The Journal of Hospital Infection - Supplement 1

Results:

When applying the CDC definitions to the cases defined as

non-Trust-apportioned using PHE definitions, there was an increase of

22% in reportable cases from 168 to 205 cases. 31/37 additional Trust-

apportioned cases according to CDC surveillance definitions were due

to previous hospitalisation in the 4 weeks prior to the positive

specimen. Of the 84 cases defined as non-Trust-apportioned, 40 (48%)

had a previous overnight hospital stay in the 4 weeks prior to the

positive specimen, and 24 (29%) within 1 week prior to the positive

specimen.

Discussion and/or Conclusion(s):

A high proportion of

C. difficile

cases defined as

‘

non-Trust

’

have recent overnight hospitalization in

our hospitals, suggesting that these cases may be attributable to the

previous episode of care. PHE should consider including previous

hospitalisation in their epidemiological definitions.

ID: 4888

Health care-associated infections in a neurologic department:

epidemiology, antibiotic resistance and risk factors. Two years

study, Mansoura University, Egypt

Rawia Badr

, Rasha Hassan

, Enas Hammad

, Hassan Salama

Faculty

of Medicine, Mansoura University,

faculty of medicine

Background:

Health care-associated infections (HAIs) are represent-

ing a major problem in neurology department with a considerable

impact on mortality and morbidity rates.

Aim(s)/Objective(s):

This is a prospective study conducted to detect

HAIs according to CDC definition, main infection sites, common

microorganisms, mortality rate and risk factors such as age and length

of stay.

Method(s):

This study was carried out in a neurology department,

faculty of medicine, Mansoura University, Egypt, from January 2014 to

December 2015. Isolation, identification and antimicrobial suscepti-

bility were carried out. An outbreak of UTI due to Candida species

was detected and RFLP was done for typing and tracing source of

infection.

Results:

Elderly patients and those having prolonged length of

hospital stay were at higher risk of acquiring HAIs. Higher mortality

rate was associated with patients having HAIs, blood stream infection,

mechanical ventilation and low Glasgow Coma Scale. HAIs were

detected in 413 samples. The three most commonly reported organi-

sms were

Klebsiella pneumoniae

107 (26%), followed by

Candida spp

67 (16%) and

MRSA

65 (15.7%). Regarding site of infection, urinary

tract infection was the commonest site 185 (44.8%) followed

by respiratory tract infection 90 (21.8%), blood stream infection 73

(17.7%) and wound infection 65 (15.7%). Imipenem was the most

effective antibiotic against Gram negative bacteria. Vancomycin and

imipenem were found to be the most effective against Gram positive

bacteria.

Discussion and/or Conclusion(s):

HAIs prevention represents a real

challenge to minimize HAIs limiting mortality, antibiotic resistance

and outbreak emergence.

ID: 4890

The epidemiology of glycopeptide resistance in coagulase negative

staphylococci in the East Midlands

Sharon Koo, Nelun Perera.

University Hospitals of Leicester NHS Trust

Background:

Coagulase negative staphylococcus (CoNS) is one of the

most frequently isolated nosocomial pathogen in clinical microbiology

laboratories. There is increasing evidence that glycopeptide resistance,

particularly to teicoplanin is increasing in CoNS.

Aim(s)/Objective(s):

This is a retrospective analysis to evaluate

teicoplanin and vancomycin susceptibilities in CoNS isolates from

sterile site samples.

Method(s):

All CoNS isolates from sterile site samples submitted to our

laboratory between January 2013 and April 2015 were evaluated.

A total of 846 CoNS were isolated during that period. The majority of

the isolates were from blood cultures (49.5%), orthopaedic (bone, joint

and wound) samples (27.7%), peritoneal dialysis fluid (6.6%) and

sternum (5%).

Etest strips were used to determine teicoplanin MICs. Vancomycin

susceptibilities were initially performed using Etest strips. This was

changed to a commercial MIC testing system (Vitek 2) in October 2013.

The MIC values were interpreted according to the 2015 recommenda-

tions of the British Society for Antimicrobial Chemotherapy. Isolates

with MIC >4.0 mg/L were considered to be resistant.

Results:

799 samples were tested for vancomycin susceptibility. 2

samples were resistant (MIC 8.0mg/L) (0.25%). Both were isolated from

blood cultures. One of the samples was tested against teicoplanin and

was deemed susceptible (MIC 2.0mg/L).

Of the 247 samples that were tested for teicoplanin susceptibility, 9

CoNS isolates were resistant (3.6%). 2 had an MIC of 6.0mg/L and 7

samples had MIC of 8.0mg/L. 8 of the samples were vancomycin

susceptible.

Discussion and/or Conclusion(s):

The glycopeptide resistance in

CoNS isolates from our population reflected the pattern of that seen by

investigators in other countries.

ID: 4924

Carbapenemase screening programme at a London Teaching

Hospital with low rates of detection

Kerry Roulston

, Yvonne Carter

, Vicky Pang

, Gemma Vanstone

Indran Balakrishnan

, Robin Smith

, Damien Mack

Royal Free London

NHS Foundation Trust,

HSL Analytics

Background:

Public Health England advises carbapenemase screening

(CS) in high-risk patients. At the Royal Free Hospital, universal CS is

performed for Intensive Care Unit (ICU) patients and risk-based CS is

performed for private, renal, liver, haematology, infectious diseases

(ID), stroke and oncology patients.

Aim(s)/Objective(s):

To review the number of patients screened for

carbapenemase-producing organisms (CPO) and cases detected.

Method(s):

The laboratory information system was searched for all

CS samples between July 2013 and December 2015. CS was per-

formed using broth enrichment cultures containing imipenem discs

until December 2013 and chromID CARBA (Biomerieux) from January

2014 onwards. Suspected CPOs were confirmed by the reference

laboratory.

Results:

CS was performed for 25,787 samples from 7,520 patients,

with an average of three samples per patient (median = 2, IQR = 1

–

4).

The majority of samples were from ICU (52%), private (15%) and renal

(12%) patients. From the 7,520 patients screened, 44 (0.59%) new cases

of CPO were identified. Detection rates per patient screened were:

0.49% (ICU); 1.16% (private); 0.12% (renal); 2.17% (haematology); 0.55%

liver; 0% (stroke, oncology and ID); 0.82% (other specialties). Detection

rates per patient screened increased significantly from 12/3145 (0.38%)

in 2014 to 29/3685 (0.8%) in 2015 (OR = 2.07, 95%CI = 1.02

–

4.46,

p = 0.031).

Discussion and/or Conclusion(s):

The cost-effectiveness of imple-

menting CS is unknown. The yield from CS of high-risk patients in our

hospital is currently low, however detection rates have increased

significantly over the last two years. CS of high-risk patients is

important due to the increased risk of infection associated with high

mortality.

ID: 4939

Light surveillance of cranial surgery procedures in a tertiary

referral hospital using an integrated electronic approach

Teresa Inkster

, Ann Kerr

, Irene Aitken

, Susan Little

, Helen O

′

Neill

Queen Elizabeth University Hospital, Glasgow,

NHSGGC infection

control team,

NHSGGC Infection control,

NHSGGC infection control

Background:

NHS GGC IPCT have revisited SSI surveillance within

cranial surgery procedures using a real-time web application designed

to assist infection surveillance.

Neurosurgical SSI surveillance is not a mandatory component of the

HPS SSI surveillance programme.

Abstracts of FIS/HIS 2016

–

Poster Presentations / Journal of Hospital Infection 94S1 (2016) S24

–

S134

S123