for cephalosporins were changedafter 2010, but the laboratory did not change to the newguidelines since the panels for the commercial system inuse did not have concentration wells that were low enoughto allow interpretation at the lower breakpoints [13].The data collected was extrapolated from Sunset® into aMicrosoft Access® database. Queries and reports of indi-vidual antimicrobial resistance and co-resistance patternswere generated to supply descriptive statistics for eachspecific study year and pathogen. Pharmacy data oninpatient antimicrobial consumption at VGH in the formof defined daily doses (DDDs) were compared againstantimicrobial resistance results. Statistical analysis wasperformed to determine linear-to-linear temporal trendsof individual antimicrobial resistance patterns. In addition,Pearson correlation coefficients were determined for asso-ciations between resistant combinations of antimicrobialagents, and Pearson Chi-square test was performed toassess for temporal trends in co-resistance patterns. P valuesof ≤0.05 were used for statistical significance in all cases.ResultsThe total number of unique blood stream isolates ofGNB was 3,280 over the 10-year study period, rangingannually from 109 (2002) to 463 (2010). The five mostcommon GNB isolated were Escherichia coli, Klebsiellapneumoniae, Klebsiella oxytoca, Enterobacter cloacae, andPseudomonas aeruginosa, and the proportion of each is il-lustrated in Figure 1. The temporal trends of resistance tocommonly used antibiotics for each of these organismsare illustrated by line graphs in Figure 2.For E. coli (Figure 2a), overall trends of increasing resist-ance to cefazolin and ceftriaxone were observed over time.Resistance to piperacillin-tazobactam remained stable at4% from 2009 to 2011; it replaced ticarcillin-clavulanateon the hospital formulary in 2009. Resistance to ciproflox-acin and trimethroprim-sulfamethoxazole (TMP-SMX)both peaked during 2006 at roughly 40% and stabilized atlower levels of 29% and 34% respectively in 2011. Compar-ing this with antimicrobial consumption data (Figure 3),use of ciprofloxacin peaked in 2007 at 23,800 DDD anddecreased to 10,100 DDD in 2011, whereas TMP-SMXconsumption peaked in 2007 at 5,100 DDD, dropped to4,200 DDD in 2009, and has increased back to 6,700 DDDWong et al. BMC Infectious Diseases 2014, 14:393 Page 3 of 10http://www.biomedcentral.com/1471-2334/14/393in 2011. A subgroup analysis of ESBL E. coli (total of 54isolates in 2010 and 2011) revealed the following resist-ance rates: ciprofloxacin (83%), piperacillin-tazobactam(17%), TMP-SMX (76%), gentamicin (52%), ceftriaxone(100%), and imipenem (0%). In contrast, non-ESBL E. coli(total of 414 isolates in 2010 and 2011) had lower resist-ance rates: ciprofloxacin (21%), piperacillin-tazobactam(2%), TMP-SMX (26%), gentamicin (8%), ceftriaxone (3%),and imipenem (0%).For K. pneumoniae (Figure 2b), overall trends of increas-ing resistance were observed with cefazolin, ceftriaxoneFigure 1 Distribution of gram-negative bacilli isolated frombloodstream isolates from 2002–2011. The total number ofunique bloodstream isolates during the study period was 3280.and piperacillin-tazobactam. Subgroup analysis for ESBLK. pneumoniae (total of 12 isolates in 2010 and 2011) re-vealed the following resistance rates: ciprofloxacin (50%),piperacillin-tazobactam (42%), TMP-SMX (50%), gentami-cin (42%), ceftriaxone (100%), and imipenem (0%). In con-trast, non-ESBL K. pneumoniae (total of 118 isolates in2010 and 2011) had lower resistance rates: ciprofloxacin(10%), piperacillin-tazobactam (4%), TMP-SMX (8%), gen-tamicin (2%), ceftriaxone (2%), and imipenem (1%).For K. oxytoca (Figure 2c), increasing rates of resist-ance to ceftriaxone, ciprofloxacin, TMP-SMX, and piperacillin-tazobactam were observed. For E. cloacae (Figure 2d), re-sistance rates increased with ceftriaxone, TMP-SMX, andciprofloxacin, but