Antibacterial Resistance In Wales

Transcription

1 A Report from Public Health Wales Healthcare Associated Infection, Antimicrobial Resistance & Prescribing Programme (HARP team) Antibacterial Resistance In Wales Authors: Maggie Heginbothom, Robin Howe & Eleri Davies Date: 01/06/2018 Status: Final Version 1: Antibacterial Resistance in Wales Page: 1 of 54 HARP team 1

2 Table of Contents Table of Contents... 2 Section 1: Introduction... 3 Section 2: Key points of interest... 4 Section 3: Methods... 6 Resistance data...6 Section 4: Monitoring Trends in Resistance... 9 UK 5 Year Antimicrobial Resistance Strategy... 9 Background...9 APRHAI Primary Data Set Section 5.1: Antimicrobial resistance rates for the most common organisms causing bacteraemia Background Escherichia coli (n=2,663 in 2017) Klebsiella spp. (n=549 in 2017) Enterobacter spp., Serratia spp., Proteus spp., and Ps. aeruginosa Enterobacter spp. (n=115 in 2017) Serratia spp. (n=91 in 2017) Proteus spp. (n=223 in 2017) Pseudomonas aeruginosa Staphylococcus aureus Meticillin Sensitive Staphylococcus aureus (n=834 in 2017) Meticillin Resistant Staphylococcus aureus (n=123 in 2017) Enterococcus spp. (n=449 in 2017) Streptococcus pneumoniae (n=406 in 2017) Section 5.2: Antimicrobial resistance rates for urinary coliforms Community Urinary Coliforms (n=100,578 in 2017) Outpatient Urinary Coliforms (n=8,512 in 2017) Inpatient Urinary Coliforms (n=17,986 in 2017) Section 5.3: Antimicrobial resistance rates for Staphylococcus aureus MSSA ( all locations n=23,519 in 2017) MRSA (all locations n=2,439 in 2017) Section 5.4: Antimicrobial resistance rates for other pathogens Haemophilus influenzae (n=11,011 in 2017) Streptococcus pneumoniae (n=3,303 in 2017) Streptococcus pyogenes (n=4,237 in 2017) Campylobacter species (n=3,245 in 2017) Neisseria gonorrhoeae (n=509 in 2017)

3 Section 1: Introduction In 2014, Lord O Neill was commissioned by the UK Prime Minister to review the global impact of antimicrobial resistance. He estimated that by 2050, 10 million lives a year and a cumulative 100 trillion USD of economic output would be at risk due to the rise of drug resistant infections if no proactive solutions were found now to slow down the rise of drug resistance. ( Antimicrobial drugs underpin the practice of modern medicine as we know it: if they lose their effectiveness due to the continuing rise in antimicrobial resistance, key medical procedures (such as gut surgery, caesarean sections, joint replacements, and chemotherapy for cancer treatment) could become compromised and too dangerous to perform. Antimicrobial resistance is an increasing problem in Wales and has already led to a small number of difficult to treat infections, leading to failed therapy and potential complications. Treatment for most infections is started empirically before antimicrobial susceptibilities are known. A particular problem with the spread of antimicrobial resistance is that it becomes more difficult to select empirical therapy that will have reliable activity. The aim of this report from the Antimicrobial Resistance Surveillance Unit of the Healthcare Associated Infections, Antimicrobial Resistance & Prescribing Programme (HARP) team, Public Health Wales is to provide data that can be used to design empirical therapy guidance, and to track antimicrobial resistance trends in Wales from 2008 to The report has had to be selective in what is presented, and concentrates on the major acute hospitals and district general hospitals in Wales, and the local community health boards. Useful links: Review on Antimicrobial Resistance May UK Antimicrobial Resistance Strategy Antimicrobial Resistance Delivery Plan (Wales) Together for Health: Tackling antimicrobial resistance & improving antibiotic prescribing. elivery%20plan.pdf 3

4 Section 2: Key points of interest UK 5 Year Antimicrobial Resistance Strategy The Wales resistance trends for drug-bug combinations reported to the Advisory Committee on Antimicrobial Prescribing, Resistance and Healthcare Associated Infection (APRHAI) as part of the UK 5 year Antimicrobial Resistance Strategy are comparable to the aggregated rates and trends for the UK (page 10). However, in some cases there is considerable variability in resistance rates between different areas and hospitals. E. coli (the commonest cause of blood stream infections in Wales) There was a statistically significant increase in resistance to co-amoxiclav and piperacillin/tazobactam in There was significant variability between hospitals, with concerning levels of resistance emerging to the most commonly used antibacterials in some areas: o o o All-Wales resistance to co-amoxiclav was 37.0%, but locally for Glangwili, Prince Philip, University Hospital Llandough, Wrexham Maelor co-amoxiclav resistance was >40%, and highest in University Hospital of Wales at 44.6%. All-Wales resistance to gentamicin was 11.0%, but locally resistance was 18.6% and 19.1% for University Hospital Llandough and Ysbyty Gwynedd respectively. All-Wales resistance to piperacillin/tazobactam was 12.9%, but locally resistance was 19.0% for Wrexham Maelor. o All-Wales resistance to third generation cephalosporins was 12.9%, but locally resistance was 21.2% and 27.1% for University Hospital Llandough and University Hospital of Wales respectively. o All-Wales resistance to fluoroquinolones was 20.3%, but locally resistance was 29.3%, 31.1% and 37.1% for Wrexham Maelor, University Hospital of Wales, and University Hospital Llandough respectively. Carbapenem resistance remains below 1% in Wales Staphylococcus aureus (the 2 nd commonest cause of bloodstream infections, and the commonest cause of wound infections in Wales) The number of Staphylococcus aureus bloodstream infections have all increased in number from For both meticillin-sensitive (MSSA) and meticillin-resistant (MRSA) S. aureus bloodstream isolates, resistance remains relatively stable, although there is a trend to increasing resistance to fusidic acid, gentamicin and tetracycline in MRSA. For MRSA from wound swabs there is also a trend to increasing resistance to fusidic acid, gentamicin and tetracycline across Wales. However there is particularly high resistance to these agents in the communities served by Wrexham Maelor, University Hospital Llandough, and Ysbyty Glan Clwyd Hospitals: o All-Wales resistance to fusidic acid was 24.4%, but locally resistance was 46.6% and 54.5% for the communities served by Ysbyty Glan Clwyd and University Hospital Llandough respectively. o All-Wales resistance to gentamicin was 10.2%, but locally resistance was 36.1% for the communities served by Wrexham Maelor. 4

