

accuracy of newer technologies such as MALDI-TOF MS. In this
retrospective study we compare MALDI-TOF MS with biochemical
identification.
Aim(s)/Objective(s):
To assess the accuracy of MALDI-TOF MS for the
identification of
Neisseria gonorrhoeae
in comparison with biochem-
ical identification (Biomerieux API NH).
Method(s):
Laboratory records of all Genitourinary Medicine culture
samples between January 2012 and October 2014 were reviewed
retrospectively. All isolates with both an API NH and MALDI-TOF MS
identification recorded were included; a subset analysis examined the
reasons for requiring repeat MALDI-TOF MS analysis.
Results:
1190 isolates were eligible for inclusion, with 1082 biochem-
ically identified as
Neisseria gonorrhoeae
. MALDI-TOF MS identified
984 (91%) of these correctly after one analysis, rising to 1081 (99.9%)
after two analyses. Failure to generate an acceptable identification
score was the reason for repeat analysis in 76% of cases; the remaining
24% of isolates were either initially identified as other
Neisseriae
or as
genetically dissimilar contaminating organisms incorrectly sampled
from culture media.
Discussion and/or Conclusion(s):
MALDI-TOF MS is an accurate
and reliable method of identifying
Neisseria gonorrhoeae
. Laboratories
with access to this technology could consider its use as a single
method of identification in routine cases.
ID: 4743
Evaluation of the feasibility of using C-reactive protein (CRP) to
optimise prescribing for lower respiratory tract infections in
primary care settings
Jacqueline Sneddon
1
, William Malcolm
2
, Gail Haddock
3
,
Simon Hurding
4
, Alastair Leanord
5
, Dilip Nathwani
6
.
1
Healthcare
Improvement Scotland,
2
NHS National Services Scotland,
3
NHS Highland,
4
NHS Lothian,
5
The Scottish Government,
6
NHS Tayside
Background:
The Scottish Antimicrobial Prescribing Group has
developed several initiatives to promote reduction of unnecessary
antibiotic use for respiratory tract infections in primary care.
Aim(s)/Objective(s):
This study was developed to evaluate the feasi-
bility of using CRP testing, as recommended by NICE, to support
clinical decision making in lower respiratory tract infections (LRTI) in
GP practices in Scotland.
Method(s):
A study steering group was established to provide advice
on methodology and governance issues. Ten GP Practices were recrui-
ted across four health board areas to take part in the study. Following
training, practices used the CRP testing on patients presenting
with suspected LRTI for at least 4 weeks during the period
November 2015
–
February 2016. Data collected during consultations
and GP feedback were analysed to determine the practical aspects of
how the test was used and its perceived impact on GP decision making
and prescribing of antibiotics.
Results:
Data suggests that CRP testing can be accommodated with
the current appointments system, by either the GP running the test
during the consultation or a practice nurse performing the test in a
nearby treatment room. The test appeared to reduce the number of
immediate prescriptions, increase use of delayed prescriptions and
provided reassurance for both prescribers and patients when no
antibiotic was used. Study results will be used to inform future strategy
for point of care testing for community infections.
Discussion and/or Conclusion(s):
CRP testing is practical in the
Scottish primary care context and offers an additional tool to
support reduction of unnecessary use of antibiotics for respiratory
infections.
ID: 4757
Evaluation of Xpert Influenza assay; easy to use but moderate
sensitivity and no subtype data
Timothy Barkham
1
, Janice Leong
1
, Masafumi Inoue
2
.
1
Tan Tock Seng
Hospital,
2
Experimental Therapeutics Centre, Agency for Science,
Technology and Research
Background:
Testing for Influenza with nucleic acid tests (NATs) is a
daily activity. The desire to know subtypes, originally driven by the
threat of H5N1, led to the introduction of a single tube NAT that
delivers results for Influenza A, B and subtypes H1N1 and H3N2. The
weakness of the system is that samples are run in batches. In contrast,
an apparent strength of the Xpert flu assay is that it is run on a random
access instrument, allowing individual samples to be processed
without waiting for a batch.
Aim(s)/Objective(s):
To compare the Xpert assay with the assay in
routine use.
Method(s):
146 clinically requested samples were run on the in use
abTEST Flu 4 QPCR I kit and the Xpert Flu/RSV XC kit in June 2015.
Analysis assumed the in use assay to be correct as accumulated quality
assurance data showed excellent performance.
Results:
The Xpert system detected 31/35 H3N2, 34/35 H1N1, 30/35
Influenza B and 1/1 Influenza A without subtype: a total of 96/106
(90%). Five of the missed cases were due to a
‘
failed
’
run. The Xpert
system did not detect influenza in any of the 40 samples reported as
‘
not detected
’
by the in use assay.
Discussion and/or Conclusion(s):
Failed runs, linked with micro-
fluidics, are a significant weakness of the Xpert system. Detection
sensitivity is 10% inferior to our current system and the Xpert does not
deliver subtypes, nor can it be updated with amended primers/probes
at short notice. This can be crucial when dealing with outbreaks of new
subtypes requiring different management.
ID: 4858
Direct Lancefield grouping for the rapid identification of
streptococci in blood cultures
Melissa Baxter
1
, Aiden Plant
1
, Marina Morgan
1
.
1
Royal Devon and Exeter
NHS Foundation Trust
Background:
Standard laboratory identification of the Lancefield
group of Streptococci is by latex agglutination for the carbohydrate
antigen primarily from a colony following incubation and growth of
the organism. We sought to determine the reliability of undertak-
ing Streptococcal grouping directly from positive blood cultures,
proposing that the rapid identification may expedite appropriate
antimicrobial therapy, infection control precautions and public health
notification.
Methods:
Positive blood cultures (BacTalert 3D) with Streptococci
identified on Gram
’
s stain were prospectively subjected to direct
Lancefield A, B, C and G grouping (Prolab diagnostics) from December
2015 to June 2016. Comparison was made with final identification on
completion of culture, either by repeat Lancefield grouping or MALDI-
TOF analysis.
Results:
Of 53 blood cultures positive for
β
-haemolytic streptococci, 32
were subject to direct Lancefield grouping. 27 agglutinated with a
single group, whilst five remained agglutination negative. All of the 27
yielded the same organismon final culture (
p
= 0.02). The five negative
agglutinations cultured Group A streptococcus on four occasions and
Group G streptococcus on one.
Discussion:
We report 84.4% sensitivity in identification of Strepto-
coccal bacteraemia with direct Lancefield grouping. Importantly there
were no false positives. We propose direct grouping is a useful adjunct
in the interpretation of positive blood cultures pending confirmatory
testing. The utility of this with regard to clinical impact is yet to be
evaluated.
ID: 4866
Finding the needle in the haystack: Screening for CPOs at a large
London Teaching Hospital using Rosco Confirm Kits
Gemma Vanstone
1
, Damien Mack
2
, Indran Balakrishnan
3
.
1
HSL
Analytics,
2
The Royal Free London NHS Foundation Trust,
3
The Royal Free
NHS Foundation Trust
Background:
Several methods for the detection of carbapenemase
producing organisms (CPOs) are now available, with different
Abstracts of FIS/HIS 2016
–
Poster Presentations / Journal of Hospital Infection 94S1 (2016) S24
–
S134
S75