FIS Resistance Surveillance: The UK Landscape. Alasdair MacGowan Chair BSAC Working Party on Antimicrobial Resistance Surveillance

Transcription

1 FIS 2013 Resistance Surveillance: The UK Landscape Alasdair MacGowan Chair BSAC Working Party on Antimicrobial Resistance Surveillance

2 A statement of the obvious Good quality surveillance data on resistant micro-organisms are essential to underpin effective interventions to counter the problems of resistance and developing guidelines on prescribing antimicrobial drugs. Such data should be epidemiologically and clinically relevant. Copenhagen Recommendations European Union, 1998

3 Problems with antimicrobial surveillance in the last 20 years routine passive surveillance was biased by disease state, age, uneven methodologies in and outside labs. However, large data sets are captured. sentinel focused surveillance using central laboratory overcomes some problems by use of standard within lab methods but not requesting bias and is costly poor dissemination of data public/private and funders providers Based on Livermore, MacGowan, Wale, 1998; BMJ 317, 614-5

4 Epidemiologically relevant data? Passive Surveillance Collection of antimicrobial resistance data in the community i.e. UK/Ireland depends on healthcare system Patient presents with infection Primary Care Doctor Primary Care Nurse Specimen sent to laboratory Other NHS Direct Accident Departments (OTC Pharmacy) Bacteria isolated, identified and susceptibility tests performed using routine methods Confirmed and repeated by central laboratory using gold standard methods Results entered into Laboratory Information Management System (LIMS) Resistance Data Base

5 Pitfalls with passive surveillance Data skewed because patients who fail initial empirial therapy more likely to be tested Age bias in LIMS towards the elderly and young Location of patient (Admission Unit vs Primary Care) Isolates from different disease states included in database disproportionately:- i.e. Acute Sinusitis vs UTI, exacerbation COPD vs CAP Specimens may be sent for non-infective reasons

6 Specimen referral by location and age for community LRTI: Sputum Proportion of patients Age Band In community In hospital <48h Lovering, unpublished

7 Assessment of Biases: Sputum Referral Single Practice 1:35 episodes of GP diagnoses Chest Infection results in sputum referral Ampicillin resistance in H.influenzae rate Laboratory database (passive surveillance) 0.20 Unselected patients with LRTI who received an antibiotic (n=20) 0.11 MacGowan et al, 1998

8 Assessment of Bias: Laboratory Database vs UTI Symptoms in UTI Group, i.e. culture >10 5 pathogen Pain on urination 0.85 Increased frequency 0.87 Blood in urine 0.25 Increased urgency 0.65 Urination at night 0.56 Abdominal pain 0.56 Back pain 0.38 Smelly urine 0.44 Patient Group TMP n %S P CI Acute UTI Laboratory data Lovering & MacGowan, 2003

9 But significantly underpowered:- If true rate of resistance in laboratory database is 24% Number of isolates required 90% power 80% power +10% % % % Patients recruited over 2-3 years : 55

10 Problem with sentile surveillance in Primary Care Acute Infection Group UTI (n=165) Cough (n=197) Throat (n=249) Urine sample (n=150) Sputum sample (n=99) Throat swab (n=243) Potential pathogen (n=67) E.Coli (n=59) Proteus (n=3) Klebsiella (n=4) Other (n=2) Potential pathogen (n=23) H.influenzae (n=13) S.pneumoniae (n=8) M.catarrhalis (n=1) Potential pathogen (n=38) Group A n=79 Group C n=11 Group G n=6

11 BSAC antimicrobial resistance surveillance contribution: Overcomes previous perceived problems: continuous since 1999 based on uniform within laboratory methodology target at relevant resistances by use of a core testing set allow cross validation with other resistance data bases data is widely available Does not overcome: bias related to passive surveillance in the community partly answers concerns about how representative sampling is (increase from 25 to 40 collecting labs)

12 BSAC Surveillance Programme : Outputs Trends in CAP: S.pneumoniae, H.influenzae Trends in HAP: S.aureus, ESBL producers Trends in Bacteraemia: MRSA, ESBL, carbapenemase producers