remained relatively stable with piperacillin-tazobactam. For P. aeruginosa (Figure 2e), resistance ratesto ceftazidime, ciprofloxacin, and piperacillin-tazobactamall increased. P. aeruginosa resistance to imipenem was4.2% during 2007–2011 (from a baseline of 0% in 2002–2006).Chi-square trend tests were performed to determinewhich temporal patterns of antimicrobial resistance fitinto a linear-by-linear association model. Linear-by-lineartrends were identified with cefazolin resistance in E. coli(X2 = 9.062, p < 0.003), ceftriaxone resistance in E. coli (X2 =13.070, p < 0.001), cefazolin resistance in K. pneumoniae(X2 = 15.183, p < 0.001), ceftriaxone resistance in K.pneumoniae (X2 = 18.066, p < 0.001), piperacillin-tazobactamresistance in K. pneumoniae (X2 = 5.485, p < 0.019), andcefazolin resistance in K. oxytoca (X2 = 4.329, p < 0.037).Other temporal patterns that approached significance in alinear-by-linear association model include ciprofloxacinresistance in K. pneumoniae (X2 = 3.656, p = 0.056), TMP-SMX resistance in K. oxytoca (X2 = 3.576, p = 0.059),ticarcillin-clavulanate resistance in K. pneumoniae (X2 =3.716, p = 0.054), and ceftazidime resistance in P. aerugi-nosa (X2 = 3.213, p = 0.073). The temporal trend modelsfor E. coli resistant to ciprofloxacin and SMX-TMP arecurvilinear with calculated departure of linearity as a chi-square for ciprofloxacin is 22.611 on 8df (p < 0.0039), andfor TMP-SMX is 15.31 on 8df (p = 0.053).The second part of the analysis involved looking atcorrelation patterns of resistance between various pairsof antibiotics, which is commonly reported in other pub-lications but does not provide information regarding co-resistance as defined in this study. In E. coli (Table 1a),high correlations of resistance were identified betweenseveral combinations of antimicrobial agents includingpiperacillin-tazobactam with ciprofloxacin (r = 0.973),ticarcillin-clavulanate with cefazolin (r = 0.885), TMP-SMX with ciprofloxacin (r = 0.871), ticarcillin-clavulanatewith TMP-SMX (r = 0.839), and ceftriaxone with cefazolin(r = 0.808). In K. pneumoniae (Table 1b), highest correlationsof resistance were seen between piperacillin-tazobactam withceftriaxone (r = 0.982), ceftriaxone with cefazolin (r = 0.935),Figure 2 Temporal patterns of resistance to selected antimicrobial agents for a) E. coli, b) K. pneumoniae, c) K. oxytoca, d) E. cloacae,and e) P. aeruginosa. Antimicrobial resistance rates were analyzed yearly for E. coli and K. pneumoniae, and combined into two time groupsof 2002–06 and 2007–11 for K. oxytoca, E. cloacae, and P. aeruginosa due to lower isolate numbers. Number of isolates for each GNB is listedbelow the corresponding year or time groups. Y-axis represents resistance in percentage. Note: Pip-tazo is piperacillin-tazobactam,TMP-SMX is trimethoprim-sulfamethoxazole, Tic-clav is ticarcillin-clavulanate.Figure 3 Inpatient antimicrobial utilization at Vancouver General Hospital in defined daily dose (DDD) by calendar year. DDD is listedon Y-axis. Note: Pip-tazo is piperacillin-tazobactam, TMP-SMX is trimethoprim-sulfamethoxazole, Tic-clav is ticarcillin-clavulanate.Wong et al. BMC Infectious Diseases 2014, 14:393 Page 4 of 10http://www.biomedcentral.com/1471-2334/14/393el. cWong et al. BMC Infectious Diseases 2014, 14:393 Page 5 of 10http://www.biomedcentral.com/1471-2334/14/393Table 1 Pearson correlation matrixes showing Pearson corrantibiotics for a) E. coli, b) K. pneumonia, c) K. oxytoca, d) Ea) Pearson correlation matrix (r) for E. coliand ticarcillin-clavulanate with cefazolin (r = 0.893). InK. oxytoca (Table 1c), highest correlations of resistancewere between ciprofloxacin with ceftriaxone (r = 0.952),ticarcillin-clavulanate with ceftriaxone (r = 0.773), andticarcillin-clavulanate with ciprofloxacin (r = 0.773). InCefazolin CeftriaxoneCefazolin (n = 1783)Ceftriaxone (n = 1783) 0.808Ciprofloxacin (n = 1783) 0.487 0.290TMP-SMX (n = 1783) 0.770 0.503Pip-tazo (n = 846) 0.702 0.623Tic-clav (n = 1063) 0.885 0.299b) Pearson correlation matrix (r) for K. pneumoniaeCefazolin CeftriaxoneCefazolin (n = 480)Ceftriaxone (n = 480) 0.935Ciprofloxacin (n = 480) 0.630 0.566TMP-SMX (n = 480) 0.719 0.689Pip-tazo (n = 163) 0.917 0.982Tic-clav (n = 295) 0.893 0.677c) Pearson correlation matrix (r) for K. oxytocaCefazolin CeftriaxoneCefazolin (n = 127)Ceftriaxone (n = 127) 0.576Ciprofloxacin (n = 127) 0.619 0.952TMP-SMX (n = 127) −0.598 −0.005Pip-tazo (n = 71) 0.011 −0.222Tic-clav (n = 65) 0.649 0.773d) Pearson correlation matrix (r) for E. cloacaeCeftriaxone CiprofloxacinCeftriaxone (n = 170)Ciprofloxacin (n = 170) 0.688TMP-SMX (n = 170) 0.611 0.172Pip-tazo (n = 83) 0.927 0.542Tic-clav (n = 96) 0.944 0.613e) Pearson correlation matrix (r) for P. aeruginosaCeftazidime CiprofloxacinCeftazidime (n = 115)Ciprofloxacin (n = 115) −0.457Pip-tazo (n = 60) 0.177 0.225Tic-clav (n = 65) 0.888 −0.100Number of isolates (n) included in each matrix is listed for each row. Pearson cothose that are bolded and italicized trended towards significance with p values >piperacillin-tazobactam, Tic-clav is ticarcillin-clavulanate; nd is not determined assusceptibilities reported.ation coefficients (r) of resistance between pairs ofloacae, e) P. aeruginosaE. cloacae (Table 1d), highest correlations of resistancewere observed between ticarcillin-clavulanate with ceftri-axone (r = 0.944), piperacillin-tazobactam with ceftriaxone(r = 0.927), and ticarcillin-clavulanate with TMP-SMX(r = 0.781). Finally, in P. aeruginosa (Table 1e), resistanceCiprofloxacin TMP-SMX Pip-tazo Tic-clav0.8710.973 0.8650.733 0.839 ndCiprofloxacin TMP-SMX Pip-tazo Tic-clav0.2740.886 0.7780.648 0.635 ndCiprofloxacin TMP-SMX Pip-tazo Tic-clav−0.184−0.349 −0.1740.773 0.013 ndTMP-SMX Pip-tazo Tic-clav0.7810.781 ndPip-tazo Tic-clavndefficients (r) which are bolded represent pairs with p values ≤ 0.05, while0.05 ≤ 0.07. Note: TMP-SMX is trimethoprim-sulfamethoxazole, Pip-tazo isisolates did not consistently have both Pip-tazo and Tic-clavWong et al. BMC Infectious Diseases 2014, 14:393 Page 6 of 10http://www.biomedcentral.com/1471-2334/14/393to ticarcillin-clavulanate was highly correlated with ceftaz-idime resistance (r = 0.888).The final part of our analysis involved selecting out agroup of isolates that were resistant to one specific anti-microbial agent and then examining that group's anti-microbial susceptibility profile–the co-resistance pattern.In E. coli (Table 2a), those that are resistant to ceftriax-one have a 73% probability of also exhibiting resistanceto ciprofloxacin. In contrast, for those that are resistantto ciprofloxacin, only 25% are also resistant to ceftriax-one. Other high co-resistance rates found in E. coli werepiperacillin-tazobactam-resistant strains that are co-resistantwith TMP-SMX (77%), piperacillin-tazobactam-resistantstrains that are co-resistant with ciprofloxacin (71%), andceftriaxone-resistant strains that are co-resistant withTMP-SMX (69%). To assess temporal patterns, PearsonChi-Square tests were performed to analyze whether co-resistance rates remained stable or changed significantlyover time. Over time comparing the periods of 2002–2006and 2007–2011, the rate of ceftriaxone-resistant E. colithat are co-resistant with ciprofloxacin has been stablearound 73% (X2 = 0.005, p = 0.941) suggesting no signifi-cant change. However, the rate of ciprofloxacin-resistantE. coli that are co-resistant with ceftriaxone has increasedfrom 16% to 30% (X2 = 12.103, p < 0.001).For K. pneumoniae (Table 2b), high co-resistance rateswere seen with ceftriaxone-resistant strains that are co-resistant with TMP-SMX (83%), ceftriaxone-resistantstrains that are co-resistant with ciprofloxacin (60%),and piperacillin-tazobactam-resistant