5 o All-Wales resistance to tetracycline was 23.7%, but locally resistance was 33.7%, 48.2% and 48.7% for the communities served by Ysbyty Gwynedd, Ysbyty Glan Clwyd, and Wrexham Maelor respectively. Klebsiella spp. (the 3 rd commonest cause of bloodstream infections in Wales) There was a general increase in resistance to most antibacterials compared with There was considerable variability across the country with particularly high levels of resistance to some antibacterials at Withybush Hospital: o Resistance to co-amoxiclav 34.8%, gentamicin 26.1%, and third generation cephalosporins 34.8% at Withybush. Urinary tract infections Coliforms (the commonest cause of urinary tract infections in Wales) Resistance to co-amoxiclav has increased in community, inpatient, and outpatient urinary coliforms (pages 36-42). Resistance to trimethoprim has levelled off in community and inpatient in urinary coliforms at 37.7% and 42.9% respectively; but it continues to rise in outpatient urinary coliforms. Trimethoprim resistance increases with the patient s age, and in the community, resistance is 47.4% in the 80 and over age group. o Trimethoprim is no longer recommended in the Wales guidelines for the treatment of UTI in the 65 and over age group. Nitrofurantoin resistance remains stable at around 11% across Wales. Locally, the communities served by the Cardiff and Swansea laboratories have high rates of resistance to co-amoxiclav, cephalosporins, and fluoroquinolones. antibiotic%20guidance.pdf 5

6 Section 3: Methods Resistance data Data presented Antimicrobial resistance data is presented for the following pathogens: Top ten bacteraemia pathogens (selected) Urinary coliforms (community & hospital) Wound swab isolates (community & hospital) o Staphylococcus aureus including MRSA from wound swabs All specimens (community & hospital) o Streptococcus pneumoniae o Streptococcus pyogenes o Haemophilus influenzae o Campylobacter spp. o Neisseria gonorrhoeae Data sources Antimicrobial susceptibility testing data was extracted from the regional DataStore systems. Community data is presented by DataStore site: For example, data for specimens processed by the laboratories at Prince Philip and Glangwili hospitals are aggregated and reported together as Carmarthen community data (J). The exception to this rule is data for A&E isolates that are reported as community data at individual hospital level. The DataStore sites, and the codes and abbreviations for community and hospital data included in this report are shown in Table 1. Data interpretation As with all surveillance schemes, appropriate interpretation of the data, with an appreciation of the potential biases, is key. The main potential biases, which should be considered in the data presented herein, are: Sampling bias o This occurs if the submission of samples to the laboratory does not represent all patients presenting with that infection, but is selective in some way. If this is the case, the published resistance rate may be skewed, and not representative of the true rate in patients presenting with uncomplicated infection. This effect is likely to be more of an issue with certain sample types. For example, bacteraemia data is felt to be representative, since most patients presenting with sepsis will have a blood culture sent. However if general practitioners only submit urine samples from patients who have failed initial therapy, the published rates of resistance will be falsely high. Selective testing o This occurs if a laboratory only tests susceptibility to a certain agent against selected organisms. For example, a laboratory might only test some agents when an organism is resistant to first-line drugs. This would result in falsely high rates of resistance. In order to reduce the effect of selective testing on the published rates, data is only included if 6

7 >80% of a given isolate from a given specimen is tested against the agent. Methodological variability o In 2012/2013 EUCAST antimicrobial susceptibility testing (AST) methodology was implemented across the laboratories in Wales. Prior to this, a number of different methods for AST had been employed and there were issues with methodological variability within the data. Duplicate testing o This occurs if a patient has multiple specimens tested from a single infection episode. Potentially this can skew the resistance data. In order reduce the effect of this; duplicate isolates are removed from analysis by a sub-routine in DataStore. Isolates are deemed duplicates if the same organism with the same antibiogram is grown from the same sample type within 14 days (for hospital inpatients) or 91 days (for community patients). All Wales data The All-Wales resistance rates for each antimicrobial comprise an aggregate of data from a number of different laboratories. All-Wales resistance rates are only presented for organisms where no testing bias occurred at individual hospital level see below. Community data is from samples referred from a general practice and hospital data is from samples submitted from hospital inpatients or outpatients as described. Individual Hospital/Laboratory data Individual hospital or laboratory resistance rates are only presented for organisms where 80% of such isolates from the given sample type was tested and where the number of isolates tested exceeds 9. Duplicates Data from duplicate isolates was removed prior to analysis. For community data, organisms from the same patient, with the same identification and susceptibility pattern isolated 91 days from the date of the initial isolate were excluded, and for hospital data the cut-off was 14 days. Antimicrobial Groups Prior to the introduction of EUCAST AST methodology in 2012/2013 there was variation AST methodologies between laboratories, and in antibiotic panels tested (e.g. differences in choice and number of third generation cephalosporins tested). In such cases, data is aggregated and resistance rates are expressed at group level. The antimicrobial groups included in this report comprise of the following aggregated susceptibility data: Fluoroquinolones ciprofloxacin &/or levofloxacin, norfloxacin Third generation cephalosporins (3GC) ceftazidime &/or cefotaxime, ceftriaxone, cefpodoxime. Carbapenems imipenem &/or meropenem, ertapenem. 7

