Introduction
Healthcare associated infections represent an enormous challenge to patient
care in the Irish hospital and healthcare system. This issue is further
compounded by the increase in antimicrobial resistance in Ireland and
around the world. There is rising concern in Ireland about what are commonly
perceived as ;superbugs’. Methicillin resistant Staphylococcus
aureus (MRSA) is a topic that has dominated headlines in medical
journals for three decades. The public have genuine concerns about the ‘level’ of
MRSA in the institutions that deliver healthcare and the consequences
of the organism for patients. Infections with MRSA may require treatment
with parenteral second-line antimicrobials, possibly more expensive and
toxic, necessitating prolonged hospital stay [1]. Bacteraemia is an important
indicator of bloodstream infection. In many cases these infections are
healthcare-associated or hospital- acquired infections [2]. There is
a lack of consensus as to how to define hospital-acquired infections
and how to measure MRSA in Irish hospitals. Two measures often quoted
include the percentage of S. aureus isolates that are MRSA and
the incidence rate of S. aureus (and MRSA) expressed per 1000
bed days used. In contrast to the United Kingdom, in Ireland there is
little data published and readily available on S. aureus bacteraemia
and specifically MRSA bacteraemia. The results of surveillance of bacteraemia
due to S. aureus are seen as indicators of the extent of MRSA
in hospitals [3]. This study examines how representative two measures
for MRSA bacteraemia are in the Mid-Western Area – the percentage
of isolates MRSA and the incidence rate of MRSA bacteraemia expressed
per 1000 hospital bed days.
Methods and Materials
Cases of S. aureus bacteraemia were identified from blood cultures
investigated at the Microbiology Laboratory at the Mid-Western Regional
Hospital, Limerick. This laboratory provided the blood culture service
for all acute hospitals in the region in 2002-2004 (St. John’s
Hospital, since July 2002). Cases were exclusively blood cultures (including
those taken through intravascular devices) and did not include any fluids
(e.g., knee aspirates) that may have been cultured in the same manner.
No duplicates were included, first isolates were taken but this did not
exclude cases where MRSA bacteraemia followed an MSSA bacteraemia. Non-residents
of the HSE Mid-Western Area were not excluded from analyses. All patient
episodes were assigned to the hospital referring the sample to the laboratory.
The incidence rate of bacteraemia is the number of new bacteraemia episodes
expressed per in-patient bed days used over a period. These data were
extracted from the Laboratory Information System at the Mid-Western Regional
Hospital.
Data on inpatient bed days used were kindly provided by the HIPE Department
of the Mid-Western Regional Hospital and St. John’s Hospital, Limerick,
Ireland.
Abbreviations:
Hospital 1 – Mid-Western Regional Hospital Ennis, Co. Clare
Hospital 2 – Mid-Western Regional Hospital, Limerick City
Hospital 3 – Mid-Western Regional Hospital Nenagh, Co. Tipperary
Hospital 4 – Mid-Western Regional Maternity Hospital, Limerick
City
Hospital 5 – Mid-Western Regional Orthopaedic Hospital, Co. Limerick
Hospital 6 – St John’s Hospital, Limerick City
MSSA – Methicillin sensitive S. aureus
MRSA – Methicillin resistant S. aureus
Results
Data on 245 episodes of S. aureus bacteraemia were collected
for the three years January 2002 to December 2004. Table 1 shows the
number of MSSA and MRSA bacteraemia in each year by hospital and the
average percentage of MRSA in each centre over the three years. Overall
in the region, the percentage of bacteraemia caused by MRSA was 44% in
2002, 56% in 2003 and 48% in 2004. Table 2 shows the relative size and
activity in each of the hospitals in the area. Hospital 2, the largest
hospital, recorded the highest number of patients with S. aureus bacteraemia
in the area and the incidence rate has fallen each year unlike the percentage
which rose from 42% in 2002 to 54% in 2003. In isolation, the percentage
MRSA in Hospital 3 and Hospital 5 appear similar but the incidence rate
for both hospitals (Table 2) shows the level of MRSA to be very different.
The annual incidence rate in Hospital 3 is rising from 2002 to 2004.
The annual trend in %MRSA in hospitals from 2002 to 2004 are
opposite to the trend shown by the incidence rate. In the largest
hospital in the area the crude annual incidence rate of MRSA
bacteraemia has fallen 40% over the three years. Year on year
for the region there is no significant difference between annual
incidence rates or percentage MRSA at the 5% level.
The figure compares the percentage of patients with S. aureus bacteraemia
that were MRSA and the incidence rate (per 1000 bed days used)
of patients with bacteraemia due to MRSA and all S. aureus in
the Mid-Western Area from 2002 to 2004 by quarter. Aggregate
data merges opposite trends in MRSA from hospitals. While there
is little change in the incidence rate of MRSA bacteraemia in
the area, the rate of all S. aureus bacteraemia (including
MSSA) fell slightly in 2003 before rising again in late 2004.