strains that areco-resistant with ceftriaxone (60%). For K. oxytoca (Table 2c),high co-resistance was seen between ciprofloxacin-resistantstrains with ceftriaxone (60%), however it is important tonote that only 5 isolates of K. oxytoca were analyzed.For E. cloacae (Table 2d), highest co-resistance rates wereobserved with piperacillin-tazobactam-resistant strainsthat are co-resistant with ceftriaxone (100%), ciprofloxacin-resistant strains that are co-resistant with ticarcillin-clavulanate (91%), ceftriaxone-resistant strains that areco-resistant with ticarcillin-clavulanate (81%), and TMP-SMX resistant strains that are co-resistant with ticarcillin-clavulanate (80%).In P. aeruginosa (Table 2e), high co-resistance rates wereobserved with piperacillin-tazobactam-resistant strains thatare co-resistant with ceftazidime (83%), and piperacillin-tazobactam-resistant strains that are co-resistant withciprofloxacin (67%). It has been previously observed thatP. aeruginosa resistant to fluoroquinolones are oftenassociated with resistance to other antibiotic classes[10,12]. Interestingly, our results do not suggest a particu-larly high co-resistance profile in ciprofloxacin-resistantP. aeruginosa with co-resistance rates of 23-24% withgentamicin, imipenem, ceftazidime, and piperacillin-tazobactam.DiscussionTo our knowledge, this is one of the largest studies com-prehensively evaluating antimicrobial co-resistance in GNBisolated from bloodstream infections. In the first partof our study, an unanticipated finding in our resultswas that both ciprofloxacin and TMP-SMX resistancepeaked in 2006 for E. coli, and has subsequently de-creased and stabilized.Our experience at VGH is that there had previouslybeen heavy consumption of ciprofloxacin during thetime period when the resistance peaked, which suggeststhis selective pressure may have contributed to this ob-servation. We reviewed consumption data of antibioticsat VGH based on inpatient utilization data (Figure 3),and determined that the (DDD) of ciprofloxacin wasstable at 23,000-24,000 from 2002–2004, decreasing to17,000 in 2005, increasing back to 23,800 in 2007, andhas since decreased to 10,100 in 2011. In comparison forceftriaxone utilization, the DDD has continually increasedfrom 3,300 to 7,200 from 2002 to 2011. The utilizationdata supports that resistance rates of E. coli to ciprofloxa-cin and ceftriaxone correlate with consumption patternsof antibiotics during that period, which peaked in 2007and 2011, respectively (Figure 2a). The DDD for TMP-SMX is also of interest in that there were two consump-tion peaks – 5,100 in 2007 and 6,700 in 2011, which againcorrelates with resistance peaks in our E. coli resistancedata and provides further support that resistance pat-terns may be associated with antimicrobial consump-tion (Figure 2a). As most of the GNB isolates werecollected from VGH, the changes in resistance patternsare most attributable to the antimicrobial consumption atthat site. In general, the antimicrobial consumption datacollected from VGH is representative of the usage at theother two smaller hospitals sites.When looking at antimicrobial resistance patterns ofE. coli in the community collected by BC Biomedical La-boratories (45 community-based patient services centresin Greater Vancouver, BC), we observe a significantjump in the resistance rates of E. coli to ciprofloxacinbetween 2002 and 2007 (10% to 22%) [14]. It is notedfrom the BC Centre for Disease Control publication thatthere was no E. coli data published for the years 2003–2006 because antibiograms were not produced when theresistance rates remained relatively similar, suggestingthat the jump in resistance may have occurred in 2007.However, unlike our data after 2007, the level of resist-ance remained relatively stable until a further increasingtrend in 2010 (26%) and 2011 (27%) [14]. One explan-ation for