8 Susceptibility results Throughout, data is presented in tables and on graphs as resistance rates with 95% confidence intervals (95% CI). 1 For the purpose of this report, susceptibility results recorded as intermediate are included in the category resistant, and in the case of penicillin susceptibility results for S. pneumoniae results recorded as intermediate, lowlevel or high-level resistance are included in the category resistant. 1. Newcombe, Robert G. "Two-Sided Confidence Intervals for the Single Proportion: Comparison of Seven Methods," Statistics in Medicine, 17, (1998). Other surveillance schemes This report focuses on comparisons of data collected for Wales in the calendar years 2008 and To provide some external context to the data presented, it has been also been compared to PHE surveillance data: Public Health England (PHE): Sentinel surveillance schemes are susceptible to potential biases, particularly selective coverage and selective reporting. Thus, any comparisons should be treated with caution. Resistance rates quoted from PHE publications relate to England, Wales and Northern Ireland (unless otherwise stated). Table 1: Codes for hospital and community data Hospital Hospital Code DataStore Site Morriston E Neath Port Talbot Princess of Wales Singleton Nevill Hall Royal Gwent T B S M Swansea Newport Wrexham Maelor H Wrexham Ysbyty Gwynedd K Bangor Ysbyty Glan Clwyd L Rhyl University Hospital of Wales University Hospital Llandough Velindre Prince Charles D F P Q Cardiff Royal Glamorgan C Pontypridd Glangwili Prince Philip N J Carmarthen Bronglais A Aberystwyth Withybush G Haverfordwest All-Wales R Z 8

9 Section 4: Monitoring Trends in Resistance UK 5 Year Antimicrobial Resistance Strategy Background In 2014, a sub-group of the Advisory Committee on Antimicrobial Prescribing, Resistance and Healthcare Associated Infection (APRHAI) was established to recommend surveillance outputs to support the UK Five Year Antimicrobial Resistance Strategy. Appendix C Monitoring Trends in Resistance of the Strategy document states: Changes in the level of resistance to antibiotics like the carbapenems, which are often the last option for hard to treat infections, will be monitored. The agreed drug-bug combinations for monitoring resistance are listed in Table 2; the combinations were ratified by the Department of Health High-Level Steering Group. Public Health Wales provide the Wales data to APRHAI for this surveillance project. Table 2: APRHAI Drug-Bug Combinations Specimen Organism Data Set Antimicrobial cefotaxime or ceftazidime Escherichia coli Primary imipenem or meropenem ciprofloxacin gentamicin Secondary piperacillin/tazobactam cefotaxime or ceftazidime Klebsiella pneumoniae Primary imipenem or meropenem ciprofloxacin gentamicin Secondary piperacillin/tazobactam Blood Culture cefotaxime or Klebsiella oxytoca Primary ceftazidime imipenem or meropenem ciprofloxacin gentamicin piperacillin/tazobactam Pseudomonas spp. Primary ceftazidime imipenem or meropenem Acinetobacter spp. Secondary colistin Enterococcus spp. Secondary vancomycin Staphylococcus aureus Secondary meticillin Streptococcus pneumoniae Primary penicillin All ceftriaxone Neisseria gonorrhoeae Primary specimens azithromycin 9

10 APRHAI Primary Data Set Table 3 shows the resistance rates for Wales compared with the UK aggregate rates for some primary data set drug-bug combinations. There are small differences in some of the resistance rates, but generally, the trends in resistance are comparable. Table 3: APRHAI Primary Data Set APRHAI Primary data set Escherichia coli Co-amoxiclav Wales *UK *41 - Cefotaxime Wales &/or *UK *11 - ceftazidime Wales Ciprofloxacin *UK *18 - Gentamicin Wales *UK *10 - Pip/tazobactam Wales *UK *12 - Imipenem &/or Wales < meropenem *UK *<1 - Klebsiella pneumoniae Cefotaxime Wales &/or *UK *12 - ceftazidime Wales Ciprofloxacin *UK *11 - Gentamicin Wales *UK *9 - Pip/tazobactam Wales *UK *18 - Imipenem &/or Wales meropenem *UK Pseudomonas spp Ceftazidime Wales *UK Imipenem &/or Wales meropenem *UK Streptococcus pneumoniae Penicillin Wales *UK *The 2016 UK rates only includes data for England and Northern Ireland. The 2017 Wales data relates to Klebsiella spp. and not only Klebsiella pneumoniae as previously reported. No UK data has been published for

11 Section 5.1: Antimicrobial resistance rates for the most common organisms causing bacteraemia Background The 2016 top ten bacteraemia report for Wales comprises the commonest organisms isolated from blood cultures in Wales, see Table 4 below. The data for 2017 is not yet published but will be available on the Public Health website later in the year. Internet: 003b246b/dc07cde91e431bb af449/$FILE/Wales%20Top%2010% pdf Intranet: 003b246b/dc07cde91e431bb af449/$FILE/Wales%20Top%2010% pdf Table 4: Top Ten Bacteraemias 2016 Rank Organism Rate per 100,000 bed days 1 Escherichia coli (E. coli) 81 2 Staphylococcus aureus (MSSA) 23 3 Klebsiella species 17 4 Enterococcus species 15 5 Streptococcus pneumoniae 14 6 Coagulase-negative Staphylococcus 13 7 Proteus species 7 8 Streptococcus viridians group 6 8 Pseudomonas aeruginosa 6 8 Streptococcus group B 6 The datasets include infections originating from community and hospital sources (inpatient and outpatient), and so may be affected by local clonal strains which can result in marked variability in resistance rates between hospitals/regions; results should be interpreted with caution. Since coagulase negative staphylococci are frequently contaminants when isolated from blood cultures, data on susceptibility are not presented here. Streptococcus viridians group has appeared in the top 10 in 2016 & 2017 but susceptibility data are not presented here. Enterobacter species, Serratia species and MRSA are no longer in the top 10 however; data is presented since they appear in previous reports. The data in this report is not presented in rank order, but rather an order to allow easy comparison of resistances for related bacteria. Figure 1 (over page) shows the number of isolates included in this data set per year ( ) for E. coli, Enterobacter spp., Klebsiella spp., Proteus spp., Pseudomonas aeruginosa, Serratia spp. individually, and collectively as the group gram-negative bacilli (GNB). There has been an upward trend in the numbers of GNB, driven by the increase in E. coli, Klebsiella spp., and Proteus spp. Figure 2 shows the number of isolates per year ( ) for Enterococcus spp., MSSA, MRSA and Streptococcus pneumoniae individually and collectively as the group gram-positive cocci (GPC). The downward trend in GPC reversed in 2015, largely due to the increase in MSSA and S. pneumoniae. 11