Up to early 2003 the two measures appear very similar but then
the percentage of patients that yield MRSA isolates from bacteraemia
fluctuates in the area from quarter to quarter. In one nine-month
period the percentage of bacteraemia caused by MRSA rose from
39% to 71% and fell again to 40%. The difference between the
two measures is not as obvious at regional level and is probably
not statistically significant.
Discussion
In Ireland, bacteraemia caused by S. aureus became a statutorily
notifiable disease from January 2004. In the United Kingdom, the publication
of the incidence of S. aureus bacteraemia (including MRSA) has
been mandatory since April 2001. The publication of incidence rates for
Trusts and hospitals in England, Wales, Scotland and Northern Ireland
has proved informative.
Interventions designed to reduce antibiotic resistance and control
MRSA in hospitals are vital to minimise morbidity and mortality
due to infections caused by resistant organisms. Prudent antibiotic
usage, handwashing/hand disinfection, active screening, contact
precautions and environmental hygiene are key aspects to minimising
MRSA in hospitals. Consistent and comparative measures will be
required to evaluate such interventions. The strategy for control
of antimicrobial resistance in Ireland (SARI) outlines considerable
data on the surveillance, infrastructure and burden of disease
in Ireland as well as proposals for the implementation of future
strategies to control antimicrobial resistance [1]. It is suggested
that studies that link information on interventions to control
and prevent MRSA with resistance rates at the level of the hospital,
region or both, may increase our understanding of the nature
of the MRSA epidemic [4].
Data on S. aureus and MRSA bacteraemia are important
indicators of healthcare associated infections because they estimate
true infections (in the majority of cases) rather than colonisation.
The number of cases of MRSA bacteraemia is a small proportion
of all MRSA infections but this indicator is less likely to be
influenced by bias due to sampling variations between centres
(e.g. differential screening policies or sites tested) [5]. Regional
variations can occur if there are outbreaks of MRSA. Different
types of MRSA may be present in the regions but in Ireland there
are no regional data on S. aureus bacteraemia published
regularly.
An appropriate and consistent measure of the ‘level’ of MRSA
in a hospital is difficult to establish. Measures vary considerably by
time and place. It is claimed that MRSA has been endemic in a number
of large hospitals in Dublin since the 1980s [1]. Large tertiary healthcare
facilities may have higher rates of MRSA given that they provide more
specialist services (e.g., dialysis, oncology and intensive care) for
patients with complex medical needs. Laboratory methods of antimicrobial
susceptibility testing can vary between regions in Ireland and standardisation
may minimise this as a cause of variability. Admission and discharge
data are often not readily available through laboratory information and
communication technology, placing the burden of surveillance on infection
control staff. Complete case ascertainment is a crucial aspect of surveillance
and electronic data extraction of all cases highlighted issues for surveillance
in one hospital.
The percentage of isolates that are MRSA varies widely between different
time periods and depends on the type of hospital. The percentage alone
does not indicate the number of MRSA bacteraemias.
The incidence rate of MRSA does not show as much variability in time
and is better as a measure between hospitals of different size and casemix
compared to the percentage of bacteraemia that are MRSA. Peaks in the
incidence rate reflect increases in MRSA and not MSSA changes.
The situation is always more complex when percentages and rates are applied
to small numbers, such data must not be over-interpreted. Confidence
intervals are wide in such circumstances.
Percentage MRSA bacteraemia may be useful at a national ‘ecological’ level.
Data on 477 cases from over 20 Irish hospitals participating
in European Antimicrobial Resistance Surveillance System (EARSS)
showed the percentage of bacteraemia that are MRSA in Ireland
was 42% in 2003. This was much lower than the percentage in the
Mid-Western Area (56%). However, the incidence rate of MRSA bacteraemia
in the Mid-Western Area in 2003 is the same as the rate nationally
(0.16). Objectively, we cannot say that one measure is superior
to the other and we may be biased to a measure that shows a less
negative aspect. This incidence rate appears reliable and useful
for comparative purposes because it takes into account the difference
in the relative size of hospitals. However, services and casemix
are not equivalent, so unqualified comparisons between hospitals
are not as helpful. Indeed the attribution of a case of MRSA
bacteraemia to a particular healthcare facility is fraught with
problems – carriage of MRSA may have preceded admission
or infection may already be advanced on admission to one hospital
from another facility. Certain agreed time limits, consistently
applied, may make surveillance data more useful and some risk
adjustment of crude rates may facilitate comparisons in future.
At the very least high quality data on S. aureus bacteraemia
(including MRSA) should be published regularly by region if not
by hospital. It would be useful to determine the trends in the
percentage of S. aureus bacteraemia due to MRSA compared
to trends in the incidence rate per 1000 bed days used in other
European countries at national and hospital level as well.
Acknowledgements
The authors acknowledge the assistance of Trina Dooley
and Anne FitzGerald, HIPE Departments, Limerick and Breda Tuohy,
Surveillance Assistant, Department of Public Health. The commitment
and dedication of all involved in the multi-disciplinary Infection
Control Teams in the Mid-Western Area is acknowledged. The co-operation
and work of the staff of the Microbiology Department of the Mid-Western
Regional Hospital is greatly appreciated.
|