1. CPE screening: seek and ye shall find!
Jon Otter, PhD FRCPath
Imperial College London
@jonotter
Blog:
You can download these slides from

2. Rising threat from MDR-GNR
% of all HAI caused by GNRs.
% of ICU HAI caused by GNRs.
Non-fermenters
Acinetobacter baumannii
Pseudomonas aeruginosa
Stenotrophomonas maltophilia
Enterobacteriaceae
Klebsiella pneumoniae
Escherichia coli
Enterobacter cloacae
CPO
CPE
Hidron et al. Infect Control Hosp Epidemiol 2008;29:
Peleg & Hooper. N Engl J Med 2010;362:

3. National trends: mandatory surveillance

4. Emergence of CPE in the UK
PHE.

5. Otter et al. Clin Microbiol Infect 2015 2015;21:1057–1066.
CPE Toolbox
Hand hygiene
Cleaning / disinfection
HCW screening
Decol.
Cohorting staff / patients
Note flagging
Education
Env. screening
Contact precautions
Active screening
Antibiotic stewardship
Otter et al. Clin Microbiol Infect ;21:1057–1066.

6. Key questions Which anatomical site to screen?
Which patient group to screen?
How to sample patients?
Is serial admission screening useful?
Is pre-emptive isolation feasible?
Should we de-isolate known carriers?
How much does screening cost?
Which lab method is best?

7. Which anatomical site to screen?
Paired rectal and perineal swabs from the same individuals yielded ESBL-E in 7.8% of rectal swabs vs. 3.8% of perineal swabs, p<0.001
Dyakova et al. Clin Microbiol Infect 2017 in press.

8. Which patient group to screen?
UK PHE CPE Toolkit screening triggers:
an inpatient in a hospital abroad, or
an inpatient in a UK hospital which has problems with spread of CPE (if known), or
a‘previously’positive case.
Also consider screening admissions to high-risk units such as ICU, and patients who live overseas.

9. Universal admission screening in London
All patients within the first 72 hrs of admission (excluding paediatrics)
Rectal swab
CRO cultured on MacConkey plus erta (reference method)
CRE cultured on Chrome plate
CPO detected by PCR (Check Direct CPE*)
Perineal swab
ESBL cultured on Chrome plate
Each patient approached and verbal consent obtained; risk factor questionnaire completed. Target sample size: 4500.
* PCR+ samples repeated on Cepheid PCR.
The study was approved by the NHS Research Ethics Committee.
Otter et al. J Antimicrob Chemother 2016;71:

10. Universal admission screening in London
4843 patients enrolled.
Rectal swabs collected from 4207 patients.
CPE cultured from 5 (0.1%) patients.
All were positive by Cepheid, 4/5 positive by CheckDirect.#
Risk factors were overseas hospitalisaiton anywhere, or in a PHE risk country.
Samples from 2 patients were PCR+/culture negative by both PCR systems (Cepheid and CheckDirect).
CPO identified in 7 (0.2%) patients.
Samples from a further 75 patients were culture negative, and PCR+ by CheckDirect but negative by Cepheid.
Working hypothesis: false positives.
Otter et al. J Antimicrob Chemother 2016;71:

11. ‘Standard of care’ screening
Risk factors
Number of approached patients
Patients in isolation %
Standard of care screening*
Number of patient with at least one risk factor (out of all approached patients)
3874
253
6.5%
250
Non-UK residents 55
0.0% 4
7.3%
Overseas travel in the past 12 months
1524 77
5.1% 78
Overnight hospital stay in the past 12 months - GSTT
1658
135
8.1% 88
5.3%
Overnight hospital stay in the past 12 months - within M25
1964
153
7.8%
112
5.7%
Overnight hospital stay in the past 12 months - North West 8 1
12.5%
Overnight hospital stay in the past 12 months - any UK hospital (including London)
2187
167
7.6%
165
7.5%
Overnight hospital stay in the past 12 months - overseas hospital (CRE risk countries) 20 2
10.0% 5
25.0%
Overnight hospital stay in the past 12 months - overseas hospital (any country) 49
10.2%
* Within the first 48 hours of admission.
Otter et al. J Antimicrob Chemother 2016;71:

12. Risk factor prevalence
n pts
% pts
Non-UK residents 55
1.2%
Overseas travel in the past 12 months
1524
32.4%
Overnight hospital stay in the past 12 months - GSTT
1658
35.3%
Overnight hospital stay in the past 12 months - within M25
1964
41.8%
Overnight hospital stay in the past 12 months - North West 8
0.2%
Overnight hospital stay in the past 12 months - any UK hospital (including London)
2187
46.5%
Overnight hospital stay in the past 12 months - overseas hospital (PHE risk countries) 20
0.4%
Overnight hospital stay in the past 12 months - overseas hospital (any country) 49
1.0%
Antibiotics in the past 6 months - any
2628
55.9%
Antibiotics in the past 6 months - one course
1399
29.8%
Antibiotics in the past 6 months - more than one course
1229
26.1%
At least one risk factor
3618
77.0%
At least one risk factor (excluding antibiotics)
2961
63.0%
n=4701.
Otter et al. J Antimicrob Chemother 2016.

13. Distant large problems vs. small local ones?
CPE introductions come from hospitals within a regional referral network, even if the prevalence in another referral network is much higher (more than 100x higher, in fact)!
Donker et al. BMC Med 2017.

14. How do I screen? Rectal swab is the best sample
Insert no more than 2cm into rectum
Twist gently and withdraw
Ideally want to see faeces on swab.
Patient and staff education as to why this is needed in order to overcome taboos
Alternate specimen is stool sample, but have to wait for the patient to ‘go’
Thank you Jon,
So how do I screen a patient for CRE?
A rectal swab provides the best results. The swab should be inserted into the rectum no more than 2 cm or an inch twisted gently and withdrawn. The aim is to stain it brown!
Unfortunately , not many people are happy to have a sample taken in this way. And Issues such as child protection have been raised.
Clear explanations are needed to gain patient understanding and informed consent and we need some public education for us all to make this more normal and acceptable.
In the meantime
If it is really not possible to get a rectal sample then a stool sample can be used. This is not quite as good as apart form anything it may be delayed while we wait for the patient to ‘perform’

15. Can I swab your rectum please?
Factors associated with patients declining to provide a rectal swab were:
younger age (odds ratio (OR) 0.99, 95% confidence interval (CI) ) female gender (OR 1.26, CI ),
transfers from other hospitals (OR 1.77, CI ) or an unknown admission route (OR 1.61, CI ),
admission before the change in study description (OR 0.39, CI )
the staff member who consented the patient (p<0.001);
ethnicity was not a significant factor.
Dyakova et al. Clin Microbiol Infect 2017 in press.

16. Improving screening compliance
Dyakova et al. Clin Microbiol Infect 2017 in press.

17. Serial admission screens?
Vella et al. J Hosp Infect 2016;94:

18. Serial admission screens?
Gram-negative bacteria
Enterobacteriaceae
Resistant Enterobacteriaceae*
CPE
* Resistant to ertapenem, meropenem, temocillin or tazocin.
Mookerjee et al. Submitted.

19. Carriage rate by screening time-point*
* = significant trend-change: p<0.05.
Mookerjee et al. Submitted.

20. Carriage rate by screening time-point (patients who received all 3 screens)
n=221. No significant trend-change for any of the organism-groups: p>0.05.
Mookerjee et al. Submitted.

21. Carriage rate for patients who received 3 screens at any time-point* * *
n=1597. * = significant trend-change: p<0.05.
Mookerjee et al. Submitted.