12 Figure 1: Gram-negative bacteraemia numbers (2008 to 2017). Figure 2: Gram-positive bacteraemia numbers (2008 to 2017). 12

13 Escherichia coli (n=2,663 in 2017) E. coli is the commonest organism grown from blood cultures in Wales and the UK. The All-Wales patterns of resistance for 2008 to 2017 are shown in Figure 3, and the individual hospital resistance rates for APRHAI primary drug set are shown in Table 5. There has been a statistically significant increase in resistance rates for co-amoxiclav (COA) and piperacillin/tazobactam (PTZ) between 2016 and Fluoroquinolone (FQ), gentamicin (GEN), and third generation cephalosporin (3GC) resistance shows no change. Imipenem and meropenem (CARB) resistance rates remain below 1% in the Wales. Figure 3: All-Wales resistance rates for E. coli bacteraemia (2008 to 2017). 13

14 Table 5: Escherichia coli Note: The range of resistance is outlined with boxes e.g. the range for co-amoxiclav (COA) was 15.4% to 44.6%; individual hospital resistance rates statistically higher than the All-Wales rate are highlighted in pink. Resistance rates are not recorded when <80% of the isolates were tested. The resistance rates for E. coli bacteraemia in Neath Port Talbot (T) were notably high to a number of agents, but the number of isolates was small (n=14). 14

15 APRHAI Primary data Set Interpretation Tables 6-10: The tables show trends in resistance to drug/bug combinations in the APRHAI primary data set at hospital level, across time. The tables use a colour gradation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. The first column in the tables show the hospital code and the median number of isolates tested across the time period e.g. in Table 6, hospital code A (64) denotes Bronglais hospital with a median number of 64 isolates tested per year across the seven-year period. Note: Individual hospital or laboratory resistance rates are only presented for organisms where 80% of such isolates from the given sample type was tested and where the number of isolates tested exceeds nine. It is important to remember when interpreting this data set that hospital level data often represents small numbers of organisms, and single isolate resistance within these numbers can produce misleadingly large changes in resistance. Table 6: Trends in co-amoxiclav for E. coli by hospital ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales The data in Tables 6-10 show regionally high resistance rates to co-amoxiclav in UHL (P) and UHW (F), Wrexham Maelor (H) and Ysbyty Gwynedd (K), and Glangwili (J) and Prince Philip (R) hospitals. For Royal Glamorgan (C), the notable downward trend in co-amoxiclav resistance continues. An increase in third generation cephalosporin resistance in E. coli isolated from blood cultures in UHL (P) between 2016 and An increase in fluoroquinoles resistance in Wrexham Maelor hospital (H), with generally higher resistance to third generation cephalosporins and fluoroquinoles in UHL (P) and UHW (F). An increase in piperacillin/tazobactam resistance at an All-Wales level, driven by a general increase in resistance across most hospitals. 15

16 Table 7: Trends in third generation cephalosporin resistance for E. coli by hospital ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales Table 8: Trends in fluoroquinolone resistance for E. coli by hospital ( ) 16

17 Table 9: Trends in gentamicin resistance for E. coli by hospital ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales Table 10: Trends in piperacillin/tazobactam resistance for E. coli ( ) 17

18 Klebsiella spp. (n=549 in 2017) The All-Wales patterns of antimicrobial resistance in Klebsiella spp. are shown in Figure 4 and Table 12. There has been a statistically significant increase in resistance rates for co-amoxiclav (COA) across time ( ). Imipenem and meropenem resistance rates remain below 1% in the UK. Figure 4: All-Wales antimicrobial resistance rates for Klebsiella species; isolated from blood culture (2007 to 2017) The Welsh resistance rates are comparable with the Klebsiella spp. bacteraemia data published by the PHE (see Table 11 below). Table 11: Antibiotic susceptibility of Klebsiella species bacteraemia (England and Northern Ireland): (PHE) Klebsiella spp Total reports: 6,224 6,153 6,618 7,279 8,413 Piperacillin/ Tazobactam Meropenem Cefotaxime Ceftazidime Ciprofloxacin Gentamicin % Non-susceptibility 13% 15% 16% 17% 17% Reports with susceptibility data 5,291 5,254 5,359 6,250 7,096 % Non-susceptibility 1% 1% 1% 1% 1.0% Reports with susceptibility data 4,357 4,478 4,918 5,993 7,099 % Non-susceptibility 10% 10% 10% 10% 10% Reports with susceptibility data 3,217 3,136 3,257 3,618 4,159 % Non-susceptibility 9% 10% 10% 10% 10% Reports with susceptibility data 4,374 4,126 4,276 5,226 6,087 % Non-susceptibility 8% 9% 9% 9% 10% Reports with susceptibility data 5,180 5,074 5,251 6,178 7,264 % Non-susceptibility 6% 7% 6% 7% 8% Reports with susceptibility data 5,590 5,488 5,720 6,606 7,628 18