13 CAP Programme: S. pneumoniae (420) % non-susceptible

14 Both Programmes: Penicillin non-susceptible S.pneumoniae - Great Britain data Year Respiratory programme Bacteraemia programme % % 6.8% % 7.1% % 8.0% % 3.0% % 2.0% % 4.2% % 4.3% % 2.2% % 3.7% % 4.3% % 3.6% % 8.3%

15 % Penicillin non-susceptibility in pneumococci: BSAC bacteraemia surveillance

16 CAP Programme: H. influenzae (516) % non-susceptible

17 HAP Programme: E. coli (274)

18 HAP Programme: ESBL prevalence non- non- non- ICU ICU ICU ICU ICU ICU Klebsiella spp. Escherichia coli Enterobacter spp. %ESBL 18% 20% 7% 9% 13% 8% % ctx-m 30% 35% 17% 33% 67% 50% %ESBL 16% 13% 17% 16% 13% 9% % ctx-m 81% 71% 85% 53% 85% 73% %ESBL 9% 6% 8% 7% 5% 6% % ctx-m 57% 60% 60% 50% 0% 0%

19 HAP Programme: S. aureus (226) % non-susceptible P <0.0005

20 Bacteramia Programme: % MRSA over time (BSAC surveillance) England Wales, Scotland, N. Ireland, Ireland

21 N of MRSA bacteraemias 60 MRSA bacteraemia in England % MRSA in S. aureus bacteraemia (BSAC surveillance) Number of MRSA bacteraemias (mandatory reports) 0

22 % Vancomycin non-susceptibility: enterococci, BSAC bacateraemia surveillance E. faecalis E. faecium

23 BSAC E. coli from bacteraemia (n = 3149; p.a.) Cipro C'taxime Gentamicin

24 E.coli - ESBL prevalence segmented logistic regression p = 0.15 OR = 0.86 p < OR = 1.76 Q Observed, mean for year Model BSAC Bacteraemia

25 Ceph-R mechanisms BSAC % Prevalence E. coli Other/undefined AmpC non-ctx-m ESBL CTX-M Livermore et al., JAC 2008; 62 2: ii41-54, updated

26 Carbapenemase-producing Enterobacteria: BSAC bacteraemia Enterobacter, KPC-4.?Same patient Klebsiella, NDM Enterobacter, IMP Enterobacter, VIM Enterobacter, IMI Klebsiella, VIM

27 Carbapenem resistant GNBs: Referrals to National Reference Laboratory IMP VIM KPC OXA- 48 NDM IMI KPC- VIM Data access

28 log2(mic) Vancomycin MIC creep bacteraemia isolates Original MICs for 271 MRSA Geometric mean with 95% CI, conventional calculation Vancomycin fitted line = log2 units/year; p = Teicoplanin fitted line = log2 units/year; p = Year

29 log2(mic) Vancomycin MICs for 271 MRSA - re-test Geometric mean with 95% CI, from interval regression Gradient of fitted line = log2 units/year; p = Year

30 Conclusions: What the BSAC Resistance Surveillance Programme has shown: Long term non-public sector surveillance of antimicrobial resistance is possible In CAP: - decline in penicillin non-susceptible S.pneumoniae with recent increase, with shift in serotype vaccine types down - significant increase in resistance to aminopenicillin H.influenzae In HAP - short term results so far - most resistance stable, ESBL producers, few carbapenem resistant Klebsiella - P.aeruginosa resistance higher in ICU especially carbapenem and piperacillin/tazobactam - S.aureus resitance not higher in ICU with drop in MRSA (MecA, ciprofloxacin, erythromycin resistance) /cont d

31 Conclusions cont d Bacteraemia - decline in MRSA - rise then plateauing and fall of CTXM - carbapenemase beginning to register - Not much happening with P.aeruginosa - vancomycin MIC shift in MRSA is not a problem.

32 Thanks to: Dr Ian Morressy, formerly of Quotient Professor David Livermore Dr Rosy Reynolds Also: - Collecting laboratories - Central laboratories - Website provider (MRS) - Working Party Members - Commercial Sponsors - BSAC