Post on 20-Jan-2020
1
Annual epidemiological commentary: Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C. difficile infections, up to and including financial year April 2018 to March 2019 11 July 2019
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
2
About Public Health England
Public Health England exists to protect and improve the nation’s health and wellbeing, and
reduce health inequalities. We do this through world-leading science, knowledge and
intelligence, advocacy, partnerships and the delivery of specialist public health services.
We are an executive agency of the Department of Health and Social Care, and a distinct
delivery organisation with operational autonomy. We provide government, local
government, the NHS, Parliament, industry and the public with evidence-based
professional, scientific and delivery expertise and support.
Public Health England
Wellington House
133-155 Waterloo Road
London SE1 8UG
Tel: 020 7654 8000
www.gov.uk/phe
Twitter: @PHE_uk
Facebook: www.facebook.com/PublicHealthEngland
Prepared by: Dimple Chudasama, Simon Thelwall, Olisaeloka Nsonwu, Graeme Rooney,
Sobia Wasti, Russell Hope
For queries relating to this document, please contact: mandatory.surveillance@phe.gov.uk
© Crown copyright 2019
You may re-use this information (excluding logos) free of charge in any format or medium,
under the terms of the Open Government Licence v3.0. To view this licence, visit OGL.
Where we have identified any third party copyright information you will need to obtain
permission from the copyright holders concerned.
Published July 2019
PHE publications
gateway number: GW-518
PHE supports the UN
Sustainable Development Goals
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
3
Contents
About Public Health England ........................................................................................................... 2
Executive summary ......................................................................................................................... 6
Bacteraemia caused by Gram-negative organisms ......................................................................... 8
E. coli bacteraemia ...................................................................................................................... 8
Total reports ............................................................................................................................. 8
Age and sex distribution ........................................................................................................... 9
Seasonal trends in E. coli ....................................................................................................... 14
Source of E. coli bacteraemia ................................................................................................ 15
Geographic distribution of E. coli bacteraemia ....................................................................... 16
Klebsiella spp. bacteraemia ....................................................................................................... 19
Total reports ........................................................................................................................... 19
Distribution of Klebsiella species ............................................................................................ 19
Age and sex distribution ......................................................................................................... 20
Seasonal trends in Klebsiella spp. ......................................................................................... 21
Source of Klebsiella spp. bacteraemia ................................................................................... 22
Geographic distribution of Klebsiella spp. bacteraemia.......................................................... 23
Pseudomonas aeruginosa bacteraemia .................................................................................... 25
Total reports ........................................................................................................................... 25
Age and sex distribution ......................................................................................................... 25
Seasonal trends in P. aeruginosa .......................................................................................... 27
Source of P. aeruginosa bacteraemia .................................................................................... 28
Geographic distribution of P. aeruginosa bacteraemia .......................................................... 30
Discussion ................................................................................................................................. 31
Epidemiological analysis of Staphylococcus aureus bacteraemia ................................................. 34
Meticillin-resistant Staphylococcus aureus bacteraemia ........................................................... 34
Total reports ........................................................................................................................... 34
Trust-assigned reports ........................................................................................................... 36
Age and sex distribution ......................................................................................................... 37
Source of MRSA bacteraemia ................................................................................................ 41
Geographic distribution of MRSA bacteraemia ...................................................................... 42
Discussion .............................................................................................................................. 43
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
4
Meticillin-susceptible Staphylococcus aureus bacteraemia ....................................................... 46
Total reports ........................................................................................................................... 46
Hospital-onset reports ............................................................................................................ 47
Age and sex distribution ......................................................................................................... 47
Source of MSSA bacteraemia ................................................................................................ 51
Geographic distribution of MSSA bacteraemia ...................................................................... 51
Discussion .............................................................................................................................. 53
Epidemiological analysis of Clostridioides difficile Infection .......................................................... 55
Hospital-onset reports ............................................................................................................ 56
Prior trust exposure ................................................................................................................ 56
Seasonal trends in CDI .......................................................................................................... 58
Age and sex distribution ......................................................................................................... 59
Trends in age and sex............................................................................................................ 60
Geographic distribution of CDI ............................................................................................... 62
Discussion .............................................................................................................................. 63
Appendix ....................................................................................................................................... 66
Background to the mandatory surveillance of MRSA and MSSA bacteraemia .......................... 66
Mandatory surveillance of C. difficile infection ........................................................................... 67
Mandatory surveillance of Gram-negative bacteraemia ............................................................ 67
E. coli bacteraemia................................................................................................................. 67
Klebsiella and P. aeruginosa bacteraemia ............................................................................. 67
A note on terminology ................................................................................................................ 68
Use of mandatory surveillance statistics .................................................................................... 68
Data included in the 2018/19 Annual Epidemiological Commentary (AEC) ............................... 68
E. coli bacteraemia................................................................................................................. 69
Klebsiella spp. bacteraemia ................................................................................................... 69
P. aeruginosa bacteraemia .................................................................................................... 69
MRSA bacteraemia ................................................................................................................ 70
MSSA bacteraemia ................................................................................................................ 70
C. difficile infections ............................................................................................................... 70
Commentary .......................................................................................................................... 70
Alternative presentations of the annual data .......................................................................... 71
Glossary ........................................................................................................................................ 72
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
5
Methods ..................................................................................................................................... 74
Inclusion criteria for reporting to the surveillance system ....................................................... 74
Methods of reporting data on the HCAI Data Capture System (DCS) .................................... 75
Deadline for entering data ...................................................................................................... 75
CCG attribution process ......................................................................................................... 76
Overview of CCG attribution .................................................................................................. 76
Algorithms for apportioning cases .......................................................................................... 76
Assignment of MRSA cases through the Post Infection Review ............................................ 78
Analysis of data ......................................................................................................................... 80
Time to onset calculations ...................................................................................................... 80
Denominator data .................................................................................................................. 81
Rate calculations .................................................................................................................... 82
Healthcare associated infections in Wales, Scotland and Northern Ireland ............................... 83
References .................................................................................................................................... 85
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
6
Executive summary
Overall rates of MRSA bacteraemia and C. difficile infection have continued to see year on
year reductions. These declines in MRSA bacteraemia and CDI rate, were in both the
community and hospital settings, testament to the success of the interventions introduced
to combat these infections. In contrast, the E. coli and MSSA rates have increased, with
the most prominent rises seen in the community-onset cases. MSSA showed a small
increase in hospital-onset cases but for E. coli this has been relatively static.
Therefore, over time the community-onset component has accounted for an increasing
proportion of cases, this is even true for the MRSA and CDI, as although rates have
reduced in both settings the reduction was steeper in the hospital setting. This highlights
the need to increase efforts to prevent cases in the community. Nevertheless, hospital-
onset infections remain important, representing thousands of cases a year.
While many of the infections were community-onset it is estimated that a large proportion
(up to 50% in the case of E. coli) have had recent healthcare interactions. Therefore, in
order to reduce infection rates further, control efforts in the hospital setting must be
maintained or strengthened, while increasing focus on interventions in the community and
the interface between hospital and community infection control teams improved.
The highest rates across all the pathogens was among those eighty-five and older
especially in males. However, there was a substantial number of cases (~80%) in the
younger age groups, and therefore prevention strategies should be focused across all
patient ages groups if we wish to minimise preventable infections and see a significant
reduction in cases.
Highlights
Rates of Escherichia coli bacteraemia have continued to rise, in 2018/19 the rate had
increased to 77.7 cases per 100,000 population (from 73.7 per 100,000 in the previous
year). The smallest annual increase in E. coli bacteraemia, since the mandatory
surveillance started in 2011, occurred between 2016/17 and 2017/18, of 2.5%. The
increase in the latest financial year (2018/19) was approximately 5%, similar to that
observed in previous years.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
7
After E. coli, Klebsiella and Pseudomonas spp. are the most common causative organisms
in Gram-negative bacteraemia but have substantially lower rates of bacteraemia than E.
coli. Rates of Klebsiella spp. increased, from 17.5 in 2017/18 to 19.1 in 2018/19. Rates of
Pseudomonas aeruginosa bacteraemia declined slightly, from 7.7 in 2017/18 to 7.5 in
2018/19. Urinary tract infection (UTI) remained the most important primary focus of E. coli,
Klebsiella spp. and Pseudomonas aeruginosa bacteraemia, causing 49.1%, 33.5% and
29.7% of cases respectively.
Rates of Meticillin-Sensitive Staphylococcus aureus (MSSA) bacteraemia continued to
increase moderately from 2011/12 when the surveillance was introduced. The rate
increased from 21.5 cases per 100,000 population in 2017/18 to 21.7 per 100,000
population in 2018/19.
In contrast those of MSSA, rates of Meticillin-Resistant Staphylococcus aureus (MRSA)
bacteraemia and Clostridioides difficile infection (CDI) have remained low in comparison to
rates at their peak, at 1.5 and 22.1 per 100,000 population, respectively in 2018/19.
The high rates of Gram-negative bacteraemia’s and diverse nature of the underlying
causes of these infections compared to MRSA and CDI present a significant challenge to
achieving the ambition to halve healthcare-associated Gram-negative bacteraemia by
2023/2024: HM Government 2018
This report constitutes a full and descriptive analysis of the data from the mandatory
surveillance of bacteraemia and CDI programmes in England. Detailed information on
each pathogen can be found in the individual sections of the report.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
8
Bacteraemia caused by Gram-negative
organisms
E. coli bacteraemia
Total reports
A total of 43,242 cases of E. coli bacteraemia were reported by NHS Trusts in England
between 1 April 2018 and 31 March 2019. Of the 43,242 E. coli cases, 7,632 (17.6%) were
hospital-onset (Table 1). The total number of cases reported in 2018/19 is an increase of
5.2% from 2017/18 (n = 41,091), and an increase of 33.8% from 2012/13 (n = 32,309).
Figure 1 shows the trends in the rates of E. coli cases from 2012/13 to 2018/19. The rate
of E. coli cases per 100,000 population has risen from 60.4 in 2012/13 to 77.7 in 2018/19.
Table 1: E. coli counts and rates by financial year, England: 2012/13 to 2018/19
Financial year
Mid-year population estimate*
All reported
cases
Rate (all reported cases
per 100,000 population)
Total bed days**
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed days)
2012/13 53,475,358 32,309 60.4 34,439,455 7,552 21.9 2013/14 53,976,973 34,286 63.5 34,327,781 7,558 22.0 2014/15 54,432,437 35,816 65.8 34,797,208 7,380 21.2 2015/16 55,018,884 38,309 69.6 34,576,351 7,742 22.4 2016/17 55,240,933 40,647 73.6 34,976,071 7,878 22.5 2017/18 55,619,430 41,091 73.9 34,749,047 7,680 22.1 2018/19 55,619,430 43,242 77.7 34,525,599 7,632 22.1 * Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy. ** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
9
Figure 1: Trends in the rate of E. coli bacteraemia in England, 2012/13 to 2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the
2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Age and sex distribution
For all age and sex analyses, cases in which the sex was missing or given as unknown
were excluded. In 2012/13, the sex was unknown for 815 cases (2.5%), while in 2018/19,
the sex was unknown for 24 cases (0.1%).
