Prevalentie en karakteristieken

ESBLAT Symposium 2018 Prevalentie en karakteristieken Veilig voedsel produceren ESBLs humaan Marc Bonten UMCU/RIVM @MarcBonten https://reflectionsipc.com

Contents 1. Prevalence of carriage with ESBL-producing bacteria. 2. Risk factors for carriage. 3. Transmission of ESBL-producing bacteria. 4. Prevalence and incidence of infections with ESBLproducing bacteria. 5. Consequences of infections with ESBL-producing bacteria.

Prevalence of ESBL carriage in healthcare settings 1. 9.5% among 337 inhabitants from 8 Long-Term Care Facilities (LTCF) 4.2% E. coli, 5,2% K. pneumoniae (all in 1 LTCF) Courtesy Linda Verhoef, RIVM, SNIV study 2. 10.9% among 643 inhabitants of 9 LTCF Willemsen et al. Antimicrob Resist Infect Control. 2014;3:26 3. 8.2% among 1,351 patients at hospital admission 8.6% when admitted from LTCF 7.9% when admitted from home Platteel et al. Clin Microbiol Infect. 2015; 21: 141-6 4. 7.7% among 1,871 ICU patients receiving SOD 4.4% among 1,928 ICU patients receiving SDD Oostdijk et al. JAMA, Oct 2014

Prevalence of ESBL carriage in Dutch population 1. 4.5% (95%CI 3.4-6.0%) daycare children and parents (van der Bunt et al, J Antimicrob Chemother. 2016;10:XXX) 2. 4.5% (95% CI 3.3-6.6) ESBL/AmpC prevalence among children attending daycare centers; (Koningstein M, et al. J Pediatric Infect Dis Soc. 2015;4(4):e93 9) 3. 4.5% (95% CI 3.7-5.4) ESBL/AmpC prevalence among adults in living in livestockdense area; (Wielders C., et al. Clin Microbiol Infect. 2017;23:120e1 8) 4. 5.1% (95% CI 3.8-6.6) overall ESBL/AmpC prevalence in a general population (high and low broiler density area); (Huijbers PMC, et al. Clin Microbiol Infect. 2013;19(6):E256 9) 5. 8.2% (95% CI 6.9-9.8) ESBL carriage elderly at hospital admission (Platteel TN, et al. Clin Microbiol Infect. 2015;21:141 146) 6. 8.6% (95% CI 6.2-12.0) ESBL carriage before travelling (Paltansing S, et al. Emerg Infect Dis. 2013) 7. 8.6% (95% CI 7.3-10.0) ESBL carriage among citizens of Amsterdam (Reuland EA, et al. J Antimicrob Chemother. 2016;10:1-7) 8. 10.6% (95% CI 9.7-11.5) patients visiting GP with gastrointestinal complaints (Reuland EA, et al. Clin Microbiol Infect. 2013;19(6):542 9)

ESBLAT Study design ESBLs in humans (Monthly repeated crosssectional) Prevalence and genotypes in the Dutch human population Risk factors for ESBL carriage in humans ESBLs in pets (Monthly repeated crosssectional) Pets Prevalence and genotypes of ESBLs in Dutch dogs and cats Risk factors for ESBL carriage in dogs and cats Persistence of ESBLs in humans and pets (follow-up after 1 and 6 months after first sample) UMCU/RIVM UU Faculty of Veterinary Medicine

ESBLAT Study design: inclusion Dutch general population Humans Invited: n=47.957 Participation y/n? Y N Willing to participate in Human Faecal sampling? Faecal sample Y N Questionnaire & faecal sample N= 4.177 (8,7%) Only questionnaire N= 4.611 (9,6%) No information N=39.169

Representativeness of ESBLAT study population Source: CBS 2017 Netherlands 2015 N=16,900,726 ESBLAT N=4,177 Difference <20 years 22.7% 14.5% -8,2% 20 to 39 years 24.5% 12.5% -12,0% 40 to 64 years 35.1% 45.0% +10,1% 65 to 79 years 13.4% 24.7% +11,3% >80 years 4.3% 3.2% -1,1% Average age population 41.3 49.8 +8,5yr Parents both born in the Netherlands 78.3% 97.5% +19,2% Gender (Male) 49.5% 45.5% +4% Degree of urbanization (adresses per km2) >= 2500 22.7% 5.4% -17,3% 1500-2500 24.8% 15.0% -9,8% 1000-1500 17.3% 19.5% +2,2% 500-1000 17.6% 27.9% +10,3% < 500 17.6% 32.2% +14,6% County Groningen / Friesland / Drenthe 10,2% 12,4% +2,2% Overijssel / Flevoland / Gelderland 21,1% 24,3% +3,3% Utrecht / N-Holland / Z-Holland 45,1% 37,5% -7,6% Zeeland / N-Brabant / Limburg 23,6% 25.,6% -2%