this observed difference is that data from BCBiomedical Laboratories is reflective of an outpatient-based setting, where ciprofloxacin continues to be heav-ily utilized for urinary tract infections. Fortunately dueto concerns regarding resistance, the 2010 InfectiousTable 2 Summary of antimicrobial co-resistance rates determined for a) E. coli, b) K. pneumoniae, c) K. oxytoca,d) E. cloacae, and e) P. aeruginosaa) E. coliNo. of isolates (n) Resistance to Co-resistance rate with antibiotic (%)2002-06 Ceftriaxone Ciprofloxacin Pip-tazo TMP-SMX Tic-clav44 Ceftriaxone 73 14 59 59194 Ciprofloxacin 16 8 62 313 Pip-tazo 33 67 67 nd193 TMP-SMX 13 63 8 3191 Tic-clav 29 67 nd 662007-11146 Ceftriaxone 73 18 72 48357 Ciprofloxacin 30 10 58 2628 Pip-tazo 57 71 79 nd373 TMP-SMX 28 56 10 2764 Tic-clav 42 61 nd 612002-11190 Ceftriaxone 73 18 69 53551 Ciprofloxacin 25 10 60 2931 Pip-tazo 55 71 77 nd566 TMP-SMX 23 58 9 30155 Tic-clav 34 65 nd 64b) K. pneumoniaeNo. of isolates (n) Resistance to Co-resistance rate with antibiotic (%)2002-2011 Ceftriaxone Ciprofloxacin Pip-tazo TMP-SMX Tic-clav20 Ceftriaxone 60 35 83 5354 Ciprofloxacin 22 22 46 2910 Pip-tazo 60 50 40 nd44 TMP-SMX 30 57 16 3719 Tic-clav 26 58 nd 53c) K. oxytocaNo. of isolates (n) Resistance to Co-resistance rate with antibiotic (%)2002-2011 Ceftriaxone Ciprofloxacin Pip-tazo TMP-SMX Tic-clav8 Ceftriaxone 38 20 25 505 Ciprofloxacin 60 0 20 252 Pip-tazo 50 0 0 nd5 TMP-SMX 40 20 0 1003 Tic-clav 67 33 nd 33d) E. cloacaeNo. of isolates (n) Resistance to Co-resistance rate with antibiotic (%)2002-2011 Ceftriaxone Ciprofloxacin Pip-tazo TMP-SMX Tic-clav43 Ceftriaxone 26 47 37 8117 Ciprofloxacin 65 17 59 919 Pip-tazo 100 11 11 nd24 TMP-SMX 67 42 6 8033 Tic-clav 67 30 nd 24Wong et al. BMC Infectious Diseases 2014, 14:393 Page 7 of 10http://www.biomedcentral.com/1471-2334/14/393mCeptthe-claWong et al. BMC Infectious Diseases 2014, 14:393 Page 8 of 10http://www.biomedcentral.com/1471-2334/14/393Disease Society of America (IDSA) guideline for uncom-plicated cystitis now recommends fluoroquinolone anti-biotics as alternative agents only when other urinary tractinfection therapies cannot be used [15]. As seen by ourutilization data at VGH, in the hospital setting where pa-tients are more ill in general, physicians are now prescrib-ing less ciprofloxacin as empiric antimicrobial therapybecause of the high rates of resistance in both inpatientand outpatient settings.A further clinical utility of determining co-resistancerates is to aid in the selection of empiric antimicrobialcoverage. For example, patients with suspected urinarytract infection are often started on TMP-SMX whilepending culture results. However, if the patient deterio-rates and is not responding clinically, one would have todecide which antibiotic to use to broaden the coverage.In the periods from 2002–2006, co-resistance of E. colito ceftriaxone was only 13%, so it may have been appro-priate to switch to this antibiotic. However, co-resistanceof E. coli to ceftriaxone was 28% in 2007–2011, thus itTable 2 Summary of antimicrobial co-resistance rates deterd) E. cloacae, and e) P. aeruginosa (Continued)e) P. aeruginosaNo. of isolates (n) Resistance to2002-2011 Ceftazidime11 Ceftazidime22 Ciprofloxacin 236 Pip-tazo 8313 Imipenem 239 Gentamicin 11Co-resistance rates were analyzed as a summary of isolates from 2002–2011 excand 2007–11 due to higher isolate numbers. Co-resistance rates are rounded toPip-tazo is piperacillin-tazobactam, TMP-SMX is trimethoprim-sulfamethoxazole, Tichave both Pip-tazo and Tic-clav susceptibilities reported.may be more reasonable based on our results to switch topiperacillin-tazobactam which has a co-resistance rate ofonly 10%. This is of particular importance with the ad-vances in bacterial identification techniques, such asMatrix Assisted Laser Desorption Ionization – Time ofFlight (MALDI-TOF) where bacterial species are identi-fied on average 1.45 days earlier than traditional methods[16]. The limitation of this technique is that susceptibilitytesting is still required, further separating the time gapbetween identification of bacteria to reporting of sus-ceptibility results. Knowledge of local antimicrobial re-sistance and co-resistance rates becomes more critical inprescribing empiric antimicrobial therapy during thisperiod. In addition, for initial empiric therapy in criticallyill patients one may want to consider combination anti-microbial coverage, especially in patients at high riskfor acquiring multidrug resistant organisms. With co-resistance data, one can calculate resistant rates ofusing various combinations of empiric antimicrobial treat-ment. For example, 15% of P. aeruginosa isolates areresistant to gentamicin (2011), and gentamicin-resistantP. aeruginosa that are co-resistant with piperacillin-tazobactam, imipenem and ceftazidime are 29, 33 and11%, respectively. The resistance rates of utilizing combin-ation gentamicin with piperacillin-tazobactam (0.15 × 0.29),imipenem (0.15 × 0.33) or ceftazidime (0.15 × 0.11) in initialmanagement are 4.4, 5.0, and 1.6%, respectively. This is aninteresting finding as one may not intuitively expect thelower resistance rates with the utilization of combinationtherapy with ceftazidime versus imipenem.A major strength of our study is that the isolates wereincluded from a diverse area within the Greater VancouverRegional District, which included the cities of Vancouver,Richmond, and North Vancouver. The inclusion criteriawere stringent to ensure that duplicate samples from thesame patient were not included in the analysis. Our use ofonly bloodstream isolates allowed us to reliably negateeffects from other possible ‘contaminant’ specimens, asined for a) E. coli, b) K. pneumoniae, c) K. oxytoca,Co-resistance rate with antibiotic (%)iprofloxacin Pip-tazo Imipenem Gentamicin45 50 27 924 23 2367 33 3338 18 2356 29 33for E. coli, where it was possible to analyze two time periods from 2002–06nearest percent, and those that are ≥50% are highlighted in bold font. Note:v is ticarcillin-clavulanate; nd is not determined as isolates did not consistentlyGNB isolated from blood samples are almost invariably in-dicative of true infections. Patients with positive blood cul-tures are generally more acutely ill and would benefit mostfrom the clinical utility of using co-resistance data in theselection of an empiric antimicrobial regimen. One limita-tion of this study is that we were unable to determinewhether these bloodstream infections were community- ornosocomially-acquired. Due to the low number of isolatesfor K. oxytoca, E. cloacae, and P. aeruginosa, the resistancedata had to be grouped into two time points, which re-duced the ability to determine annual variations that mayhave occurred. Another limitation was that we were un-aware of whether a patient received any recent or concur-rent antimicrobial therapies when blood cultures weredrawn, which may have yielded additional information re-garding the incidence of failed empiric therapy. In addition,we did not explicitly determine the source of bacteremia,and so are unable to tell if resistance strains and patternsintermediate susceptibility would generally not be used ifmost of the antibiotic agents.Wong et al. BMC Infectious Diseases 2014, 14:393 Page 9 of 10http://www.biomedcentral.com/1471-2334/14/393Besides ongoing surveillance of co-resistance patterns,one potential application for the future would be to lookat co-resistance patterns in various medical services andward locations. It has been proposed that unit-based anti-biograms and combination antibiograms may be more use-ful than national or hospital-specific antibiograms [17].Unit-based antibiograms refers to cumulative antibioticsusceptibility reports