22. Who’s carrying CPE on admission?
Author
Year
Location
Setting
n patients
n carriers
% carriers
Lin
2013
Chicago, USA
Long term acute care hospitals
391
119
30.4 –
Short stay hospital ICU
910 30
3.3
Day
Pakistan
Patients attending a military hospital
175 32
18.3
Adler
2015
Israel
CPE carriage in post-acute hospitals, 2008
1147
184
16.0
CPE carriage in post-acute hospitals, 2013
1287
127
9.9
Girlich
2014
Morocco
Hospitalized patients 77 10
13.0
Armand-Lefèvre
Paris, France
ICU patients 50 6
12.0
Rai
East Delhi, India
Outpatients
242 24
Birgand
Patients repatriated or recently hospitalized in a foreign country
132 9
6.8
Zhao
Fujian, China
Stool samples from hospitalized patients
303 20
6.6
Swaminathan
New York
All admissions to 7 units, including ICU, of 2 hospitals
5676
306
5.4
Wiener-Well
2010
Jerusalem, Israel
298 16
Villar
Buenos Aires, Argentina
Non-hospitalized individuals
164 8
4.9
Kothari
New Delhi, India.
Healthy neonates 75 1
1.3
Dyakova
2016
London
‘High-risk’ admissions
7666 37
0.5
Mack
‘High-risk’ inpatients and admissions.
2077 7
0.3
Kim
Seoul, Korea
ICU admissions
347
Nüesch-Inderbinen
Zurich, Switzerland
Healthy community residents and outpatients
605
0.0
All hospital admissions
4218 5
0.1
For refs see:

23. Simple, stark, sobering sums
0.5%1 x 186,393 = 932 (!) 0.1%2 x 186,393 = % x m* = 15,892
Taking our carriage rate of CPE – 0.5% and applying it to our yearly patient admissions, means we can see close to 900 CPE positive patients
0.1% which is a conservative estimate, apply that, and its 186
0.1 is the conservative estimate, applying that to all NHS hospitals, 16,000 pos patients, currently reporting around1600, 10 fold underestimate of CPE positive patients
* Admissions to NHS acute hospitals, Financial Year 14/15. NHS Confederation, Key Statistics on the NHS,
Mookerjee et al. ECCMID 2016.
Otter et al. J Antimicrob Chemother 2016 in press.

24. (In)feasibility of pre-emptive isolation
Vella et al. J Hosp Infect 2016;94:

25. Deisolation? Author Year Setting N pts Organism
Duration of colonization
Bird1
1998
Elderly care facilities, Scotland 38
ESBL K. pneumoniae
Mean 160 days (range 7-548)
Pacio2
2003
Long term care facility, USA 8
Resistant Gram-negative rods
Median 77 days (range )
Zahar3
2010
Paediatric hospital, France 62
ESBL Enterobacteriaceae
Median 132 days (range )
O'Fallon4
2009 33
Median 144 days (range 41–349)
Zimmerman5
2013
Patients discharged from hospital, Israel 97
CRE
Mean 387 days
Bird et al. J Hosp Infect 1998;40:
Pacio et al. Infect Control Hosp Epidemiol 2003;24:
Zahar et al. J Hosp Infect 2010;75:76-78.
O'Fallon et al. Clin Infect Dis 2009;48:
Zimmerman et al. Am J Infect Control 2013;41:

26. ‘Selective’ digestive decontamination
20 CRE colonized patients in each arm given gentamicin + polymyxin (SDD arm) or placebo (Control arm)
Control
SDD
Saidel-Odes et al. Infect Control Hosp Epidemiol 2012;33:14-19.

27. Antibiotics have a profound and sustained effect on the human microbiome (even those that are typically associated with no or few side effects). This results in a reduction in both diversity and change in composition, which is bad news for human health. In particular, this leave the gut more open to colonization with unwanted intruders aka antibiotic resistant bacteria.

