17 / 150
ID: 4865
Antimicrobial management of E. coli bacteraemia
Mariha Hamid
1
, Paul Wade
2
, William Newsholme
2
.
1
King
’
s College
London,
2
Guy
’
s and St Thomas
’
NHS Foundation Trust
Background:
Inappropriate antimicrobial use is a significant risk
factor in the spread of antimicrobial resistance (AMR); this, coupled
with a recent increase in the rate of E. coli bacteraemia and associated
antibiotic resistance, reinforces the need to monitor adherence to
clinical antibiotic guidelines.
Aim(s)/Objective(s):
To explore the cases of E. coli bacteraemia in
Guy
’
s and St Thomas
’
NHS Foundation Trust reported between March
and September 2015, in order to determine whether Trust guidelines
were followed with regards to empirical antibiotic therapy, and
whether antibiotic sensitivity patterns were used to adjust antibiotic
choice.
Method(s):
The patient sample was obtained from the Public Health
England surveillance programme data. Data on prescribing and
antibiotic sensitivity patterns were obtained from Trust electronic
systems and compared to Trust antimicrobial guidelines.
Results:
The analysis included 89 patients. Overall, 63% of cases
adhered to guidelines. 31% of patients required an antibiotic
adjustment based on sensitivity patterns, and appropriate alternatives
were given in 75% of these cases. Co-amoxiclav was frequently used as
monotherapy for cases where empiric guidance recommended dual
antimicrobials (such as gentamicin e.g. in severe sepsis). The majority
of patients who did not receive an appropriate antibiotic adjustment
had been given co-amoxiclav despite reported resistance.
Discussion and/or Conclusion(s):
Routinely documenting the
acknowledgment of blood culture results may prompt clinicians to
identify resistance patterns and adjust subsequent management
decisions, reducing the number of patients who are continued on
inappropriate antibiotics. This will be important for adherence to the
72-hour review component of the new AMR Commissioning for
Quality and Innovation guidance.
ID: 4894
Acquired colistin resistance found in 2% of human
enterobacteriaceae in Singapore
Pooja Rao, Jen Mee Long, Wen Ying Tang, Timothy Barkham.
Tan Tock
Seng Hospital
Background:
Colistin has increased in importance over the last decade
as multi drug resistant Gram negative bacteria have become more
numerous and more difficult to treat. The report of acquired colistin
resistance due to MCR-1, carried on a plasmid, heralds an era of
therapeutic poverty and stimulated a search for colistin resistance in
many countries to define the extent of MCR-1 in their populations.
Having acquired a positive control for MCR-1 in May 2016, we found
MCR-1 in three out of four enterobacteriaceae that had been flagged as
colistin resistant on the Vitek system in mid-June 2016.
Aim(s)/Objective(s):
The aim was to define the possible extent of
colistin resistance in our population.
Method(s):
We extracted colistin MIC data from the Vitek suscepti-
bility testing system for a sixmonth period fromDec 2015 toMay 2016.
Data for organisms naturally resistant to colistin were discarded.
Colistin resistance was defined as an MIC >2 mg/L.
Results:
Datawere available for 3887 enterobacteriaceae. 78 (2%) were
colistin resistant with MICs as follows. MIC < 0.5 mg/L, 3768; 1 mg/L,
22; 2 mg/L, 19; 4 mg/L, 17; 8 mg/L, 34;
≥
16 mg/L, 27.
Discussion and/or Conclusion(s):
Polymyxin is used instead of
colistin in Singapore, and polymyxin MICs are measured with an E
test, when clinically indicated, and interpreted with CLSI breakpoints.
Colistin results, although on our Vitek system, did not flow through to
the laboratory information system and constituted important but
hidden data. Many of these colistin resistant bacteria are not multi
drug resistant. Further work will define what proportion carry MCR-1
and how widespread this is in the community.
ID: 4961
Relationship between resistance in coliform bacteraemia and
antimicrobial exposure
Jennifer Wood
1
, Peter Davey
2
, Bruce Guthrie
2
,
Virginia Hernandez-Santiago
2
, Charis Marwick
3
.
1
University of Dundee,
2
University of Dundee,
3
University of Dundee/NHS Tayside
Background:
Antimicrobial resistance (AMR) is a serious public
health threat, with emerging resistance among coliforms particularly
concerning. Antimicrobial use contributes to this but associations
between individual past exposure and risk of AMR infection are not
well understood.
Aim(s)/Objective(s):
The aim is to quantify associations between
antimicrobial resistance in coliform bacteraemia on admission to
hospital (positive blood culture on day 0
–
2) and antimicrobial
exposure in the prior 12 months, in a case-control study using linked
routine data enhanced by case-note review. Cases are those with
resistant infections and controls those with susceptible infections.
Method(s):
The study includes all NHS Tayside residents admitted
with coliform bacteraemia (first episode if >1) from Jan 2011 to Dec
2015. Patient-level data on hospital admissions, microbiology results,
GRO deaths, and prescriptions dispensed from community pharmacies
are anonymously linked by the Health Informatics Centre (HIC).
Hospital inpatient prescribing data are not captured electronically
so extracted by case-note review, then anonymised and linked.
Logistic regression (univariate then multivariate) will generate odds
ratios for the likelihood of resistant versus susceptible (e.g. to co-
amoxiclav) bacteraemia associated with prior exposure (e.g. any
exposure/cumulative exposure/most recent exposure to beta-lactam/
beta-lactamase inhibitor combinations). Potential confounders in
multivariate analyses are: age, gender, comorbidity, deprivation, care
home residence, previous hospital admissions.
Results:
There are 372 eligible bacteraemia episodes, 204 (54.8%) with
admissions in the previous year. The complete dataset is almost
assembled foranalysis and the resultswill be available for presentation.
Discussion and/or Conclusion(s):
Findings will help to inform
prescribing policy, enabling tailored therapy for at risk individuals.
ID: 4963
A new scoring system to predict the risk of multiresistant
pathogens in pneumonia and sepsis validated using
‘
big data
’
analyses
Michael Wilke
1
, Nadine Tränkner
1
, Kerstin Worf
1
, Silvia Wilke
1
,
Wolfgang Heinlein
1
, Klaus-Friedrich Bodmann
2
.
1
Inspiring-Health,
Dr. Wilke GmbH,
2
Klinikum Barnim, Werner-Forßmann Krankenhaus
Background:
Hospital-acquired pneumonia and sepsis pose a major
challenge in intensive care medicine. A key factor for the survival of
patients is the early administration of adequate initial antibiotic
therapy, which covers the causative pathogens, especially if there are
pathogens with multiple resistances (MDR). The correct assessment
of the risk for MDR presence is a key strategy in antimicrobial
stewardship. Multiple risk factors are described in the literature, hence
few studies exist that did a systematic validation of those factors by
using large data collections.
Aim(s)/Objective(s):
The aim of this study is to develop and validate
a risk scoring system that is able to predict the presence of MRE in
these disease entities.
Method(s):
A systematic literature research based on the PICO-
principle was used to identify the risk factors for hospital-acquired
pneumonia and sepsis described in the literature. A number of risk
factors could be identified. The risk factors were
–
if possible
–
described by the use of routine data (OPS, ICD, etc.) that are easily
available and will be validated on a large amount of patient cases (55
hospitals with 3.7 Million hospitalizations over 5 years). Using the
valid risk factors a score will be developed using multivariate
regression models and Receiver-operator Curves (ROC) to allow a
better risk stratification in the initial phase of the therapy.
Abstracts of FIS/HIS 2016
–
Oral Presentations / Journal of Hospital Infection 94S1 (2016) S11
–
S21
S12