Figure 2 compares the age and sex structure of E. coli bacteraemia cases in England
between 2012/13 and 2018/19. For both years, E. coli bacteraemia rates were highest
among patients ≥ 85 years of age. In 2012/13, the rate among males was 718.2 cases per
100,000 population and 482.1 among females. In 2018/19, the rate among males was
922.8 per 100,000 population and 623.3 among females. , however it must be noted that
the population for ≥ 85 is much lower than the other age groups, particularly for males. It
should also be noted that the ≥ 85 years group accounts for 22.9% of all E. coli
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
10
bacteraemia’s, whereas the rest is made up by the other age groups, with the 75-84 year
olds with the highest proportion of E. coli bacteraemias at 27.9% and the 1-14 age group
with the lowest of 0.5%.
Figure 2: Age and sex structure of E. coli rates per 100,000 population, England*
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
In general, the rates of E. coli bacteraemia are greater among males than among females,
and particularly so among older age groups. However, there is a notable exception to this.
In young adulthood (15-44) the rate of E. coli bacteraemia is considerably higher among
females (24.4 per 100,000 population) than among males (7.7 per 100,000 population),
giving a male-to-female rate ratio of 3 (95% CI: 2.9-3.4).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
11
Although the rates of E. coli bacteraemias have increased over time, the percentage gap
in rates between sexes within an age group have not changed appreciably (Table 2 & 3).
To compare disease incidence between two groups, one can calculate the ratio of rates
between the two groups to produce a rate ratio. The rate ratio for E. coli between male and
females in 2012/13 was 0.9 (95% CI: 0.9-1). In comparison, the rate ratio in 2018/19 was 1
(95% CI: 0.9-1).
Table 2: E. coli counts and rates by age group and sex, England: 2012/13
Age group (years) Population n cases Rate, per 100,000
Male Female Male Female Male Female
<1 353,705.1 336,393.0 306 195 86.5 58.0 1-14 4,503,696.0 4,294,577.8 81 136 1.8 3.2 15-44 10,724,195.2 10,658,627.9 712 2,087 6.6 19.6 45-64 6,669,558.9 6,835,423.2 2,944 2,878 44.1 42.1 65-74 2,347,921.0 2,530,753.9 3,520 2,889 149.9 114.2 75-84 1,320,659.4 1,677,572.3 4,577 4,311 346.6 257.0 ≥ 85 408,947.6 813,326.7 2,937 3,921 718.2 482.1
Table 3: E. coli counts and rates by age group and sex, England: 2018/19
Age group (years) Population n cases Rate, per 100,000
Male Female Male Female Male Female
<1 335,035 318,432 420 283 125.4 88.9 1-14 4,813,341 4,581,557 110 119 2.3 2.6 15-44 10,749,133 10,572,163 829 2,578 7.7 24.4 45-64 7,014,072 7,205,186 3,934 3,905 56.1 54.2 65-74 2,646,091 2,849,090 4,930 4,170 186.3 146.4 75-84 1,435,728 1,747,546 6,236 5,816 434.3 332.8 ≥ 85 487,653 864,403 4,500 5,388 922.8 623.3
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
Figure 3 shows the trends in rate of E. coli bacteraemia by age group and by sex. For both
sexes the rate of E. coli bacteraemia has increased most dramatically amongst the ≥ 85
age group. However, there have also been increases in rate in the <1 age group.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
12
Figure 3: Trend in age and sex structure of E. coli cases and rate per 100,000
population, England, 2012/13 to 2018/19
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Percentage changes in the age & sex-specific rates of E. coli bacteraemia relative to the
first year of surveillance are shown in Figure 4. Except for females between the ages of 1-
14, rates of E. coli bacteraemia have increased since the start of surveillance.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
13
The proportionally largest increases in rate are observed in those less than 1 years old for
both males and females.
In contrast to the females, rates among males 1-14 years old have shown increases in rate
almost as great as the <1s.
Figure 4: Change in age- and sex-specific rates since 2011/12
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
14
Seasonal trends in E. coli
Seasonality trend analysis, seen in Figure 5, show a clear seasonality for E. coli
bacteraemia, with the highest percentages reported in the July to September quarter (Q2).
This appears to be the same for both hospital-onset and community-onset cases, however
it is more consistent in community-onset cases. More variable is seen in hospital-onset
cases compared to community-onset, particularly from 2014/15 to 2018/119.
Figure 5: Trends in the seasonality of E. coli bacteraemia, 2011/12 to 2018/19
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
15
Source of E. coli bacteraemia
The provision of data on the most likely primary focus of E. coli infection information is
voluntary. In 2012/13 a total of 27,610 (85.5%) of records had entries for the source of
bacteraemia. By 2018/19 a total of 27,747 (64.2%) had entries for the most likely primary
focus of the bacteraemia (Table 4).
Urinary tract infection (UTI) has consistently been the most frequent primary focus for E.
coli bacteraemia cases. In 2012/13, 48.9% of cases had a most likely primary focus of UTI,
by 2018/19 it was 49.1%. Conversely, the percentage of records for which the primary
focus was reported as unknown has decreased from 20.2% in 2012/13 to 14.3% in
2018/19.
Table 4: E. coli bacteraemia counts and rates by primary focus of bacteraemia,
England: 2018/19
Financial
year GI * n (%) HB** n (%) UTI n (%) Respiratory tract n (%)
Other*** n (%)
Unknown n (%) Total n (%)
2012/13 1,782 (6.5) 3,756 (13.6) 13,501 (48.9) 1,050 (3.8) 1,936 (7.0) 5,585 (20.2) 27,610 (100.0) 2013/14 1,711 (6.0) 3,855 (13.6) 13,393 (47.3) 1,016 (3.6) 1,873 (6.6) 6,452 (22.8) 28,300 (100.0) 2014/15 1,640 (5.7) 3,818 (13.3) 13,087 (45.6) 1,099 (3.8) 1,851 (6.4) 7,233 (25.2) 28,728 (100.0) 2015/16 1,491 (5.6) 3,556 (13.4) 12,219 (46.2) 1,068 (4.0) 1,703 (6.4) 6,408 (24.2) 26,445 (100.0) 2016/17 1,237 (5.4) 3,277 (14.4) 10,724 (47.2) 1,027 (4.5) 1,553 (6.8) 4,907 (21.6) 22,725 (100.0) 2017/18 1,718 (6.7) 4,034 (15.8) 12,564 (49.1) 1,575 (6.1) 1,756 (6.9) 3,963 (15.5) 25,610 (100.0) 2018/19 1,934 (7.0) 4,604 (16.6) 13,633 (49.1) 1,674 (6.0) 1,929 (7.0) 3,973 (14.3) 27,747 (100.0)
* Gastrointestinal not including hepatobilary
** Hepatobilary
****Other” includes the following options in the HCAI DCS; bone and joint, central nervous system, genital tract
(including prostate), indwelling intravascular device, other, respiratory tract, skin/soft tissue, no clinical signs of bacteraemia.
When the analysis of the most likely primary focus is restricted to cases where date of
admission was available, some differences in distribution are noted (Figure 6). The most
common primary focus is UTIs, accounting for just under half of all cases with an
admission date.
The primary focus of E. coli bacteraemia also varies according to time to onset. UTI were
the primary focus for over half (51.1%) of patients where the time to onset was < 2 days
from admission. In contrast, in cases where the time to onset was two to six days from
admission, only 33.2% of cases had a primary focus that was UTI.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
16
Figure 6: Distribution of primary focus E. coli bacteraemia, by time to onset,
England 2018/19
Geographic distribution of E. coli bacteraemia
Some geographical variation in rates of E. coli is noted (Figure 7). Areas with high rates
tended to be towards the North of England (NHS England North (Cumbria and North East)
Q74: 104.7 E. coli bacteraemia per 100,000 population, NHS England North (Cheshire and
Merseyside) Q75: 91.4 and NHS England Midlands and East (North Midlands) Q76: 89.9).
In contrast, lower rates were observed in the south of England, with the lowest rate in
2018/19 occurring in NHS England London (Q71) with a rate of 64.6 bacteraemia per
100,000 population.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
17
Figure 7: Geographic distribution of E. coli rates per 100,000 population, England
2018/19*
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
Table 5 gives NHS local office-specific rates. All local offices have experienced increases
in E. coli incidence over time. The largest increase in incidence was observed in NHS
England South West (South West South) Q85, which saw the incidence rise from 57.2 per
100,000 population in 2012/13 to 88.9 in 2018/19. However, between 2017/18 and
2018/19, drops in rate were observed for two local offices: Q88 NHS England South East
(Kent, Surrey and Sussex) from 77.2 to 75.1 and Q87 NHS England South East
(Hampshire, Isle of Wight and Thames Valley) from 71.1 to 70.2 bacteraemia per 100,000
population.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
18
Table 5: E. coli rates by NHS local office, England per 100,000 population: 2012/13 to
2018/19*
Code NHS local office
20
12
/13
20
13
/14
20
14
/15
20
15
/16
20
16
/17
20
17
/18
20
18
/19
Q78 NHS England Midlands and East (Central Midlands) 56.0 58.8 61.6 64.7 65.3 66.0 67.5 Q79 NHS England Midlands and East (East) 52.1 55.5 57.6 62.5 68.4 69.1 72.3 Q76 NHS England Midlands and East (North Midlands) 72.5 79.3 82.6 83.1 86.4 81.5 89.9 Q77 NHS England Midlands and East (West Midlands) 60.7 62.5 66.3 67.3 68.6 67.6 74.9 Q75 NHS England North (Cheshire and Merseyside) 71.9 86.2 76.7 81.4 89.1 91.4 91.4 Q72 NHS England North (Yorkshire and Humber) 71.5 73.6 76.1 80.1 84.9 82.4 89.0 Q83 NHS England North (Greater Manchester) 66.0 66.5 63.4 68.7 72.6 70.3 79.0 Q74 NHS England North (Cumbria and North East) 77.0 74.0 82.9 88.1 92.5 96.6 104.7 Q84 NHS England North (Lancashire and South Cumbria) 67.2 70.4 74.9 78.4 82.5 85.5 87.3 Q87 NHS England South East (Hampshire, Isle of Wight and
Thames Valley) 48.8 51.8 56.6 60.0 65.8 71.1 70.2
Q88 NHS England South East (Kent, Surrey and Sussex) 57.7 62.0 62.3 68.2 75.9 77.2 75.1 Q86 NHS England South West (South West North) 50.6 52.4 57.5 59.1 58.4 57.0 65.1 Q85 NHS England South West (South West South) 57.2 60.2 66.6 72.6 79.6 81.1 88.9 Q71 NHS England London 53.2 55.7 56.6 60.9 63.0 62.6 64.6 National 60.4 63.5 65.8 69.6 73.6 73.8 77.7
* 2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
19
Klebsiella spp. bacteraemia
Total reports
A total of 10,638 cases of Klebsiella spp. bacteraemia were reported by NHS Trusts in
England between 1 April 2018 and 31 March 2019. Of the 10,638 Klebsiella spp. cases,
3,182 (29.9%) were hospital-onset (Table 6).