for patients in a particular ward overa specified period of time, whereas combination antibio-grams provide information on percent susceptibility toother antibiotics if it is resistant to one particular antibiotic[17], or essentially the co-resistance rate that we investi-gated in this study. Unit-based combination antibiogramscan be considered for certain wards, such as the leukemia/bone marrow transplant ward where patients can be critic-ally ill, as this population may benefit from combinationantibiotic empiric coverage depending on local susceptibil-ity patterns. The ability of MALDI-TOF technology to ex-pedite identification of pathogens emphasizes the need andimportance of having accurate and up to date antimicro-bial susceptibility data.ConclusionsIn summary, increasing antimicrobial resistance was ob-served for several GNB over a 10-year period, and maypartially be associated with antimicrobial consumption.The study of co-resistance rates in multidrug resistantGNB may provide further insight into the epidemiologyof resistance acquisition. Further applications for co-resistance data include its utility as a clinical tool to aidin the prescription of empiric antimicrobial therapy.Competing interestsAll authors have completed ICMJE forms and disclose no potential financialor non-financial competing interest.Authors’ contributionsPW participated in study design, interpreted data analysis, producedgraphical illustrations of data, and drafted the manuscript. MVK participatedin study design, and acquired and analyzed majority of data. DR participatedin study design, provided raw data, and has been involved in revising themanuscript. TL participated in study design, provided raw data, and hasbeen involved in revising the manuscript. MY interpreted the data andthere were other effective alternatives. Overall, the inci-dence of intermediate susceptible isolates was generallylow during the study period ranging from 0% to 2% forwere more commonly associated with certain types of in-fections. Although we classified all intermediate susceptibleisolates as being resistant which would seemingly elevatethe overall resistance rates, this methodology is consistentwith clinical practice in which an antimicrobial agent withperformed the statistical analysis. WB participated in the design andcoordination of the study, and helped draft and revise the manuscript. Allauthors have read and approved the final manuscript.AcknowledgementsWe would like to thank Salomeh Shajari for compiling data on DDD forantimicrobial utilization at VGH over the study period.FundingThis study was an unfunded project.Author details1Division of Infectious Diseases, Department of Medicine, Faculty ofMedicine, The University of British Columbia (UBC), 452D, Heather PavilionEast, VGH, 2733 Heather Street, Vancouver, BC V5Z 3J5, Canada. 2Faculty ofPharmaceutical Sciences, UBC, Vancouver, Canada. 3Department of Pathologyand Laboratory Medicine, Faculty of Medicine, UBC, Vancouver, Canada.4Division of Medical Microbiology and Infection Control, Vancouver GeneralHospital (VGH), Vancouver, Canada. 5Brain Research Centre, UBC, Vancouver,Canada.Received: 17 March 2014 Accepted: 10 July 2014Published: 12 October 2014References1. Kallen AJ, Srinivasan A: Current epidemiology of multidrug-resistantgram-negative bacilli in the United States. Infect Control Hosp Epidemiol2010, 31(Suppl 1):S51–54.2. D'Agata EM: Rapidly rising prevalence of nosocomial multidrug-resistant,Gram-negative bacilli: a 9-year surveillance study. Infect Control HospEpidemiol 2004, 25(10):842–846.3. Gootz TD: The forgotten Gram-negative bacilli: what genetic determinantsare telling us about the spread of antibiotic resistance. Biochem Pharmacol2006, 71(7):1073–1084.4. Courvalin P, Trieu-Cuot P: Minimizing potential resistance: the molecularview. Clin Infect Dis 2001, 33(Suppl 3):S138–146.5. Spellberg