19 Table 12: Klebsiella spp. Note: Resistance rates are not recorded if the organisms are intrinsically resistant to an antibacterial agent e.g. for amoxicillin. The range of resistance is outlined with boxes e.g. the range for co-amoxiclav (COA) was 0.0% to 34.8%; individual hospital resistance rates statistically higher than the All-Wales rate are highlighted in pink. Locally, gentamicin and third generation cephalosporin resistance rates for Klebsiella spp. bacteraemia from Withybush hospital (G) were notably higher than the rest of Wales. 19

20 APRHAI Primary data Set Interpretation Tables 13-17: The tables show trends in resistance to drug/bug combinations in the APRHAI primary data set at hospital level, across time. The tables use a colour gradation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. Note 1: The data is different to that shown in Table 3 (page 11); Table 3 shows the rates for Klebsiella pneumoniae as per APRHAI instructions, but the following tables show the data for all Klebsiella species to allow comparisons with previous reports. Note 2: Individual hospital or laboratory resistance rates are only presented for organisms where 80% of such isolates from the given sample type was tested and where the number of isolates tested exceeds nine. It is important to remember when interpreting this data set that hospital level data often represents small numbers of organisms, and single isolate resistance within these numbers can produce misleadingly large changes in resistance. Table 13: Trends in third generation cephalosporin resistance for Klebsiella spp. ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales The data in Tables show a small but continuing increasing trend in resistance to third generation cephalosporins, fluoroquinolones, and piperacillin/tazobactam at an All-Wales level, driven by small increases in resistance to these agents across many hospitals. Imipenem and/or meropenem resistance was not reported in Klebsiella spp. in

21 Table 14: Trends in fluoroquinolone resistance for Klebsiella spp. ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales Table 15: Trends in gentamicin resistance for Klebsiella spp. ( ) 21

22 Table 16: Trends in piperacillin/tazobactam resistance for Klebsiella spp. ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales Table 17: Trends in imipenem/meropenem resistance for Klebsiella spp. ( ) 22

23 Enterobacter spp., Serratia spp., Proteus spp., and Ps. aeruginosa Table 18: Enterobacter spp., Serratia spp., Proteus spp., and Ps. aeruginosa Note: Resistance rates are not recorded if the organisms are intrinsically resistant to an antibacterial agent e.g. for amoxicillin. 23

24 Enterobacter spp. (n=115 in 2017) The All-Wales patterns of antimicrobial resistance for Enterobacter spp. are shown in Figure 5 & Table 18 with no statistically significant changes across time. Figure 5: All-Wales antimicrobial resistance rates for Enterobacter species; isolated from blood culture (2008 to 2017) Serratia spp. (n=91 in 2017) The All-Wales patterns of antimicrobial resistance for Serratia spp. are shown in Figure 6 and Table 18. There has been no significant change in resistance rates between 2014 and Figure 6: All-Wales antimicrobial resistance rates for Serratia species; isolated from blood culture (2008 to 2017) 24

25 Proteus spp. (n=223 in 2017) The All-Wales patterns of antimicrobial resistance in Proteus spp. are shown in Figure 7 & Table 18. There has been no significant change in resistance rates between 2014 and Note: The reliability of susceptibility testing of carbapenems with automated systems is uncertain and resistance rates may be unreliable. Figure 7: All-Wales antimicrobial resistance rates for Proteus species; isolated from blood culture (2008 to 2017) Pseudomonas aeruginosa (n=190 in 2017) The All-Wales patterns of antimicrobial resistance in Pseudomonas aeruginosa are shown in Figure 8 & Table 18. There has been a statistically significant decrease in ciprofloxacin resistance between 2011 and Figure 8: All-Wales antimicrobial resistance rates for Pseudomonas aeruginosa; isolated from blood culture (2008 to 2017) 25

26 Staphylococcus aureus The All-Wales resistance rates for Staphylococcus aureus at hospital level are shown in Table 19, the data includes all Staphylococcus aureus both MSSA and MRSA. In 2017, gentamicin and tetracycline resistance rates for S. aureus bacteraemias in Wrexham Maelor (H) were notably higher than the All-Wales rate. Flucloxacillin resistance reflects the proportion of S. aureus bacteraemias that were MRSA; the proportions of MRSA bacteraemias were notably higher in this locality than other acute hospitals in Wales. Figure 9 shows the numbers of Staphylococcus aureus bacteraemias from 2008 to In 2017, the rates for Staphylococcus aureus bacteraemias both MRSA and MSSA have increased in number. Figure 9: All-Wales Staphylococcus aureus bacteraemia numbers (2008 to 2017) 26

27 Table 19: Staphylococcus aureus (MSSA & MRSA) Note: The range of resistance is outlined with boxes e.g. the range for clindamycin (CLI) was 5.3% to 29.1%; individual hospital resistance rates statistically higher than the All-Wales rate are highlighted in pink. Locally, gentamicin and tetracycline resistance rates for S. aureus bacteraemias in Wrexham Maelor (H) were notably higher than the All-Wales rate. 27

28 APRHAI Primary data Set Interpretation Table 20: The table show trends in resistance to a drug/bug combination in the APRHAI primary data set at hospital level, across time. The tables use a colour gradation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. Table 20: Trends in meticillin resistance for Staphylococcus aureus ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL Q = Velindre R = Prince Philip S = Singleton T = Neath Port Talbot Z = All-Wales The data in Table 20 shows a notable reduction is meticillin resistance across time in Staphylococcus aureus isolated from blood cultures taken in Ysbyty Gwynedd (K) i.e. a reduction in the proportion of MRSA from 42.6% in 2011 to 18.2% in The proportion of MRSA remains high in Wrexham Maelor (H) at 28.3%. 28