Table 6: Klebsiella spp. bacteraemia counts and rates by financial year, England:
2018/19
Financial year
Mid-year population estimate
All reported
cases
Rate (all reported cases
per 100,000 population)
Total bed days**
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed days)
2017/18 55,619,430 9,744 17.5 34,749,047 2,902 8.4 2018/19 55,619,430 10,638 19.1 34,525,599 3,182 9.2
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18were
used as a proxy. * Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Distribution of Klebsiella species
Klebsiella pneumoniae was the most frequently reported Klebsiella species. Species
distribution was similar between hospital-onset cases and community-onset cases (Table 7).
Table 7: Counts and percentages of Klebisella species, England 2018/19
Species All cases Hospital-onset
cases Community-onset
cases
K. pneumoniae 7,850 (73.8%) 2,279 (71.6%) 5,571 (74.7%) K. oxytoca 1,743 (16.4%) 532 (16.7%) 1,211 (16.2%) K. aerogenes 131 (1.2%) 64 (2.0%) 67 (0.9%) Other named species 396 (3.7%) 142 (4.5%) 254 (3.4%) Not speciated 518 (4.9%) 165 (5.2%) 353 (4.7%)
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
20
Age and sex distribution
Rates of Klebseilla spp. bacteraemia were higher among males than among females. This
was particularly true among age groups greater than 65 years old (Figure 8). A further
breakdown of age distribution can be seen in Table 8. The rate ratio between males and
females in the ≥ 85 age group was 3.1 (95% CI: 2.8-3.3).
Figure 8: Age and sex structure of Klebsiella spp. rates per 100,000 population,
England
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
21
Table 8: Klebsiella spp. counts and rates by age group and sex, England: 2018/19
Age group (years) Population n cases Rate, per 100,000
Male Female Male Female Male Female
<1 335,035 318,432 109 84 32.5 26.4 1-14 4,813,341 4,581,557 92 48 1.9 1.0 15-44 10,749,133 10,572,163 372 425 3.5 4.0 45-64 7,014,072 7,205,186 1,391 950 19.8 13.2 65-74 2,646,091 2,849,090 1,624 920 61.4 32.3 75-84 1,435,728 1,747,546 1,798 894 125.2 51.2 ≥ 85 487,653 864,403 1,219 708 250.0 81.9
Seasonal trends in Klebsiella spp.
Seasonality analysis of Klebsiella spp. bacteraemia (Figure 9), show a higher percentage
of cases in the quarter July to September (Q2), similar to that seen in E. coli bacteremia. In
contrast to E. coli bacteraemia, where there is little difference the proportional seasonal
shifts in hospital-onset cases, for the Klebseilla spp. infections the effect was more
pronounced.
Figure 9: Trends in the seasonality of Klebsiella spp. bacteraemia, 2017/18 to
2018/19
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
22
Source of Klebsiella spp. bacteraemia
The most frequent primary focus of Klebsiella spp. bacteraemia was UTI, constituting
33.5% of cases with a reported primary focus of infection. The primary focus of infection
was not reported in 46.8% of cases (Table 9).
Table 9: Klebsiella spp. bacteraemia counts and rates by primary focus of
bacteraemia, England: 2018/19
Financial
year UTI GI* HB** Respiratory
tract Others Unknown Total
n % n % n % n % n % n %
2017/18 1677 32.9 374 7.3 1031 20.2 469 9.2 601 11.8 944 18.5 5096 2018/19 1893 33.5 483 8.5 1083 19.1 559 9.9 747 13.2 891 15.8 5656
* Gastrointestinal not including hepatobilary ** Hepatobilary ****Other” includes the following options HCAI DCS; bone and joint, central nervous system, genital tract (including prostate), indwelling intravascular device, other, respiratory tract, skin/soft tissue, no clinical signs of bacteraemia.
Amongst patients with an admission date and a time to onset of less than 2 days, UTI
forms a major primary focus of bacteraemia (37.8%). However, in case with an onset ≥ 7
days after admission, this declined to 20.1%
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
23
Figure 10: Distribution of primary focus of Klebsiella spp. bacteraemia, by time to
onset, England 2018/19
Geographic distribution of Klebsiella spp. bacteraemia
Generally, rates of Klebsiella spp. bacteraemia were higher in the North of England than in
the South. (Figure 11). However, with the exceptions of: Q83 NHS England North (Greater
Manchester) had a rate of 18.7 cases per 100,000 population and Q84 NHS England
North (Lancashire and South Cumbria) had a rate of 19.5, both showing slightly lower
rates. Also, two southerly offices showing higher rates, Q85 NHS England South West
(South West South) and Q88 NHS England South East (Kent, Surrey and Sussex) with a
rate of 21.3 and 20.2, respectively (Table 10).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
24
Figure 11: Geographic distribution of Klebsiella spp. Rates per 100,000 population,
England 2018/19*
* Figure 9: Geographic distribution of Klebsiella spp. rates per 100,000 population, England 2018/19
Table 10: Klebsiella spp. rates by NHS local office, England: 2018/19
Code NHS local office 2017/18 2018/19
Q76 NHS England Midlands and East (North Midlands) 19.6 21.0 Q77 NHS England Midlands and East (West Midlands) 14.9 16.5 Q78 NHS England Midlands and East (Central Midlands) 16.1 17.7 Q79 NHS England Midlands and East (East) 16.4 17.7 Q72 NHS England North (Yorkshire and Humber) 17.8 20.5 Q74 NHS England North (Cumbria and North East) 24.3 25.6 Q75 NHS England North (Cheshire and Merseyside) 20.1 21.9 Q83 NHS England North (Greater Manchester) 16.8 18.7 Q84 NHS England North (Lancashire and South Cumbria) 18.6 19.5 Q87 NHS England South East (Hampshire, Isle of Wight and Thames Valley) 15.6 16.8 Q88 NHS England South East (Kent, Surrey and Sussex) 19.4 20.2 Q85 NHS England South West (South West South) 18.0 21.3 Q86 NHS England South West (South West North) 15.1 16.4 Q71 NHS England London 16.6 17.9 National 17.5 19.1
* 2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
25
Pseudomonas aeruginosa bacteraemia
Total reports
A total of 4,185 cases of P. aeruginosa bacteraemia were reported by NHS Trusts in
England between 1 April 2018 and 31 March 2019. Of the 4,185 P. aeruginosa cases,
1,517 (36.2%) were hospital-onset (Table 11).
Table 11: P. aeruginosa bacteraemia counts and rates by financial year, England:
2018/19
Financial
year Mid-year
population estimate*
All reported cases
Rate (all reported cases
per 100,000 population)
Total bed days**
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed days)
2017/18 55,619,430 4,299 7.7 34,749,047 1,624 4.7 2018/19 55,619,430 4,185 7.5 34,525,599 1,517 4.4
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy. ** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Age and sex distribution
The age and sex distribution of P. aeruginosa bacteraemia were higher among older
people and particularly among older men Figure 12.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
26
Figure 12: Age and sex structure of P. aeruginosa rates per 100,000 population,
England 2018/19*
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
The rate ratio between men and women in the ≥ 85 age group was 3.4 (95% CI: 2.9-3.9). A further breakdown of age distribution can be seen in Table 12
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
27
Table 12: P. aeruginosa counts and rates by age group and sex, England: 2018/19
Age group (years) Population* n cases Rate, per 100,000 Male Female Male Female Male Female
<1 335,035 318,432 34 25 10.1 7.9 1-14 4,813,341 4,581,557 65 55 1.4 1.2 15-44 10,749,133 10,572,163 183 153 1.7 1.4 45-64 7,014,072 7,205,186 525 370 7.5 5.1 65-74 2,646,091 2,849,090 631 354 23.8 12.4 75-84 1,435,728 1,747,546 724 327 50.4 18.7 ≥ 85 487,653 864,403 483 254 99.0 29.4
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place
Seasonal trends in P. aeruginosa
Seasonality analysis of P. aeruginosa bacteraemia (Figure 13), show a higher percentage
of cases in the quarter July to September (Q2) for community-onset cases in 2017/18,
whereas although an increase is seen around the same time in hospital-onset cases, a
peak is seen in the October - December 2017 (Q3) month. However, there is a shift in the
2018/19, where both peaks for hospital-onset and community-onset are seen in Q2,
similarly to the trend seen in E. coli and Klebsiella spp. bacteraemia.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
28
Figure 13: Trends in the seasonality of P. aeruginosa bacteraemia, 2017/18 to
2018/19
Source of P. aeruginosa bacteraemia
The most frequent primary focus of bacteraemia for P. aeruginosa was UTI, constituting
29.7% of cases with a reported primary focus of infection (Table 13). However, the primary
focus was not reported for half (49.6%) of the cases.
Table 13: P. aeruginosa counts and rates by primary focus of bacteraemia, England:
2018/19
Financial year
GI* HB** UTI Respiratory tract
Others Unknown Total
n % n % n % n % n % n %
2017/18 125 6 102 5 637 30 300 14 491 23 471 22 2,126 2018/19 152 7 100 5 626 30 275 13 545 26 413 20 2,111
Among patients with an available admission date, those with a time to onset of less than 2
days, UTI was the most commonly reported primary focus of P. aeruginosa bacteraemia
(34.0%). However, for patients where the time to onset was ≥ 7 days from admission, the
percentage of infections with a UTI primary focus declined to 22.6% (Figure 14).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
29
Figure 14: Distribution of primary focus of P. aeruginosa bacteraemia by time to
onset, England 2018/19
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
30
Geographic distribution of P. aeruginosa bacteraemia
Figure 15: Geographic distribution of P. aeruginosa rates per 100,000 population,
England 2018/19*
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Figure 15 and Table 14 gives NHS local office-specific rates. Broadly speaking there is no
north-south divide in the rates of P. aeruginosa, unlike that observed for the other Gram-
negative bacteraemias. The highest rate was observed in Q88 NHS England South East
(Kent, Surrey and Sussex) (9.7 cases per 100,000 population) and Q76 NHS England
Midlands and East (North Midlands) (8.3 cases per 100,000 population) (Table 14).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
31
Table 14: P. aeruginosa bacteraemia rates by NHS local office, England: 2018/19*
Code NHS local office 2017/18 2018/19
Q76 NHS England Midlands and East (North Midlands) 8.9 8.3 Q77 NHS England Midlands and East (West Midlands) 7.0 7.6 Q78 NHS England Midlands and East (Central Midlands) 7.1 7.6 Q79 NHS England Midlands and East (East) 7.4 7.6 Q72 NHS England North (Yorkshire and Humber) 7.4 7.1 Q74 NHS England North (Cumbria and North East) 8.0 8.0 Q75 NHS England North (Cheshire and Merseyside) 8.4 7.2 Q83 NHS England North (Greater Manchester) 5.8 4.8 Q84 NHS England North (Lancashire and South Cumbria) 6.7 6.3 Q87 NHS England South East (Hampshire, Isle of Wight and Thames Valley) 8.2 7.5 Q88 NHS England South East (Kent, Surrey and Sussex) 9.0 9.7 Q85 NHS England South West (South West South) 7.0 5.7 Q86 NHS England South West (South West North) 7.8 6.2 Q71 NHS England London 8.1 8.2 National 7.7 7.5
* 2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Discussion
This report provides the first two years of data for the expanded Gram-negative
surveillance and is the second annual commentary in which onset status has been
reported for E. coli bacteraemia. Data on bacteraemia caused by E. coli, Klebsiella spp.
and P. aeruginosa showed both similarities and differences in the epidemiology of these
infections.