Participants aggregated per municipality

ESBLAT Results Prevalence of ESBL carriage: 186/4177 4.5% (95%CI: 3.9-5.1%) E. coli: 178 (93.7%) K. pneumonia: 10 (5.2%) E. cloacae complex: 2 (1.1%) Adjusted for: Prevalence (%) 95% CI Gender 4.5 3.9-5.2 Age 4.3 3.7-5.0 County 4.5 3.9-5.2 Degree of urbanisation 4.8 4.1-5.5 Nationality 4.9 4.3-5.6 E. coli genotypes: blactx-m15: 85 (44.5%) blactx-m1: 30 (15.7%) blactx-m14: 31 (16.2%) blactx-m27: 27 (14.1%) E. coli sequence types: ST131: 42 (23.6%) ST38: 22 (12.4%) ST10: 14 (7.9%)

ESBL positieve participants

Univariately tested variables (of 86 variables) General characteristics: - Gender - Country of birth is other than the Netherlands - Age - Social Economic Status (SES) - Degree of urbanization - Season of sampling - Year of sampling - Level of education Health care related: - Admission to hospital in 4 weeks prior to sample collection - Hospitalization in the past 12 months - Used antibiotics in 8 weeks prior to sample collection - Used antibiotics in the past (never, <6 months, 6-12 months, >12 months) - Used of proton pump inhibitors in the past 6 months Exposure related: - Attending day care of at least 1 child in household - Ate raw or undercoocked meat 1 week in the past week - Eat no meat - Eat pork - Eat in a restaurant more often than 20 times a year - Being abroad: - in the 4 weeks prior to sample collection - in the past 12 months - Animals in or around the household - Not changing the kitchen towel on a daily basis - Swam: - in sea or ocean ('salt' water) in the past 4 weeks - in sea or ocean ('salt' water) in the past 12 months - in open fresh water (lakes, river, etc.) in the past 4 weeks - in open fresh water (lakes, river, etc.) in the past 12 months

ESBLAT: Risk factors identified univariate multivariate* Variable OR 95% CI OR 95% CI Gender (Male) 1,25 0,93 1,68 1,44 1,01 2,05 Country of birth is other than the Netherlands 2,32 1,28 4,19 2,34 1,18 4,61 Eat in a restaurant more often than 20 times a year 1,96 1,33 2,90 1,74 1,09 2,78 used antibiotics in the past never <6 months 2,43 1,28 4,60 2,06 1,02 4,19 6-12 months 1,66 0,96 2,87 1,64 0,73 3,68 >12 months 1,64 0,77 3,48 1,43 0,79 2,59 Ate pork 0,61 0,37 1,02 Swam in sea or ocean ('salt' water) in the past 12 months 1,56 1,14 2,13 1,46 1,0 2,14 Not changing the kitchen towel on a daily basis 2,20 1,41 3,43 2,15 1,27 3,62 Travelled (in the past 12 months): Not travelled to Europe 1,19 0,83 1,69 0,83 0,54 1,29 to Africa 3,85 1,73 8,5 3,03 1,23 7,46 to Asia 3,73 2,18 6,35 2,00 1,02 3,90 to North-America 1,34 0,52 3,46 0,17 0,02 3,62 *controlled for the following variables: age (0-4, 5-12, 12-19, 20-39, 40-64, 95-79, >80) and season (autumn(september November), winter (December February), spring (March May), summer (June August)).

Prevalence ESBLAT: Traveling to Africa and Asia in the past 12 months 35,00 30,00 25,00 20,00 15,00 10,00 5,00 0,00 ESBL prevalence in participants travelled in the past year % ESBL Continent ESBL N % ESBL Africa 8 65 12,31 Asia 22 184 12,00 Europe 100 2410 4,15 North- America 5 107 4,67 Oceania 0 11 0 South- America 1 22 4,55 Continent

ESBLAT: other risk factors Eat more than 20 times a year in a restaurant Associated with travel, higher SES Swam in sea or ocean ('salt' water) in the past 12 months Associated with travelling Not changing the kitchen towel on a daily basis Associated with Not changing the dish towel on a daily basis, Not always cleaning hands before preparing food, Not always cleaning hands after toilet visit