29 Meticillin Sensitive Staphylococcus aureus (n=834 in 2017) The All-Wales pattern of antimicrobial resistance in MSSA is shown in Figure 10 and Table 21; with no statistically significant changes between 2008 and Figure 10: All-Wales antimicrobial resistance rates for Meticillin Sensitive Staphylococcus aureus (MSSA) isolated from blood culture (2008 to 2017) Meticillin Resistant Staphylococcus aureus (n=123 in 2017) The All-Wales pattern of antimicrobial resistance in MRSA is shown in Figure 11. There has been a statistically significant increase in tetracycline resistance across time ( ), and gentamicin resistance continues on an upward trend. Figure 11: All-Wales antimicrobial resistance rates for Meticillin Resistant Staphylococcus aureus (MRSA) isolated from blood culture (2008 to 2017) 29

30 Table 21: Meticillin Sensitive Staphylococcus aureus (MSSA) Note: The numbers of MRSA are too small at hospital level to present in a table. 30

31 Enterococcus spp. (n=449 in 2017) The All-Wales pattern of antimicrobial resistance in Enterococcus spp. is shown in Figure 12 and Table 22; with no statistically significant changes in resistance between 2008 and Figure 12: All-Wales antimicrobial resistance rates for Enterococcus spp. isolated from blood culture (2007 to 2016) In 2017, the All-Wales resistance rate for amoxicillin was 43% Susceptibility to amoxicillin is a guide to speciation of the organism, E. faecalis being normally susceptible and E. faecium being normally resistant, and suggests that in 2017, 57% of entercoccal bacteraemias were due to E. faecalis. Table 22: Enterococcus spp. 31

32 Note: Locally resistance rates for vancomycin ranged from 31.8% in Prince Charles hospital (N) to 0% in Withybush, Wrexham Maelor (H) and Ysbyty Gwynedd (K). Amoxicillin resistance ranged from 25% in Withybush hospital (G) to 55% in Morriston (E). Amoxicillin resistance rates may simply reflect variation in the proportion of E. faecalis to E. faecium. APRHAI Primary data Set Interpretation Table 23: The table show trends in resistance to a drug/bug combination in the ARHAI primary data set at hospital level, across time. The tables use a colour gradation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. The number following the hospital code e.g. (9) represents the median number of isolates per year over the data set. Resistance rates are only shown when the number of isolates were 10 or more for any one year e.g. for hospital A data is only shown for 2016 and Table 23: Trends in vancomycin resistance for Enterococcus spp. ( ) Key: A= Bronglais B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles P = UHL S = Singleton Z = All-Wales The data in Table 23 shows an increase in vancomycin resistance in Enterococcus spp. isolated from blood cultures taken in Prince Charles hospital (N) indicating an increase in the proportion VRE, and a notable reduction in the proportion of VRE in Ysbyty Gwynedd (K). 32

33 Streptococcus pneumoniae (n=406 in 2017) The All-Wales pattern of antimicrobial resistance is shown in Figure 13 & Table 24; with no statistically significant change in resistance across time. Figure 13: All-Wales antimicrobial resistance rates for Streptococcus pneumoniae isolated from blood culture (2008 to 2017) Table 24: Streptococcus pneumoniae 33

34 APRHAI Primary data Set Interpretation Table 25: The table show trends in resistance to a drug/bug combination in the APRHAI primary data set at hospital level, across time. The tables use a colour gradation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. Table 25: Trends in penicillin resistance for S. pneumoniae ( ) Key: B = Princess of Wales C = Royal Glamorgan D = Royal Gwent E = Morriston F = UHW G = Withybush H = Wrexham Maelor J = Glangwili K = Ysbyty Gwynedd L = Ysbyty Glan Clwyd M = Nevill Hall N = Prince Charles R = Prince Philip S = Singleton Z = All-Wales The data in Table 25 shows relatively low penicillin resistance in Streptococcus pneumoniae isolated from blood cultures in 2017 across Wales. 34

35 Section 5.2: Antimicrobial resistance rates for urinary coliforms For the purposes of this report, the term coliform refers to organisms that were reported as a coliform by the laboratory, or when identified further, were reported as one of the genera belonging to the family Enterobacteriaceae. The genera included in this section of the report comprise: Citrobacter Edwardsiella Enterobacter Escherichia Hafnia Klebsiella Kluyvera Morganella Pantoea Proteus Providencia Rahnella Salmonella Serratia Yersinia It should be noted that data from routinely submitted urine specimens is more prone to bias than data from blood culture isolates due to variable sampling by clinicians. Thus, resistance rates quoted here are likely to be higher due to increased sampling from patients who are more likely to have resistant organisms (e.g. patients with recurrent infections or infections that have failed to respond to initial therapy). This should be factored into any use of the data presented for the design of empiric treatment guidance. The generation of more specific data reports (e.g. different patient age groups) can be discussed with the Antimicrobial Resistance Surveillance Unit contact HARP@wales.nhs.uk. 35

36 Community Urinary Coliforms (n=100,578 in 2017) The All-Wales pattern of antimicrobial resistance for community urinary coliforms is shown in Figure 14 & Table 26. Figure 14: All-Wales antimicrobial resistance rates for coliforms from community urine samples (2008 to 2017) There has been a statistically significant increase in co-amoxiclav resistance in Fluoroquinolone resistance continues on an upward trend, and trimethoprim resistance has levelled off. Amoxicillin resistance, which is not represented in this graph because of its continuing high levels of resistance, remains at around 55%. Figures 15 and 16 over page show antimicrobial resistance by age group and gender, and age group and time. Resistance to trimethoprim has levelled off largely influenced by the rates for the older age groups. High resistance has been noted in the 80+ age group (47% in males and females). Whilst resistance to trimethoprim continues to rise in the lower age groups, the high rates of resistance may reflect an element of selective testing within the community. The true rate of resistance to trimethoprim in patients presenting with uncomplicated UTI in the community is likely to be considerably lower, and trimethoprim remains the suggested first-line empirical therapy for most of these patients. However, in the elderly, or patients who have received antibiotics within the last 3 months, the likelihood of infection with a resistant organism is higher, and an alternative antibiotic should be considered. Alternatives include: o o o Nitrofurantoin 100mg m/r BD for 3 days in women and 7 days in men (if GFR over 45mL/min). Pivmecillinam 400mg for 3 days in women and a minimum of 7 days in men. Fosfomycin 3g stat dose in women, and a 3g stat plus 3g dose 3 days later in men. 36