E. coli bacteraemia had the highest rate of all the bacteraemias caused by Gram-negative
organisms, causing 77.7 cases per 100,000 population in 2018/19. Klebsiella spp. and P.
aeruginosa bacteraemias had much lower rates of infection (19.1 and 7.5 respectively).
Differences were also observed in the proportion of cases that were hospital-onset (E. coli
- 18%, Klebsiella spp. - 30% and P. aeruginosa - 36%).
While the UTIs remains a major source of bacteraemia for all Gram-negative infections (E.
coli - 49%, Klebsiella spp. - 33% and P. aeruginosa – 30%), the respiratory tract is the
focus for 13% of P. aeruginosa infections but only 6% in E. coli and 10% in Klebsiella spp.
infections.
The collection of information on primary focus of bacteraemia changed in April 2018/19
with the introduction of mandatory surveillance of Klebsiella spp. and P. aeruginosa. For
this reason, the data on primary focus of infection for E. coli bacteraemia presented in this
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
32
report may not match those in previous reports. Despite the differences in epidemiology
noted above, there are similarities in the age and sex distribution of all three bacteraemias.
Rates of infection are higher among older age groups (65-74, 75-84 and ≥ 85) and those
under one year old for all Gram-negative bacteraemias. Despite this similarity, the rate of
bacteraemia caused by either Klebsiella spp. or P. aeruginosa is much greater among
men than among women. This contrasts with rates of bacteraemia caused by E. coli where
the rate of bacteraemia is similar between men and women. It should, however, be noted
that rates of E. coli bacteraemia are still higher among males than females for most age
groups.
Seasonality trends appeared similar across all three bacteraemias, particularly in regard to
the highest infection rates being seen in Q2 of the financial year. Community-onset cases
in general, appeared to see the larget and most consistent seasonal shifts in the
proportion of infections. However, seasonal changes were still clearly evident in the
hospital-onset cases especially in P. aeruginosa.
Rates of E. coli and Klebsiella spp. showed substantial inter-regional variation and similar
geographic distributions in England with higher rates in the North of England and generally
lower rates in the South of England. In contrast, rates of P. aeruginosa bacteraemia are
more evenly spread across England, except for relatively higher rates in the South East of
England. Interestingly, for all infections, the Manchester region has lower rates of infection
compared to its surrounding areas.
Long term trends are available for E. coli and reveal increasing rates of incidence since the
first full financial year of mandatory surveillance in 2012/13.
The rate of increase in counts of E. coli bacteraemia has been consistent between
2012/13 and 2016/17, averaging 5% per year. However, between 2016/17 and 2017/18,
the count of cases only increased by 1%, the first time it had been that low in the 6 years
of the mandatory surveillance of E. coli bacteraemia. However, between 2017/18 and
2018/19 the annual increase in counts of cases rate returned to 5%. Despite the increase
in all-case incidence, the incidence of hospital-onset cases has remained approximately
stable, suggesting that much of the increase in incidence is occurring in community-onset
cases. However, as community-onset cases include patients with recent hospital
admissions and other healthcare interactions, increases of cases within this category
should not be seen as coterminous with non-healthcare related infection. For E. coli
bacteraemia approximately half of the community-onset cases will be healthcare-
associated (E. coli bacteraemia sentinel surveillance group 2017, NHSI 2017).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
33
Examination of the trends in the age group show increases in incidence rate among the
older age groups (65-74 and ≥ 85) for males and females but also among the under ones.
Although the highest rates of E. coli bacteraemia occur in the ≥ 85 population, the
proportionally greatest increases have occurred among the <1 age groups. The generally
uniform increase in rate across the age groups suggest that there is no particular change
in the age related risk factor that is driving the increase in E. coli bacteraemia. The
exception to this is the constantly low rate of E. coli bacteraemia among males and
females in the 1-14 age group. It is not clear why rates have reduced and remained the
lowest amongst this group.
As described above, rates of E. coli bacteraemia have been rising for several years. The
number of Gram-negative bacteraemias far exceeds those caused by S. aureus
bacteraemias. In 2017/18, a total of 5,865 people died within 30 days of having E. coli
bacteraemia. For these reasons, the Secretary of State for Health introduced an ambition
to reduce healthcare-associated Gram-negative bacteraemia by March 2021(NHSI 2017,
HM Government 2018) which has now been extended to March 2024. The reduction in
rates of C. difficile infection and MRSA bacteraemia show what is possible with targeted
interventions. However, there are important differences in the epidemiology of E. coli
bacteraemia and that of MRSA bacteraemia and CDI which means that the type of
interventions introduced to control MRSA bacteraemia and CDI may not be effective in
controlling E. coli bacteraemia.
For both CDI and MRSA bacteraemia at their peak, cases were more likely to occur in the
hospital setting. This means that the patient’s environment could be more carefully
controlled and healthcare interventions readily improved. With many of the cases of E. coli
being community-onset, altering clinical practice to reduce infection rates is likely to be
harder than it was for MRSA and CDI. Urinary tract infections were found to be the main
source of Gram-negative bacteraemia cases and thus should be targeted by infection
prevention control programs if substantial reductions are to be achieved. If urinary tract
infections can be reduced or where they do occur detected and resolved quickly then a
concomitant reduction in bacteraemias should follow (P. Wilson 2019).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
34
Epidemiological analysis of Staphylococcus
aureus bacteraemia
A total of 12,878 Staphylococcus aureus bacteraemia cases were reported to PHE in
2018/19 through both the methicillin-resistant S. aureus (MRSA) bacteraemia and
meticillin-susceptible S. aureus (MSSA) bacteraemia surveillance schemes. This
represents a 0.6% increase in the numbers of bacteraemia caused by S. aureus from
2017/18 (n = 12,797) and a 30.3% increase from 2011/12 (n = 9,883) when MSSA
reporting was made mandatory.
In 2018/19, 6.3% (n = 805) of S. aureus bacteraemia reports were caused by MRSA. This
is a 44.6% decrease from 2011/12, in which 11.3% (n = 1,116) of reports were caused by
MRSA and a 5.8% decrease from 2017/18 in which 6.6% (n = 849) of reports were caused
by MRSA. At its peak MRSA bacteraemias accounted for approximately 40% of all S.
aureus bacteraemia cases in England (Johnson AP 2005).
The following sections will describe the epidemiology of MRSA and MSSA in England
separately.
Meticillin-resistant Staphylococcus aureus bacteraemia
Total reports
A total of 805 cases of MRSA bacteraemia were reported by NHS acute trusts in England
between 1 April 2018 and 31 March 2019. This is a decrease of 5.2% from 2017/18 (n =
849), and a decrease of 81.9% from 2007/08 (n = 4,451). Figure 16 shows the trends in
rates of MRSA cases for all cases and hospital-onset cases from 2007/08 to 2018/19. The
rate of all MRSA cases per 100,000 population, per year has fallen from 8.6 in 2007/08 to
1.4 in 2018/19.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
35
Figure 16: Trends in the all case and hospital-onset rate of MRSA bacteraemia in
England 2007/08-2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy. ** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Although the MRSA all-case rate for the most recent year (1.4 per 100,000 population) is
less than the rate in 2007/08 (8.6 per 100,000 population), it has remained approximately
stable for the past four years (Figure 16).
Of the 805 total cases reported in FY 2018/19, 271 were hospital-onset (0.8 per 100,000
bed days), remaining steady from 2017/18 (Table 15).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
36
Table 15: MRSA counts and rates by financial year, England: 2007/08 to 2018/19
Financial year
Mid-year population estimate
All reported
cases
Rate (all reported cases per 100,000
population) Total bed
days
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed days)
2007/08 51,594,959 4,451 8.6 37,320,817 2008/09 51,803,017 2,935 5.7 37,700,812 1,606 4.3 2009/10 52,306,371 1,898 3.6 37,326,212 1,004 2.7 2010/11 52,757,040 1,481 2.8 35,094,388 688 2.0 2011/12 53,312,604 1,116 2.1 34,502,306 473 1.4 2012/13 53,475,358 924 1.7 34,439,455 398 1.2 2013/14 53,976,973 862 1.6 34,327,781 364 1.1 2014/15 54,432,437 800 1.5 34,797,208 285 0.8 2015/16 55,018,884 823 1.5 34,576,351 298 0.9 2016/17 55,240,933 825 1.5 34,976,071 315 0.9 2017/18 55,619,430 849 1.5 34,749,047 276 0.8 2018/19 55,619,430 805 1.4 34,525,599 271 0.8
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
Trust-assigned reports
Prior to the introduction of the PIR process, cases were reported according to groups in
the apportioning algorithm, categorising cases by their time to onset in relation to patient
on the date of admission, patient location and the date of the specimen. Since only four
years of PIR data are available, of which the first year’s categorisation differed to the most
recent three, this report will also present cases grouped according to the apportioning
algorithm. Detailed data on the number of cases by PIR assignment group can be found in
Table 16.
The PIR data show a consistent decline in the rates of Clinical Commissioning Group
(CCG)-assigned cases from 0.8 cases per 100,000 population in 2013/14 to 0.4 cases per
100,000 population in 2017/18. In contrast, rates of trust-assigned MRSA bacteraemias fell
from 1.2 cases per 100,000 bed days in 2013/14 to 0.9 in 2014/15, after which trust-
assigned rates remained steady at 0.8 in 2017/18. The rate of Third Party assigned cases
has increased from 0.2 per 100,000 population to 0.6 per 100,000 population. Some of the
decline in rates of CCG-assigned cases and trust-assigned cases will be due to the
introduction of Third Party assignment, which resulted in some cases that would have
been otherwise assigned to either an acute trust or CCG, being assigned to the new
category.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
37
Table 16: MRSA counts and rates by PIR assignment, England: 2013/14 to 2018/19
Mid-year population estimate
CCG assigned Trust-assigned Third Party
Financial year
cases rate* Total bed
days cases rate** cases rate*
2013/14 53,976,973 450 0.8 34,327,781 412 1.2 2014/15 54,432,437 370 0.7 34,797,208 315 0.9 115 0.2 2015/16 55,018,884 280 0.5 34,576,351 298 0.9 245 0.4 2016/17 55,240,933 214 0.4 34,976,071 316 0.9 295 0.5 2017/18 55,619,430 224 0.4 34,749,047 293 0.8 332 0.6
* Rates for CCG and Third-party cases are given per 100,000 population. ** Rates for trust-assigned cases are given per 100,000 bed days.