B, Blaser M, Guidos RJ, Boucher HW, Bradley JS, Eisenstein BI,Gerding D, Lynfield R, Reller LB, Rex J, Schwartz D, Septimus E, Tenover FC,Gilbert D: Combating antimicrobial resistance: policy recommendationsto save lives. Clin Infect Dis 2011, 52(Suppl 5):S397–428.6. Pop-Vicas AE, D'Agata EM: The rising influx of multidrug-resistant gram-negativebacilli into a tertiary care hospital. Clin Infect Dis 2005, 40(12):1792–1798.7. McGowan JE Jr: Resistance in nonfermenting gram-negative bacteria:multidrug resistance to the maximum. Am J Med 2006,119(6 Suppl 1):S29–36. discussion S62-70.8. Lockhart SR, Abramson MA, Beekmann SE, Gallagher G, Riedel S, DiekemaDJ, Quinn JP, Doern GV: Antimicrobial resistance among Gram-negativebacilli causing infections in intensive care unit patients in the UnitedStates between 1993 and 2004. J Clin Microbiol 2007, 45(10):3352–3359.9. Andrade SS, Jones RN, Gales AC, Sader HS: Increasing prevalence ofantimicrobial resistance among Pseudomonas aeruginosa isolates inLatin American medical centres: 5 year report of the SENTRYAntimicrobial Surveillance Program (1997–2001). J AntimicrobChemother 2003, 52(1):140–141.10. Trouillet JL, Vuagnat A, Combes A, Kassis N, Chastre J, Gibert C:Pseudomonas aeruginosa ventilator-associated pneumonia: comparisonof episodes due to piperacillin-resistant versus piperacillin-susceptibleorganisms. Clin Infect Dis 2002, 34(8):1047–1054.11. Paramythiotou E, Lucet JC, Timsit JF, Vanjak D, Paugam-Burtz C, Trouillet JL,Belloc S, Kassis N, Karabinis A, Andremont A: Acquisition of multidrug-resistant Pseudomonas aeruginosa in patients in intensive care units:role of antibiotics with antipseudomonal activity. Clin Infect Dis 2004,38(5):670–677.12. Kriengkauykiat J, Porter E, Lomovskaya O, Wong-Beringer A: Use of an effluxpump inhibitor to determine the prevalence of efflux pump-mediatedfluoroquinolone resistance and multidrug resistance in Pseudomonasaeruginosa. Antimicrob Agents Chemother 2005, 49(2):565–570.13. CLSI: Performance Standards for Antimicrobial Susceptibility Testing:Twentieth Informational Supplement. CLSI document M100-S20. Wayne,PA: Clinical and Laboratory Standards Institute; 2010.14. BC Centre for Disease Control: Antimicrobial Resistance Trends in theProvince of British Columbia. 2012:1–52 [http://www.bccdc.ca/util/about/annreport/default.htm#heading6]15. Gupta K, Hooton TM, Naber KG, Wullt B, Colgan R, Miller LG, Moran GJ,Nicolle LE, Raz R, Schaeffer AJ, Soper DE: International clinical practiceguidelines for the treatment of acute uncomplicated cystitis andpyelonephritis in women: A 2010 update by the Infectious DiseasesSociety of America and the European Society for Microbiology andInfectious Diseases. Clin Infect Dis 2011, 52(5):e103–120.16. Tan KE, Ellis BC, Lee R, Stamper PD, Zhang SX, Carroll KC: Prospectiveevaluation of a matrix-assisted laser desorption ionization-time of flightmass spectrometry system in a hospital clinical microbiology laboratoryfor identification of bacteria and yeasts: a bench-by-bench study forassessing the impact on time to identification and cost-effectiveness.J Clin Microbiol 2012, 50(10):3301–3308.17. Paterson DL: Impact of antibiotic resistance in gram-negative bacillion empirical and definitive antibiotic therapy. Clin Infect Dis 2008,47(Suppl 1):S14–20.doi:10.1186/1471-2334-14-393Cite this article as: Wong et al.: Antimicrobial co-resistance patterns ofgram-negative bacilli isolated from bloodstream infections: a longitudinalepidemiological study from 2002–2011. BMC Infectious Diseases 2014 14:393.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionWong et al. BMC Infectious Diseases 2014, 14:393 Page 10 of 10http://www.biomedcentral.com/1471-2334/14/393Submit your manuscript at www.biomedcentral.com/submit

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0223638/manifest

Comment

Related Items