37 Figure 15: All-Wales trimethoprim resistance rates for coliforms from community urine samples by age group and gender (2017) Figure 16: All-Wales trimethoprim resistance rates for coliforms from community urine samples by age group and time ( ) 37

38 Table 26: Community Urinary Coliforms Note: The range of resistance is outlined with boxes e.g. the range of resistance to amoxicillin was 45.8% %; individual hospital rates statistically higher than the All- Wales rate are highlighted in pink. Note: Urinary coliforms from patients attending A&E are included in the community data set. The resistance rates for community urinary coliforms vary considerably across Wales with many hospitals reporting rates statistically higher than the all Wales data. Locally, the communities served by the Cardiff and Swansea laboratories have high rates of resistance to co-amoxiclav, cephalosporins, and fluoroquinolones. The Cardiff community also has higher rates of trimethoprim resistance. 38

39 Outpatient Urinary Coliforms (n=8,512 in 2017) The All-Wales pattern of antimicrobial resistance for outpatient urinary coliforms is shown in Figure 17 & Table 27. Figure 17: All-Wales antimicrobial resistance rates for coliforms from outpatient urine samples (2008 to 2017) There has been a statistically significant increase in trimethoprim resistance across time, and a statistically significant increase co-amoxiclav resistance from 2013 to Trimethoprim and fluoroquinolone resistance continue on an upward trend. Amoxicillin resistance, which is not represented in this graph because of its continuing high levels of resistance, remains at around 55%. 39

40 Table 27: Hospital Outpatient Urinary Coliforms Note 1: The range of resistance is outlined with boxes e.g. the range of resistance to amoxicillin was 43.8% %; individual hospital rates statistically higher than the All- Wales rate are highlighted in pink. Note 2: Individual hospital or laboratory resistance rates are only presented for organisms where 80%. Locally, the UHW (F) outpatients have high rates of resistance to co-amoxiclav, cephalosporins, fluoroquinolones and trimethoprim. 40

41 Inpatient Urinary Coliforms (n=17,986 in 2017) The All-Wales pattern of antimicrobial resistance for inpatient urinary coliforms is shown in Figure 18 & Table 28. Figure 18: All-Wales antimicrobial resistance rates for coliforms from inpatient urine samples (2007 to 2016) Trimethoprim resistance levelled off in 2017, and remains the same as the 2016 rate of 42.9%. Co-amoxiclav resistance shows a statistically significant increase between 2013 and Amoxicillin resistance, which is not represented in this graph because of its continuing high levels of resistance, remains > 60%. 41

42 Table 28: Hospital Inpatient Urinary Coliforms In 2016, the All-Wales resistance rates for inpatients urinary coliforms were statistically higher for all the agents listed than those for community or outpatients (Tables 26, 27 & 28). In 2017, a number of hospitals had resistance rates for urinary coliforms that were statistically higher than the All-Wales rate; the most notable was University Hospital Llandough (P) with statistically high rates to co-amoxiclav, third-generation cephalosporins, fluoroquinolones, and trimethoprim. 42

43 Section 5.3: Antimicrobial resistance rates for Staphylococcus aureus The data in this section is presented to reflect the antimicrobial susceptibility of organisms causing skin and soft tissue infections occurring in the community, and is based on the specimen description wound swab 43

44 MSSA ( all locations n=23,519 in 2017) Community MSSA (n=15,873 in 2017) The All-Wales pattern of antimicrobial resistance for MSSA from community wound swabs are shown in Figure 19 and Table 29 with a general upward trend in resistance to erythromycin (ERY), fusidic acid (FUS) and tetracycline (TET). Figure 19: All-Wales antimicrobial resistance rates for MSSA from community Wound swabs (2008 to 2017) In 2017, the All-Wales resistance rates for community, outpatients and inpatients MSSA were comparable for most of the antimicrobials see Tables 29, 30 and 31. At different times in the ten-year period 2008 to 2017, there were increases in resistance to different agents in different geographical areas, but there was no set pattern of increasing or high resistance in any particular community or hospital, and this probably reflects the varying presence of epidemic strains. Vancomycin resistance remained undetected in MSSA between 2008 &

45 Tables 29: Meticillin Sensitive Staphylococcus aureus (MSSA) from community wound swabs Table 30: Meticillin Sensitive Staphylococcus aureus (MSSA) from outpatient wound swabs Tables 31: Meticillin Sensitive Staphylococcus aureus (MSSA) from inpatient wound swabs 45

46 MRSA (all locations n=2,439 in 2017) Community MRSA (n=1,501 in 2017) The All-Wales pattern of antimicrobial resistance for MRSA from community wound swabs is shown in Figure 20 and Table 32. Figure 20: All-Wales antimicrobial resistance rates for MRSA from community Wound swabs (2008 to 2017) There has been a statistically significant increase in resistance rates for fusidic acid (FUS), gentamicin (GEN) and tetracycline (TET) over time. Locally, there was wide variability in resistance rates within Wales; with notably high rates of gentamicin and tetracycline resistance in communities served by the North Wales Laboratories in Wrexham Maelor (H), Ysbyty Gwynedd (K) and Ysbyty Glan Clwyd (L). High rates of fusidic acid resistance in University Hospital Llandough (P) and Ysbyty Glan Clwyd communities. A similar picture is seen in hospital patients - see Table 33 and 34. Tables 32: Meticillin Resistant Staphylococcus aureus (MRSA) from community wound swabs 46