Age and sex distribution
For all age and sex analyses, cases in which the sex was missing or given as unknown
were excluded. In 2007/08, 52 cases (1.2%) gave the sex as unknown. In 2018/19, 0
cases (0%) gave the sex as unknown.
Figure 17, compares the age and sex structure of MRSA bacteraemia cases in England
between 2007/08 and 2018/19. For all ages and sexes there has been a considerable
decline in rates. Despite the decline in rates, the age and sex structure has remained
consistent. For both years, MRSA bacteraemia rates were highest among patients ≥ 85
years of age. In 2007/08, the rate among males was 147.4 per 100,000 population and
44.2 among females. In 2018/19, the rate among males was 22.1 per 100,000 population
and 7.2 among females.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
38
Figure 17: Age and sex specific rates of MRSA rates per 100,000 population,
England
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
In general, the rates of MRSA bacteraemia are greater among males than among females,
and particularly so among older age groups. Although the rates of MRSA bacteraemia
have declined over time, the proportional difference in rates between sexes within an age
group have not changed appreciably. The rate ratio for males and females in ≥ 85 age
group is 3.3 (95% CI: 2.9-3.8) in 2007/08. In comparison, the rate ratio in 2018/19 was 3.1
(95% CI: 2.3-4.2) (Table 17 & 18).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
39
Table 17: MRSA counts and rates by age group and sex, England: 2007/08
Age group (years) Population n count Rate, per 100,000
Male Female Male Female Male Female
<1 333,164.7 317,035.9 31 28 9.3 8.8 1-14 4,369,016.7 4,166,420.1 22 14 0.5 0.3 15-44 10,728,560.1 10,721,325.4 252 149 2.3 1.4 45-64 6,322,972.5 6,463,536.8 594 306 9.4 4.7 65-74 2,011,279.4 2,205,682.2 583 269 29.0 12.2 75-84 1,208,494.0 1,655,252.5 884 436 73.1 26.3 ≥ 85 337,199.9 755,018.8 497 334 147.4 44.2
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
Table 18: MRSA counts and rates by age group and sex, England: 2018/19
Age group (years) Population n count Rate, per 100,000
Male Female Male Female Male Female
<1 335,035 318,432 7 7 2.1 2.2 1-14 4,813,341 4,581,557 10 10 0.2 0.2 15-44 10,749,133 10,572,163 59 60 0.5 0.6 45-64 7,014,072 7,205,186 124 52 1.8 0.7 65-74 2,646,091 2,849,090 95 42 3.6 1.5 75-84 1,435,728 1,747,546 112 57 7.8 3.3 ≥ 85 487,653 864,403 108 62 22.1 7.2
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18
were used as a proxy.
Trends in the age- and sex-specific rates of MRSA bacteraemia are shown in Figure 18.
Rates of MRSA bacteraemia have fallen across all age and sex groups since the start of
the surveillance in 2007/08. Despite this, some fluctuations in rate are seen for both sexes
in the < 1 age group in which rates increased in 2015/16 and 2016/17 (from 1.5 per
100,000 population in females in 2013/14 to 2.2 in 2016/17 and from 2.3 in males in
2013/14 to 5.0 in 2016/17). Rates have also increased in males ≥ 85 from 21.1 in 2016/17
to 22.1 in 2018/19.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
40
Figure 18: Trend in age- and sex-specific rates of MRSA bacteraemia per 100,000
population, England, 2007/08 to 2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
41
Source of MRSA bacteraemia
The HCAI Data Capture System provides acute trust users the opportunity to add
information regarding the likely source of bacteraemia. Source of bacteraemia refers to the
likely cause of the bacteraemia, such as an intravenous catheter, rather than an organ
where the infection first arose as in primary focus for Gram-negative bacteraemia. The
provision of this information is voluntary for each of the bacteraemia collections (MRSA,
MSSA, and Gram-negative bacteraemias) and has declined over time for MRSA.
In 2007/08 a total of 2,414 (54.2%) of MRSA records had entries (including “Unknown”) for
the source of bacteraemia. By 2018/19 a total of 287 (35.7%) MRSA records had entries
for the source of bacteraemia.
There have been large declines in the percentage of MRSA cases in which the source of
bacteraemia was a catheter or line. In 2007/08, catheters or lines were the source of
25.6% of cases. By 2018/19, this had declined to 12.9% of cases. In contrast, the
percentage of cases caused by skin and soft tissue infections has increased from 16.4% in
2007/08 to 33.4% in 2018/19. Between 2007/08 and 2014/15, the percentage of cases for
which the source of bacteraemia was pneumonia increased from 6.6% to 11.2%. Between
2014/15 and2017/18, the percentage of cases for which pneumonia was the source
fluctuated between 15.4% and12.3%. Trends in sources of bacteraemia are shown in
Table 19.
Table 19: MRSA counts and proportion by source of bacteraemia, England: 2018/19
Financial
year Catheters & lines* n (%)
SSTI n (%)
Pneumonia n (%) Other** n (%)
Unknown n (%)
Total n (%)
2007/08 617 (25.6) 395 (16.4) 160 (6.6) 705 (29.2) 537 (22.2) 2,414 (100.0) 2008/09 346 (22.5) 276 (17.9) 113 (7.3) 552 (35.8) 254 (16.5) 1,541 (100.0) 2009/10 178 (19.5) 191 (20.9) 63 (6.9) 328 (35.8) 155 (16.9) 915 (100.0) 2010/11 118 (17.5) 146 (21.6) 47 (7.0) 251 (37.1) 114 (16.9) 676 (100.0) 2011/12 71 (14.7) 98 (20.3) 41 (8.5) 177 (36.7) 95 (19.7) 482 (100.0) 2012/13 72 (18.3) 74 (18.8) 34 (8.6) 128 (32.5) 86 (21.8) 394 (100.0) 2013/14 39 (13.3) 57 (19.4) 33 (11.2) 100 (34.0) 65 (22.1) 294 (100.0) 2014/15 30 (11.9) 53 (20.9) 39 (15.4) 64 (25.3) 67 (26.5) 253 (100.0) 2015/16 38 (15.5) 56 (22.9) 25 (10.2) 89 (36.3) 37 (15.1) 245 (100.0) 2016/17 51 (20.0) 80 (31.4) 21 (8.2) 88 (34.5) 15 (5.9) 255 (100.0) 2017/18 50 (15.4) 101 (31.1) 40 (12.3) 118 (36.3) 16 (4.9) 325 (100.0) 2018/19 37 (12.9) 96 (33.4) 30 (10.5) 114 (39.7) 10 (3.5) 287 (100.0)
“Catheters and lines” includes the following options from the HCAI DCS question: dialysis lines, central venous catheter (CVC) associated, peripheral venous catheter (PVC) associated and intravenous (IV) lines. “*Other includes the following options HCAI DCS: endocarditis, osteomyelitis, other, prosthetic joint, surgical site infection (SSI), septic arthritis, urinary tract infection (UTI) and ventilator-associated pneumonia.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
42
Geographic distribution of MRSA bacteraemia
Some geographical variation in rates of MRSA is noted (Figure 19). The local office with
the highest rate of MRSA in 2018/19 was Q86 NHS England South West (South West
North): 2.8 cases per 100,000. The local offices with the lowest rates were Q84 NHS
England North (Lancashire and South Cumbria) and Q76 NHS England Midlands and East
(North Midlands) 0.9 cases per 100,000.
Figure 19: MRSA rates by NHS local office, England: 2018/19
*2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Table 20 gives NHS local office-specific rates. MRSA bacteraemia rates for all local offices
have declined between 2009/10 and 2018/19. However, increases in rates have been
observed for some local offices in recent years. The three regions with the largest
increases since 2016/17 are Q86 NHS England South West (South West North) , Q79
NHS England Midlands and East (East)andQ85 NHS England South West (Soyth West
South) .
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
43
Table 20: MRSA rates by NHS local office, England: 2018/19*
Code NHS local office
20
09
/10
20
10
/11
20
11
/12
20
12
/13
20
13
/14
20
14
/15
20
15
/16
20
16
/17
20
17
/18
20
18
/19
Q76 NHS England Midlands and East (North Midlands)
4.2 2.6 1.7 1.2 1.0 1.5 1.2 1.2 1.1 0.9
Q77 NHS England Midlands and East (West Midlands)
2.5 2.6 1.6 1.7 1.1 1.6 1.2 1.1 0.7 1.0
Q78 NHS England Midlands and East (Central Midlands)
2.4 2.4 1.7 0.9 1.1 1.3 1.1 1.1 1.1 1.1
Q79 NHS England Midlands and East (East) 3.1 1.9 1.8 1.4 1.2 1.2 1.5 1.7 1.8 2.0 Q72 NHS England North (Yorkshire and Humber) 4.2 3.4 2.5 2.0 1.9 1.3 1.3 1.5 1.6 1.5 Q74 NHS England North (Cumbria and North East) 3.1 2.4 1.3 1.9 1.2 1.6 1.5 1.6 1.4 1.0 Q75 NHS England North (Cheshire and Merseyside) 4.6 3.5 2.7 2.6 2.2 1.6 1.8 1.8 1.7 1.6 Q83 NHS England North (Greater Manchester) 4.6 3.5 2.4 2.3 1.9 1.5 1.9 2.1 1.6 1.9 Q84 NHS England North (Lancashire and South
Cumbria) 3.5 2.8 1.9 1.5 1.4 1.4 1.7 1.4 1.1 0.9
Q87 NHS England South East (Hampshire, Isle Of Wight and Thames Valley)
2.7 2.0 1.8 1.1 1.5 1.2 1.5 1.1 1.4 1.2
Q88 NHS England South East (Kent, Surrey and Sussex)
3.9 3.0 2.5 1.8 1.9 1.4 1.4 1.6 1.8 1.6
Q85 NHS England South West (South West South) 2.6 2.1 1.6 1.3 1.0 1.0 1.2 0.9 1.2 1.2 Q86 NHS England South West (South West North) 4.2 2.0 2.1 2.2 1.7 2.3 2.3 2.5 2.7 2.8 Q71 NHS England London 4.6 3.7 2.7 2.2 2.3 1.8 1.8 1.7 1.9 1.6 National 3.6 2.8 2.1 1.7 1.6 1.5 1.5 1.5 1.5 1.4
* 2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Discussion
MRSA bacteraemia rates have declined year on year from 2007/8 to 2014/15, however in
2015/6 the rate increased slightly but remained the same in 2016/17. In 2018/19 the rate
has decreased for all cases, and hospital-onset cases.