Percentage ESBLAT: PPI not a risk factor 45 40 Reuland EA, et al. J Antimicrob Chemother. 2016;10:1-7 35 30 25 20 15 10 5 0 ESBLAT: Proton Pump Inhibitor Use per age category 0-4 5-12 13-19 20-39 40-64 65-79 >80 Age category and 5,3% in PPI (32/599) % PPI use Subgroup analysis in 40+ group: ESBL + 4,4% (105/2395) in non-ppi

R 0 R 0 is the average number of new cases of an infection caused by one infected individual in a population of susceptibles R 0 > 1 : each infected spreads the infection to more than one other person/animal: chain reaction = epidemic R 0 < 1 : on average an infected does not replace itself in the infected population: infection cannot grow

633 (34 3%) of 1847 travellers had acquired ESBL-E during international travel, especially in southern Asia (75 1%). Median duration of colonisation after travel was 30 days. 65 (11 3%) of 577 remained colonised at 12 months. Onward transmission was found in 13 (7 7%) of 168 household members. The probability of transmitting ESBL-E to another household member was 12%. Calculated effective R0 was around 0.2 (Martin Bootsma personal communication), which might include some overestimation due to false-positive transmission events (no molecular typing).

Acquisition of ESBL carriage in Dutch hospitals Rectal carriage of ESBL-E Prevalence at admission % [95% CI/CrI] Prevalence at discharge % [95% CI /CrI] Hospital-acquired prevalence at discharge % [95%CI/Crl] Acquisition rate n/1000 patientdays at risk [95% CI/CrI) SoM study 14 hospitals 8,400 admissions 7.4% [6.2%-8.7%] 9.9% [8.2%-11.8%] 2.5% [1.7%-3.6%] 3.2 [2.2-4.5] R-GNOSIS study Single hospital 5,450 admissions 6.4% [5.2%-7.8%] 8.7% [6.8%-11.0%] 2.3% [1.3%-3.6%] Results: MCMC model 28.5% cross-transmission 71.5% endogenous selection R= 0.06 2.7 [1.4-4.2] Mathematical model (data R- GNOSIS study) 7.0% [6.2%-7.8%] 9.3% [8.6%-10.0%] 2.3% [1.7%-2.9%] 3.8 [2.9-4.9] Kluytmans-van de Bergh, Mens, et al. ICHE 2017

% Resistamt to 3rd gen cephalosporins Trends in prevalence of ESBL-producing bacteria as a cause of bloodstream infections in ISIS-AR 14 12 10 8 6 4 2 0 2010 2011 2012 2013 2014 2015 2016 2017 Data from 21 labs E. coli: 3,257-4,035/yr K. pneumoniae: 507-719/yr E. coli K. pneumoniae

Effect of national policy implemented in 2010 Sales data of antimicrobials for use in animals in the Netherlands DDD/AY for different sectors 50 45 40 DD/DJ en DDDAnat 35 30 25 20 15 68.9% reduction (2007-2016) 64.4% reduction in 2016 to reference year 2009 Fluoroquinolones and 3 rd /4 th -gen cefalosporines usage reduced to a minimum 68% reduction in use of colistin (2011-2015) 10 5 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Purple: turkeys Blue: veal Orange: broilers Light green: pigs Dark green: dairy Jaar 26

Clinical suspicion of GNB infection (BC + start Abs) N=9,422 GNB bacteremia N=773 (8,2%) GNB: 3GCeph-S N=709 (7,2%) GNB: 3GCeph-R N=64 (0,7%) 3GCeph-R/all GNB N=64/773 (9%) Rottier et al. CID, 2015

Courtesy: W. Rottier The association between antibiotic resistance and outcome M 2 : mortality M 2 G 1 : appropriateness of empiric therapy G 1 R 0 : resistance of bacterial strain R 0

U -1 X* -1 M 2 : mortality M 2 E -1 T 1 X* -1 : underlying disease U -1 : measurable X* -1 E -1 : exposure to healthcare A -1 : exposure to antibiotics C* -1 : colonizing bacterial strain K -1 : known colonization with bacterial strains A -1 C* -1 K -1 C* 0 F 0 P 0 V* 0 R 0 D 1 G 1 C* 0 : bacterial strain causing infection K 0 : known properties of C* 0 F 0 : infection source P 0 : sepsis severity V* 0 : virulence of bacterial strain R 0 : resistance of bacterial strain K 0 D 1 : source control T 1 : empiric antibiotic therapy G 1 : appropriateness of empiric therapy

To retrospectively - determine associations between appropriateness of initial antibiotic treatment and other factors associated with day 30 mortality in patients with ESBL bacteremia in 8 Dutch hospitals. 232 consecutive patients (bacteremia due to E. coli, K. pneumoniae and Enterobacter spp) between 2008 and 2010.