47 Hospital Inpatient and Outpatient MRSA (n=938 in 2017) The trends in antimicrobial resistance for both hospital inpatient and outpatient MRSA are similar to those seen in the community. Table 33: Meticillin Resistant Staphylococcus aureus (MRSA) from outpatient wound swabs Some of the same local patterns of resistance seen in the community were seen in hospital patients from the same geographical area, with notably high gentamicin and tetracycline rates in both inpatients and outpatients from Wrexham Maelor (H). High fusidic acid and tetracycline resistance rates in patients from Ysbyty Glan Clwyd (L) and University Hospital Llandough (P). High tetracycline resistance rates in outpatients from Ysbyty Gwynedd (K): See Tables 33 and 34. Tables 34: Meticillin Resistant Staphylococcus aureus (MRSA) from inpatient wound swabs There were no confirmed cases of vancomycin intermediate/resistant MRSA (VISA) between 2007 and

48 Table 35: Trends in fusidic acid resistance for Meticillin Resistant Staphylococcus aureus (MRSA) from community wound swabs ( ) Key: A= Bronglais C = Royal Glamorgan D = Royal Gwent F = UHW G = Withybush H = Wrexham Maelor J = Glangwili L = Ysbyty Glan Clwyd N = Prince Charles P = UHL S = Singleton Z = All-Wales Interpretation Tables 35-36: The tables use a graduation based on the lowest resistance to the highest resistance figures, to highlight local patterns of resistance across time. The data in table 35 shows resistance to fusidic acid has remained high in the communities served by Royal Glamorgan (C), Wrexham Maelor (H), Glangwili (J) and Ysbyty Glan Clwyd (L); and an increase in resistance in the University Hospital Llandough (P) community. The data in table 36 shows resistance to tetracycline has remained high in the communities served by Wrexham Maelor (H), Ysbyty Glan Clwyd (L), and University Hospital Llandough (P). Table 36: Trends in tetracycline resistance for Meticillin Resistant Staphylococcus aureus (MRSA) from community wound swabs ( ) 48

49 Section 5.4: Antimicrobial resistance rates for other pathogens. The data in this section of the report comprises other pathogens which may commonly cause important infections other than bacteraemia. The data is for all specimens from all locations (community, inpatient and outpatient). Haemophilus influenzae Streptococcus pneumoniae Streptococcus pyogenes Campylobacter species Neisseria gonorrhoeae 49

50 Haemophilus influenzae (n=11,011 in 2017) The All-Wales pattern of antimicrobial resistance for Haemophilus influenzae from all specimens/locations is shown in Figure 21 & Table 37; the increase in resistance to co-amoxiclav and amoxicillin seen between levelled off in Figure 21: All-Wales antimicrobial resistance rates for H. influenzae; all specimens and all locations (2007 to 2016) Locally, there was variability in co-amoxiclav and amoxicillin resistance rates within Wales with higher rates of resistance being seen in the hospital and the community served by the laboratory at Withybush (G), and Glangwili (J). Table 37: Haemophilus influenzae - all specimens and all locations 50

51 Streptococcus pneumoniae (n=3,303 in 2017) The All-Wales pattern of antimicrobial resistance for Streptococcus pneumoniae from all specimens and all locations is shown in Figure 22 & Table 38; with no statistically significant differences in resistance rates between 2016 and 2017, but with a notable trend in increasing resistance to penicillin from 2010 onwards. Figure 22: All-Wales antimicrobial resistance rates for S. pneumoniae; All specimens and all locations (2007 to 2017) The rates for all three agents are higher than the rates for S. pneumoniae isolates from blood culture: See Figure 11. Locally there was some variation in the resistance across Wales with higher rates for both erythromycin and penicillin seen in University Hospital Llandough (P). Table 38: Streptococcus pneumoniae - all specimens and all locations 51

52 Streptococcus pyogenes (n=4,237 in 2017) The All-Wales pattern of antimicrobial resistance for Streptococcus pyogenes from all specimens/locations is shown in Figure 23; with no statistically significant changes in resistance across time. Figure 23: All-Wales antimicrobial resistance rates for S. pyogenes; all specimens and all locations (2008 to 2017) There were no validated cases of penicillin resistant S. pyogenes in Wales from 2008 to

53 Campylobacter species (n=3,245 in 2017) The All-Wales pattern of antimicrobial resistance for Campylobacter spp. from all locations is shown in Figure 24 & Table 39; with a statistically significant increase in ciprofloxacin resistance across time ( ). Figure 24: All-Wales antimicrobial resistance rates for Campylobacter spp.; all specimens and all locations (2008 to 2017) Table 39: Campylobacter spp. - all specimens and all locations Locally there was some variation in the resistance across Wales with higher rates for ciprofloxacin in the University Hospital Wales and its community (F). 53

54 Neisseria gonorrhoeae (n=509 in 2017) The All-Wales pattern of antimicrobial resistance for Neisseria gonorrhoeae from all specimens and all locations is shown in Figure 25; with a statistically significant decrease in ciprofloxacin resistance between 2016 and Figure 25: All-Wales antimicrobial resistance rates for N. gonorrhoeae; all specimens and all locations (2013 to 2017) High-level resistance to azithromycin cannot be determined by routine susceptibility testing methods, and most isolates from Wales are referred to PHE for testing. PHE publish data on trends in antimicrobial resistance and decreased susceptibility in gonococcal infection in England and Wales in the GRASP report (see link below): The points of interest for Wales are that in 2017, no ceftriaxone resistance was identified, and no high-level azithromycin resistance was identified in isolates from Wales. 54