Over time, numerous interventions aimed at reducing the incidence of MRSA bacteraemia
and other infections have been introduced. These include the DoH policy document
‘Winning Ways’, published in 2003 (Department of Health 2003b); The ‘CleanYourHands’
campaign launched by the National Patient Safety Agency in 2014 (Stone et al. 2012); the
‘Saving Lives’ programme launched by the DoH in 2005 which included the ambition to
halve MRSA rates by 2008 (Department of Health 2005); the 2006 Health Act which
introduced a code of practice to provide guidance on reducing HCAI including MRSA
(Department of Health 2006); and the Health and Social Care Act, 2008 which requires the
code of practice to be regularly updated (Department of Health 2008).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
44
Thus, the epidemiology of MRSA has changed since its peak, with the greatest proportion
of cases now being community-onset. This switch in setting is most likely due to the
majority of MRSA bacteraemia interventions being targeted to the acute care setting and
thus largest reductions in MRSA bacteraemias were seen in hospital-onset cases. The
duration of stay of hospital patients is also on the decline. In 2001, the average length of
stay for a hospitalised patient in the UK was 7.4 days, this decreased to 5.9 days in 2013
(Organisation of Economic Co-operation and Development 2018). The reduced hospital
stay lessens the risk of acquiring a hospital-acquired infection. A declining length of stay
could also mean patients may acquire an MRSA infection in an acute care setting and yet
not display symptoms until they have returned to the community. Such cases would then
be readmitted, at which point the time-to-onset would be less than 2 days, according to the
definitions used here. It is also possible that early detection of MRSA bacteraemia is
improving with advances in diagnostics as well as general improvement in clinical
awareness of sepsis (Sepsis Trust, 2013).
The PIR process was introduced in April 2013 to aid the zero-tolerance policy for MRSA
bacteraemia (NHS England 2014b). Since then, rates of CCG-associated cases have
fallen each year. Trust-assigned rates fell between 2013/14 and 2014/15 but then
remained stable between 2014/15 and 2017/18. The PIR process has now been made a
local-only process and will not be reported further in these reports.
The percentage of MRSA bacteraemia where the likely source of infection was a catheter
or a line, has shown steady decrease between 2007/8 and 2014/15, with a sudden sharp
rise between 2015/16 to 2017/18, and subsequently returned to levels seen prior to
2015/16. Declines in the proportion of MRSA bacteraemia where the most likely source of
infection is a catheter or line have been noted previously and may be due to greater
clinical awareness of the importance of this route of infection and the introduction of care
bundles aimed at reducing infections in intra-vascular lines and selective decolonization of
patients with MRSA carriage.
The percentage where the primary focus is something other than catheters or intravenous
lines have fluctuated considerably over time. The percentage of infections whose likely
source is skin or soft tissue infection does appear to have increased over time, from 16.4%
in 2007/8 to 33.4% in 2018/19. However, in 2018/19 approximately 35.7% of records had
information on the likely source of bacteraemia and therefore, interpretation of these data
should be approached cautiously.
Considered together, the stabilisation in the rate of trust-assigned cases and the increase
in the percentage of infections due to catheters and lines since 2014/15 highlight the need
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
45
for continued vigilance and continued efforts on infection prevention initiatives to reduce
hospital-onset cases further. In addition, to maintain good practice in an acute trust setting,
interventions are required in the community setting, considering the majority of MRSA
cases are now community-onset. The geographical distribution of MRSA bacteraemia
across England does not show any obvious pattern. Highest rates are observed in the
South West of England, and lowest rates in the West Midlands. Elsewhere, the rates of
MRSA bacteraemia are relatively evenly distributed across the country.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
46
Meticillin-susceptible Staphylococcus aureus bacteraemia
Total reports
A total of 12,073 cases of MSSA bacteraemia were reported by NHS acute trusts in
England between 1 April 2018 and 31 March 2019. This is broadly similar to 2017/18 (n =
11,948), and an increase of 37.7% from 2011/12 (n = 8,767). Figure 20 shows the trends
in rates of MSSA cases for all cases and hospital-onset cases from 2011/12 to 2018/19.
The rate of all MSSA cases per 100,000 population, per year has risen from 16.4 in
2011/12 to 21.7 in 2018/19.
Figure 20: Trends in the rate of MSSA bacteraemia in England
*Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy. ** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
47
Hospital-onset reports
Of the 12,073 total cases reported in FY 2018/19, 3,322 were hospital-onset (9.6 per
100,000 bed days). Similar to all-reported MSSA cases, the incidence rate for hospital-
onset MSSA cases has increased steadily (from 7.8 in 2012/13 to 9.6 in 2018/19, a
change of 22.7%). Despite the increase in hospital-onset rates, the percentage of cases
that were hospital-onset has declined from 32.6% in 2011/12 to 27.5% in 2018/19 (Table
21).
Table 21: MSSA counts and rates by financial year, England: 2011/12 to 2018/19
Financial year
Mid-year population estimate
All reported
cases
Rate (all reported cases per 100,000
population) Total bed
days
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed days)
2011/12 53,312,604 8,767 16.4 34,502,306 2,854 8.3 2012/13 53,475,358 8,812 16.5 34,439,455 2,700 7.8 2013/14 53,976,973 9,290 17.2 34,327,781 2,696 7.9 2014/15 54,432,437 9,862 18.1 34,797,208 2,807 8.1 2015/16 55,018,884 10,608 19.3 34,576,351 2,921 8.4 2016/17 55,240,933 11,497 20.8 34,976,071 3,098 8.9 2017/18 55,619,430 11,948 21.5 34,749,047 3,152 9.1 2018/19 55,619,430 12,073 21.7 34,525,599 3,322 9.6
Age and sex distribution
For all age and sex analyses, cases in which the sex was missing or given as unknown
were excluded. In 2011/12, 265 cases (3%) gave the sex as unknown. In 2018/19, 7 cases
(0.1%) gave the sex as unknown.
Figure 21 compares the age and sex structure of MSSA bacteraemia cases in England
between 2011/12 and 2018/19. For both years, MSSA bacteraemia rates were highest
among patients ≥85 years of age. In 2011/12, the rate among males was 134.5 per
100,000 population and 65.6 among females. In 2018/19, the rate among males was 186
per 100,000 population and 77.7 among females. The rate of MSSA bacteraemia is also
particularly high among <1s (186 per 100,000 population and 77.7 in 2018/19).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
48
Figure 21: Age and sex structure of MSSA rates per 100,000 population, England
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
In general, the rates of MSSA bacteraemia are greater among males than among females,
and particularly so among older age groups. As the rates of MSSA bacteraemia have
increased over time, the rate ratio between sexes has increased somewhat. The rate ratio
for MSSA between males and females 85 years and over in 2011/12 is 2.0 (95% CI: 1.8-
2.3) compared with 2.4 (95% CI: 2.2-2.6) in 2018/19. Rates are also much higher in <1
age group, much more than what is seen with MRSA in this age group. Age group data is
tabulated in Table 22 and 23.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
49
Table 22: MSSA counts and rates by age group and sex, England: 2011/12
Age group (years)
Population* n cases Rate, per 100,000
Male Female Male Female Male Female
<1 350,817.8 333,993.3 251 189 71.5 56.6 1-14 4,476,176.1 4,269,121.6 283 174 6.3 4.1 15-44 10,799,091.6 10,738,537.1 849 482 7.9 4.5 45-64 6,678,566.3 6,837,399.2 1,409 731 21.1 10.7 65-74 2,241,803.6 2,422,555.7 947 516 42.2 21.3 75-84 1,295,415.4 1,666,592.1 964 646 74.4 38.8 ≥ 85 394,885.2 807,649.1 531 530 134.5 65.6
Table 23: MSSA counts and rates by age group and sex, England: 2018/19
Age group (years)
Population* n cases Rate, per 100,000
Male Female Male Female Male Female
<1 335,035 318,432 213 139 63.6 43.7 1-14 4,813,341 4,581,557 306 225 6.4 4.9 15-44 10,749,133 10,572,163 1,300 814 12.1 7.7 45-64 7,014,072 7,205,186 2,117 1,039 30.2 14.4 65-74 2,646,091 2,849,090 1,499 687 56.6 24.1 75-84 1,435,728 1,747,546 1,358 790 94.6 45.2 ≥ 85 487,653 864,403 907 672 186.0 77.7
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Rates of MSSA bacteraemia among older males have increased substantially since
2013/14. In 2013/14 the rate of MSSA bacteraemia was 136.6 among those ≥ 85. By
2018/19, this rate was 186.0 (Figure 22 below). Interestingly, rates appear to have
reduced in the <1 age group, and have increased in all the populations >14, when
comparing the 2007/08 figures to 2018/19.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
50
Figure 22: Trend in age and sex structure of MSSA cases and rate per 100,000
population, England, 2011/12 to 2018/19
* *Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
51
Source of MSSA bacteraemia
In 2011/12 a total of 3,305 (37.7%) records had entries for the source of bacteraemia. By
2018/19 a total of 3,724 (30.8%) had entries for the source of bacteraemia.
The percentage of cases caused by skin and soft tissue infections has increased from
20.3% in 2011/12 to 27.6% in 2018/19. The percentage of infections caused by
pneumonia has also risen substantially from 6% in 2011/12 to 12.1% in 2018/19. There
have been fluctuations in the percentage of MSSA cases in which the source of
bacteraemia was from either a catheter or line. In 2011/12, catheters or lines were the
source of 17.1% of cases. By 2018/19, this had declined to 15.4% of cases.
The percentage of cases for which the source of infection was unknown has decreased
from 23.6% in 2011/12 to 2.1% in 2018/19. Trends in sources of bacteraemia are shown in
Table 24.
Table 24: MSSA counts and rates by source of bacteraemia, England: 2018/19
Financial
year Catheters & lines* n (%) SSTI n (%)
Pneumonia n (%)
Other** n (%)
Unknown n (%) Total n (%)
2011/12 565 (17.1) 670 (20.3) 197 (6.0) 1093 (33.1) 780 (23.6) 3,305 (100.0) 2012/13 492 (15.1) 699 (21.4) 232 (7.1) 1088 (33.3) 755 (23.1) 3,266 (100.0) 2013/14 435 (13.4) 684 (21.1) 218 (6.7) 1124 (34.7) 775 (23.9) 3,236 (100.0) 2014/15 445 (13.1) 706 (20.8) 305 (9.0) 1087 (32.0) 855 (25.2) 3,398 (100.0) 2015/16 493 (15.3) 769 (23.9) 306 (9.5) 1169 (36.3) 487 (15.1) 3,224 (100.0) 2016/17 499 (15.7) 871 (27.4) 365 (11.5) 1279 (40.2) 169 (5.3) 3,183 (100.0) 2017/18 516 (14.6) 1003 (28.4) 445 (12.6) 1440 (40.8) 125 (3.5) 3,529 (100.0) 2018/19 575 (15.4) 1028 (27.6) 450 (12.1) 1594 (42.8) 77 (2.1) 3,724 (100.0)
“Catheters and lines” includes the following options from the HCAI DCS question: dialysis lines, central venous catheter (CVC) associated, peripheral venous catheter (PVC) associated and intravenous (IV) lines. “^Other includes the following options HCAI DCS: endocarditis, osteomyelitis, other, prosthetic joint, surgical site infection (SSI), septic arthritis, urinary tract infection (UTI) and ventilator-associated pneumonia.