28. Decolonisation using faecal microbiota transplantation (FMT)
82 year old colonised with CPE.
Carriage was delaying her admission to a nursing home.
Single dose of FMT decolonised her at 7 and 14 days.
Laiger et al. J Hosp Infect 2015 in press.
Buffie & Pamer. Nat Rev Microbiol 2013;13:

29. Screening costs – an illustration
Over a 10 month period, ~10,000 screens were taken during the management of an NDM outbreak.
The laboratory calculated £8 per screen, based on laboratory materials.
This figure is heavily dependent on prevalence, because the cost of processing increases: CPE > CRE > Negative.
It was assumed that it took a B5/6 nurse 15 minutes to collect a specimen.
Total of ~£80k actual expenditure, and £50k opportunity cost in staff time.
Also, £40k spent on gloves and aprons, and £24k on stock disposal related to isolation.
Otter et al. Clin Microbiol Infect 2017.

30. Cost hierarchy
Economic evaluation of a 40 case outbreak of CPE. Error bars represent range
Otter et al. Clin Microbiol Infect 2017.

31. Bubble size represents the total cost of the outbreak
Costs in context
Bubble size represents the total cost of the outbreak
£1m

32. Screening cost-effectiveness
Lapointe-Shaw et al. Eur J Clin Microbiol Infect Dis 2017;36:

33. Which lab method? Rectal swab Agar plate AST MADLDI-TOF MS WGS NAAT
(PCR)
NAAT = nucleic acid amplification techniques
AST = antimicrobial susceptibility testing
MALDI-TOF = Matrix-assisted laser desorption /ionization – time of flight mass spectrometry
WGS = whole genome sequencing

34. Before you throw away the agar plates…
Molecular diagnostics offer more sensitivity and shorter TAT:
but TAT may be longer in the real world than on paper;
are expensive;
rely on validation of carriage sites;
do not tell you about phenotypic susceptibility;
have a limit of detection often around a couple of logs – and sensitivity may be comparable to culture;
do not deal with changing epidemiology; struggle with target variability;
need to manage shared resistance genes between species, especially for MDR-GNR;
and is PCR+ culture- (as) clinically significant?
See further details in talk by Dr Dan Diekema

35. Summary
CPE combine resistance, virulence, the potential for rapid spread, and high costs.
Prevalence in the US and Europe appears to be patchy, but increasing; rates in parts of S. Europe are high.
We do not yet know what is effective in terms of prevention and control, but screening and isolation of carriers seems prudent.
Key questions around screening are: which patient groups to screen, the role of rapid diagnostics, and cost-effectiveness.
CPE is here to stay in London!

36. Agar plates
MacConkey
Selective for all Gram-negative bacteria (including Enterobacteriaceae and non-fermenters)
Chromogenic Media
Selective for resistant Enterobacteriaceae only (ESBL or CPE options available); several options

37. Antimicrobial susceptibility testing (AST)
Quantitative
(MIC or breakpoint)
Agar or broth dilution (manual or autmoated), E-tests
Qualitative
(R/I/S)
Disc diffusion; supplemental tests for mechanisms (e.g. ESBL, CPE)

38. MALDI-TOF, WGS, NAAT (PCR & Arrays)
Rapid and accurate speciation of bacteria from a colony; potential to detect resistance genes
WGS
Whole genome sequence; gold standard typing method; costs coming down; can detect abx resistance genes
PCR
PCR can be used to detect a single or multiple genes of interest from a pure colony
Array Chips
>100 PCRs on a single chip for simultaneous detection of a range of genes and markers

39. Which lab method? Rectal swab Agar plate AST MADLDI-TOF MS WGS NAAT
(PCR)
NAAT = nucleic acid amplification techniques
AST = antimicrobial susceptibility testing
MALDI-TOF = Matrix-assisted laser desorption /ionization – time of flight mass spectrometry
WGS = whole genome sequencing

40. NAAT direct from clinical specimens
PCR
Rapid real-time PCR kits available to detect resistance genes direct from clinical specimens; point of care tests coming
Rapid sequencing kits
Kits available for rapid sequence-based simultaneous detection of common organisms and resistance genes