Variable Variables associated with day-30 mortality Unadjusted OR in univariate analysis Adjusted OR in multivariable analysis Appropriate therapy No (vs Yes) 1.12 (0.57 2.19) 1.53 (0.68 3.45) Charlson index >3 (vs <3) 2.16 (1.13 4.16) 2.80 (1.21 6.51) Patient age >75 yr (vs <75) 2.35 (1.17 4.72) 3.81 (1.55 9.39) Hospital ward at bacteremia onset ICU (vs other) 3.99 (1.85 8.64) 2.88 (1.05 7.85) Bacteremia source Non-UTI (vs UTI) 4.31 (1.91 9.71) 4.79 (1.74 13.16) Severe sepsis Yes (vs No) 7.24 (3.53 14.71) 5.24 (2.36 11.60)

GRAND-ABC: Burden of Infections Caused by Antibiotic Resistant Gram-Negative Bacteria in the Netherlands Courtesy: Wouter Rottier

Conclusions (1) 1. Prevalence of human carriage with ESBL-producing bacteria. 4,5% in open population 8% in LTCF 8% in hospital 4% in ICU 2. Risk factors for carriage in open population. Travelling (esp Asia and Africa) Salt water exposure (linked to travelling) Restaurant exposure (linked to travelling) Antibiotic use Non-hygienic kitchen behavior

Conclusions (2) 3. Transmission of ESBL-producing bacteria. Within household: R 0,2 Within hospital: R 0,06 4. Prevalence and incidence of infections with ESBLproducing bacteria. 5% among invasive E. coli infections 8% among invasive K. pneumoniae infections 0.7% when starting empiric antibiotic therapy for sepsis Prevalence stable during last 7 years 5. Consequences of infections with ESBL-producing bacteria. Transmission of ESBL-producing bacteria. No evidence of increased mortality or increased healthcare burden

Acknowledgements Department of Medical Microbiology, University Medical Center Utrecht (UMCU), Utrecht, the Netherlands Gerrita van der Bunt A.C. Fluit L. Hidalgo J. Scharringa J. Vlooswijk L.J.L. Muller G.M.A. Riemens-Zetten Centre for Infectious Disease Control, National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands: W. Van Pelt L. Mughini-Gras A. Mulder Central Veterinary Institute (CVI) of Wageningen University, Lelystad, the Netherlands D.J. Mevius Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands J. Hordijk M.P. Spaninks A.J. Timmerman J.A. Wagenaar Workgroup ESBL Attribution (ESBLAT): Utrecht University-Institute for Risk Assessment Sciences Utrecht University Department of Infectious Diseases & Immunology University Medical Centre Utrecht National Institute for Public Health and the Environment GD-Animal Health VION-FOOD Group van Drie Group

Hypothesis: Infections caused by antibiotic resistant bacteria increase the risk of recurrent infetion. Aim: To quantify the association between antibiotic resistance and the occurrence of recurrent bacteremia. Data source: All blood cultire isolates of Enterobacteriaceae in ISIS- AR, covering 60% of all microbiological culture results since 2010 (analysis till Sept 2016).

ESBL ESBL+CRE ESBL ESBL+CRE The absolute risk of a recurrent infection caused by ESBL or ESBL/CRE after a first episode caused by a susceptible isolate was 0.1% (74/67 608 episodes). The estimated number of recurrent bacteremia episodes with ESBL or ESBL/CRE in the Netherlands is 19/year.

we used genomics to analyze a systematic 11-yr hospitalbased survey of E. coli associated with bacteremia using isolates collected from across England by the British Society for Antimicrobial Chemotherapy and from the Cambridge University Hospitals NHS Foundation Trust.

Proportions of STs during the 11-yr sampling framework. Teemu Kallonen et al. Genome Res. 2017;27:1437-1449 2017 Kallonen et al.; Published by Cold Spring Harbor Laboratory Press

% Resistamt to 3rd gen cephalosporins 15 10 5 0 E. coli K. pneumoniae

Prevalence and relative contributions of hospital acquisition, travel acquisition, within-household transmission, and between-household transmission to the prevalence of ESBL. B/W, between-household transmission rate defined as a fraction of the within-household transmission rate.