Geographic distribution of MSSA bacteraemia
Some geographical variation in rates of MSSA is noted (Figure 23). The local offices with
the highest rates of MSSA were in the north of England Q75 NHS England North
(Cheshire and Merseyside): 30.4, followed by both Q84 NHS England North (Cumbria and
North East) and Q81 NHS England North (Lancashire and South Cumbria): both reporting
28 MSSA bacteraemia per 100,000 population. However, there is one local office that
bucks the trend of higher rates in the north: Q85 NHS England South West (South West
South) had a rate of 25.8 per 100,000 population (Figure 23).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
52
The local office with the lowest rate in 2018/19 was Q71 NHS England London with a rate
of 16.6 cases per 100,000 population.
Figure 23: Geographic distribution of MSSA rates per 100,000 population, England
2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Table 25 gives NHS local office-specific rates. For most local offices, rates of MSSA
bacteraemia followed the national trend; one of increasing incidence over time. The
exception of Q83 NHS England North (Greater Manchester) and Q84 NHS England
(Lancashire and South Cumbria) showing a slight decrease on last year’s rates.
The largest increase in incidence was observed in Q74 NHS England North (Cumbria and
North East) which saw the incidence rise from 17.3 cases per 100,000 population in
2011/12 to 28 in 2018/19. Q75 NHS England North (Cheshire and Merseyside) also saw a
large increase in incidence from 20.7 cases per 100,000 population in 2011/12 to 30.4 in
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
53
2018/19. In contrast, two regions (NHS England North (Cumbria and North East) and NHS
England London) close to stable rates of MSSA over time.
Table 25: MSSA rates by NHS local office per 100,000 population, England: 2011/12
to 2018/19*
Code NHS local office
2011
/12
2012
/13
2013
/14
2014
/15
2015
/16
2016
/17
2017
/18
2018
/19
Q76 NHS England Midlands and East (North Midlands)
17.9 19.5 19.1 21.6 23.4 24.7 24.3 24.4
Q77 NHS England Midlands and East (West Midlands)
17.2 16.6 17.6 17.3 16.4 18.6 20.0 20.2
Q78 NHS England Midlands and East (Central Midlands)
15.4 14.6 13.4 15.8 16.7 18.4 18.9 18.1
Q79 NHS England Midlands and East (East) 14.0 13.2 13.9 14.7 17.6 17.9 18.8 18.9 Q72 NHS England North (Yorkshire and Humber) 20.7 19.3 20.3 20.4 23.2 24.1 23.5 25.4 Q74 NHS England North (Cumbria and North East) 17.3 17.6 20.1 19.6 25.0 26.2 26.6 28.0 Q75 NHS England North (Cheshire and Merseyside) 20.7 20.2 21.1 24.7 24.3 28.4 27.6 30.4 Q83 NHS England North (Greater Manchester) 16.2 18.3 19.3 17.2 20.5 21.3 24.9 23.9 Q84 NHS England North (Lancashire and South
Cumbria) 17.6 20.5 20.4 22.8 24.8 23.7 26.5 24.3
Q87 NHS England South East (Hampshire, Isle Of Wight and Thames Valley)
13.7 13.7 15.2 17.6 17.2 19.3 19.4 19.8
Q88 NHS England South East (Kent, Surrey and Sussex)
14.1 14.8 15.4 17.3 17.9 20.4 21.2 20.8
Q85 NHS England South West (South West South) 18.6 16.8 20.1 21.9 22.9 25.1 27.4 25.8 Q86 NHS England South West (South West North) 16.3 15.6 16.3 17.6 17.1 19.8 17.8 20.9 Q71 NHS England London 14.6 15.5 15.6 15.0 14.9 16.0 17.1 16.6 National 16.4 16.5 17.2 18.1 19.3 20.8 21.5 21.7 * Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Discussion
Mandatory surveillance of MSSA bacteraemia was introduced in January 2011. In the first
year, MSSA bacteraemia rates were stable (2011/12 to 2012/13), before increasing each
subsequent year. In contrast, hospital-onset rates fell slightly in the first year of the
mandatory surveillance programme being introduced; beyond this rates have been
increasing slightly year on year.
Age and sex structure of MSSA bacteraemia show rates increase with age group, with the
highest rate seen in the 85 and over age group, furthermore males also show higher rates
of MSSA bacteraemia compared to females. However, rates of MSSA bacteraemia are
also seen to be very high in the youngest age group, <1 years old, compared to young
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
54
patients (1-14 years old) and adults (15 to 74 years old). This contrasts with MRSA
bacteraemia in which rates among the <1 age group are close to the rates in the 1-14 age
group.
Research using data gathered by the mandatory surveillance found that cases arising in
very young patients were most likely attributable to healthcare-associated infections, rather
than community-associated infections and were related to intravascular devices
(Abernethy et al. 2017). Despite the increase in MSSA bacteraemia rates over the years,
little change has been observed in rates between males and females, with the rate in men
85 years or greater being 2.4 times greater in 2018/19 than amongst women of the same
age, 186.0 and 77.7 cases per 100,000 population, respectively.
Levels of completion of the likely source of infection for MSSA have declined over time. In
2017/18, only 30.8% of records had a source of infection information available, a slight
increase on 29.3% from last year. There has been a gradual increase over the years in the
percentage of cases where the source has been either, skin or soft tissue or pneumonia,
however a slight decrease has been observed for both in 2018/19, which is also mirrored
in MRSA bacteraemias. However it must be noted that likely source is not a mandatory
field in the DCS.
Geographically there appears to be no trends or similarities in rates of infections between
MRSA and MSSA. With a couple of exceptions MSSA appears to be most prevalent in the
north. This contrasts with the geographic distribution of MRSA in which no particular North-
South distribution was observed.
The epidemiology of MRSA and MSSA bacteraemia also show differences in mortality.
The 30-day mortality in 2017/18 was reported as 25.9% for MRSA and 19.5% for MSSA
(PHE 2018).
As rates of MRSA continue to fall steadily, we see the MSSA bacteraemia rates have
slowly increased, in both hospital- and community-onset cases. The increase in MSSA
cases appears to be driven by the increases in community-onset cases. In 2018/19, 72.5%
of all reported MSSA bacteraemias were community-onset, reasons for this are unclear.
This indicates that the focus for interventions needs to placed on community-onset MSSA
cases.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
55
Epidemiological analysis of Clostridioides
difficile Infection
A total of 12,275 cases of Clostridioides difficile infection were reported by NHS trusts in
England between 1 April 2018 and 31 March 2019. This is a small decrease of 7.7% from
2017/18 (n = 13,299) and a decrease of 77.9% from 2007/08 (n = 55,498). Figure 24
shows the trends in rates of CDI cases for all cases and hospital-onset cases from
2007/08 to 2018/19. The rate of all CDI cases per 100,000 population, per year has fallen
from 107.6 in 2007/08 to 22.1 in 2018/19.
Figure 24: Trends in the rate of C. difficile infection in England
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy. ** bed day data were not available for quarter 4 of FY 2018/19 (January to March, 2019). As a result, the 2018/19 bed day data is an aggregate of quarters 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
56
Hospital-onset reports
Of the 12,275 total cases reported in FY 2018/19, 4,201 were hospital-onset (12.2 per
100,000 bed-days). It should be noted that CDI cases are considered hospital-onset if they
occur ≥ 4 days after admission to an acute trust, where day of admission is day 1. This is
in contrast to ≥ 3 days for bacteraemia cases. The incidence rate for hospital-onset CDI
cases mirrors the trends in incidence for all cases, with declining rates from 2007/08 to
2013/14 which then remained approximately stable to 2017/18. The rate of hospital-onset
CDI cases decreased slightly from 13.6 in 2017/18 to 12.2 in 2018/19, a change of 10.8%
(Table 26).
Table 26: CDI counts and rates by financial year, England: 2007/08 to 2018/19
Financial year
Mid-year population estimate
All reported
cases
Rate (all reported cases per 100,000
population) Total bed-
days
Hospital-onset cases
Rate (Hospital-onset cases per
100,000 bed-days)
2007/08 51,594,959 55,498 107.6 37,320,817 33,434 89.6 2008/09 51,803,017 36,095 69.7 37,700,812 19,927 52.9 2009/10 52,306,371 25,604 49.0 37,326,212 13,220 35.4 2010/11 52,757,040 21,707 41.1 35,094,388 10,417 29.7 2011/12 53,312,604 18,022 33.8 34,502,306 7,689 22.3 2012/13 53,475,358 14,694 27.5 34,439,455 5,980 17.4 2013/14 53,976,973 13,362 24.8 34,327,781 5,034 14.7 2014/15 54,432,437 14,193 26.1 34,797,208 5,233 15.0 2015/16 55,018,884 14,143 25.7 34,576,351 5,162 14.9 2016/17 55,240,933 12,847 23.3 34,976,071 4,622 13.2 2017/18 55,619,430 13,299 23.9 34,749,047 4,740 13.6 2018/19 55,619,430 12,275 22.1 34,525,599 4,201 12.2
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy. ** Bed day data population data were not available for quarter 4 of FY 2018/19 (January to March 2019). As a result, the 2018/19 bed day data is an aggregate of quarter 1, 2 and 3 of 2018/19 and quarter 4 of 2017/18.
Prior trust exposure
In April 2017, the mandatory surveillance captured information on whether a patient with
CDI had been previously admitted to the reporting trust in the past 84 days. These data
have previously been reported as monthly counts in the monthly tables produced by the
mandatory surveillance team but have not previously been published as annual rates.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
57
With the prior trust exposure, cases are split into the following groups:
Hospital-onset, healthcare associated (HOHA) - Cases where the date of onset is two or
more days after the date of admission, where date of admission is day zero and the
patient was admitted to an NHS acute trust at time of specimen.
Community-onset, healthcare associated (COHA) - Cases which are not HOHA but have
previously been discharged from the reporting organisation within the 28 days prior to
specimen date.
Community-onset, indeterminate association (COIA) - Cases which are not HOHA but
have previously been discharged from the reporting organisation but have been
discharged from the reporting organization between 28 and 84 days prior to specimen
date.
Community-onset, community associated (COCA) - Cases which are not HOHA and have
not previously been discharged from the reporting organisation within the past 84 days.
Cases for which “Don’t know” was recorded against the prior trust exposure status are
grouped into “Unknown” and cases for which no information about the prior trust exposure
was entered are grouped into “No information”. Both of these groups are included in the
community-onset group as they are known not to be HOHA cases (Table 27).
Table 27: Prior healthcare exposure of CDI cases by financial year, England: 2017/18
to 2018/19*
Financial year
Total bed-days
HOHA
COHA COIA COCA Unknown* No
information** Total cases Count Rate
2017/18 34,749,047 5,465 15.7 1,890 1,064 2,857 262 1,761 13,299 2018/19 34,525,599 4,871 14.1 2,271 1,330 3,467 331 5 12,275 *The record indicates that it is unknown whether the patient was admitted to the reporting organisation in the past three months. **No information was entered for the prior trust exposure. HOHA – Hospital-onset, healthcare associated COHA - Community-onset, healthcare associated COIA - Community-onset, indeterminate association COCA – Community-onset, community associated
The count of HOHA dropped from 5,465 in 2017/18 to 4,871 in 2018/19. In contrast, each
of the community-onset categories increased in counts of cases between 2017/18 and
2018/19. The count of cases for which no prior trust information was provided, dropped
from 1,761 in 2017/18 to 5 in 2018/19
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
58
Seasonal trends in CDI
Data from previous Annual Epidemiological Commentaries have indicated a shift in the
seasonality of hospital-onset CDI. In 2010/11, 28.8% of hospital-onset cases were
reported in the first quarter of the financial year, with declining percentages reported in the
subsequent quarters of the financial year (Figure 25). hospital-onset cases in 2011/12
showed a similar distribution. However, in 2014/15, hospital-onset cases were more evenly
distributed throughout the year. Cases reported in the first quarter of the year made up
23.0% of cases and cases reported in the fourth quarter of the year made up 26.1%. From
2015/16 onwards, the second quarter of the year (July-September) saw the greatest
proportion of cases reported (between 26% and 29% of cases each year).
In contrast, community-onset cases have always shown a peak in the second quarter of
the financial year, with this quarter forming 28% to 29% of cases for the financial year.
Figure 25: CDI counts by financial year and quarter, England 2010/11 to 2018/19
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
59
Age and sex distribution
For all age and sex analyses, cases in which the sex was missing or given as unknown
were excluded. In 2007/08, the sex was reported as unknown for 667 cases (1.2%). In
2018/19, the sex was reported as unknown for 22 cases (0.2%).
Figure 26 compares the age and sex structure of CDI cases in England between 2007/08
and 2018/19. For both years, CDI rates were highest among patients ≥ 85 years of age. In
2007/08, the rate among men was 1,504.7 per 100,000 population and 1,490.2 among
females. In 2018/19, the rate among males was 215.7 per 100,000 population and 221
among fenales.
Figure 26: Age and sex structure of CDI rates per 100,000 population, England
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Table 28 and 29 show the comparison of the sex structure of CDI rates reveals very little
gap in rates between men and women. The rate ratio between men and women ≥85 years
old is 1.0 (95% CI: 0.9-1.1).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
60
Table 28: CDI counts and rates by age group and sex, England: 2007/08
Age group (years)
Population n cases Rate, per 100,000
Male Female Male Female Male Female
2-14 4,046,149.6 3,858,142.2 178 157 4.4 4.1 15-44 10,728,560.1 10,721,325.4 1,251 1,539 11.7 14.4 45-64 6,322,972.5 6,463,536.8 3,479 3,333 55.0 51.6 65-74 2,011,279.4 2,205,682.2 4,970 4,904 247.1 222.3 75-84 1,208,494.0 1,655,252.5 7,974 10,721 659.8 647.7 ≥ 85 337,199.9 755,018.8 5,074 11,251 1,504.7 1,490.2
Table 29: CDI counts and rates by age group and sex, England: 2018/19
Age group (years)
Population n cases Rate, per 100,000
Male Female Male Female Male Female
2-14 4,467,295 4,252,796 162 128 3.6 3.0 15-44 10,749,133 10,572,163 400 672 3.7 6.4 45-64 7,014,072 7,205,186 903 1,118 12.9 15.5 65-74 2,646,091 2,849,090 1,149 1,370 43.4 48.1 75-84 1,435,728 1,747,546 1,424 1,965 99.2 112.4 ≥ 85 487,653 864,403 1,052 1,910 215.7 221.0
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Trends in age and sex
Although CDI rates are greatest among the elderly, as a percentage of cases reported to
the mandatory surveillance, there has been an increase among younger age groups, and
a concomitant decline in the percentage of cases that are from older age groups. In
2007/08, among females, the percentage of cases that occurred among those ≥85 years
old was 35.3%. In 2018/19, this had declined to 26.7%. Trends in the percentage of cases
among females between 75 and 84 years old has also declined over the same time period.
In 2007/08, among females, the percentage of cases that occurred among those between
75 and 84 years old was 33.6%. In 2018/19, this had declined to 27.4%.
In contrast, the percentage of cases that occurred among females between 65-74 years
old was 15.4% in 2007/08 and 19.1% in 2018/19. Younger age groups have also seen
increases in the percentage of cases which they form.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
61
The increase in percentage of cases among younger age groups has not been seen in
males. In 2007/08, 22.1% of cases among males occurred in the ≥85 age group. In
2018/19, this was 20.7%. A small increase has been observed among the 15-44 age
group in males from 5.5% in 2007/08 to 7.9% in 2018/19.
It is important to note that this only reflects percentages of cases and that, because of the
age and sex structure of the population of England, the incidence rate by age and sex can
be very different.
Figure 27: Trend in age and sex structure of CDI cases ant rate per 100,000
population, England, 2007/08 to 2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
62
Geographic distribution of CDI
The geographical distribution of CDI incidence in England shows a north-south divide
(Figure 28). The local offices with the highest rates of CDI are in the north of England Q74
NHS England North (Cumbria and North East): 29.2 CDI per 100,000 population, Q75
NHS England North (Cheshire and Merseyside): 29.7 and Q84 NHS England North
(Lancashire and South Cumbria): 28.9). In contrast, the local office with the lowest rate in
2018/19 was Q71 NHS England London with a rate of 15.2 cases per 100,000 population.
Figure 28: Geographic distribution of CDI rates per 100,000 population, England
2018/19
* Mid-year population estimates for 2018/19 were not available at time of publication and so population data for 2017/18 were used as a proxy.
Table 30 gives NHS local office-specific rates. All local offices showed declining trends in
the rate of CDI since 2009/10. The greatest declines occurred in Q74 NHS England North
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
63
(Cumbria and North East) and Q75 NHS England North (Cheshire and Merseyside). For
Q74 NHS England North (Cumbria and North East) the rate of CDI was 74.6 for 2009/10
and 29.2 for 2018/19. For Q74 NHS England North (Cheshire and Merseyside) the rate of
CDI was 74.2 for 2009/10 and 29.7 for 2018/19.
Table 30: CDI rates by NHS local office, England: 2011/12 to 2018/19*
Code NHS local office
20
11
/12
20
12
/13
20
13
/14
20
14
/15
20
15
/16
20
16
/17
20
17
/18
20
18
/19
Q76 NHS England Midlands and East (North Midlands) 50.8 45.3 32.3 31.4 28.5 30.7 30.5 25.0 Q77 NHS England Midlands and East (West Midlands) 55.5 47.9 45.9 30.0 26.8 25.8 25.6 24.9 Q78 NHS England Midlands and East (Central Midlands) 45.9 35.1 26.2 24.1 22.8 23.1 22.6 20.1 Q79 NHS England Midlands and East (East) 40.6 31.1 24.0 22.5 20.4 26.0 24.8 22.8 Q72 NHS England North (Yorkshire and Humber) 46.2 43.5 39.8 31.8 29.6 28.7 28.1 26.7 Q74 NHS England North (Cumbria and North East) 74.6 46.4 37.9 35.0 28.0 32.4 34.1 27.4 Q75 NHS England North (Cheshire and Merseyside) 74.2 55.9 39.0 30.8 28.6 31.5 31.0 29.4 Q83 NHS England North (Greater Manchester) 67.0 58.3 42.9 35.5 29.4 30.9 29.5 28.3 Q84 NHS England North (Lancashire and South Cumbria) 58.9 46.2 35.3 30.5 28.0 33.9 34.7 25.1 Q87 NHS England South East (Hampshire, Isle Of Wight
and Thames Valley) 45.5 39.5 34.5 24.3 21.7 21.1 21.0 20.2
Q88 NHS England South East (Kent, Surrey and Sussex) 42.4 38.6 28.4 25.8 23.8 23.1 23.1 24.5 Q85 NHS England South West (South West South) 43.8 43.5 43.2 27.5 25.0 28.7 29.3 23.5 Q86 NHS England South West (South West North) 53.0 41.4 37.1 33.1 31.6 27.9 28.3 25.2 Q71 NHS England London 33.5 31.1 25.5 20.1 17.9 19.9 18.7 16.5 National 49.0 41.1 33.8 27.5 24.8 26.1 25.7 23.3
* 2018/19 population data were not available at time of preparation and 2017/18 population data were used in place.
Discussion
Between 2007/08 and 2012/13 rates of CDI fell rapidly. Since 2012/13 rates have
continued to decline, but much less rapidly. The rapid decline in the rate of all cases of
CDI has been mirrored in hospital-onset cases. However, the decline in community-onset
cases has not been so rapid and community-onset cases now constitute 66% of cases,
where as in 2011/12 it only accounted for 57.3% . Many of the interventions aimed at the
reduction of CDI rates were targeted at the hospital setting, which likely explains the
greater reduction in hospital-onset cases compared to community-onset cases. However,
the division of cases into hospital-onset and community-onset cases ignores the effect of
any prior admissions to hospital which could increase the risk of CDI. For this reason, and
to better align surveillance in England with that performed by ECDC and CDC, information
on prior trust exposure was introduced in April 2017.
The prior trust exposure classification groups cases according to whether the patient has
previously been admitted to the reporting trust in the past three months (84 days). Since
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
64
the prior trust exposure classification of cases has only been running for two years, it is
hard to make comments about whether observed differences between years represent
long-term trends. Hospital-onset, healthcare-associated cases formed the largest group by
the new apportioning method, but showed a considerable decline from 5,465 cases in
2017/18 to 4,871 cases in 2018/19. In contrast, the counts of cases for community-onset
groups have each increased, at least partly because of a change in the provision of
information on prior trust exposure. The proportion of cases which lack any information on
prior trust exposure declined from 13% in 2017/18 to <0.1% in 2018/19, probably as a
result of making the questions around prior trust exposure mandatory.
Although the prior trust exposure only records admission to the reporting organisation,
analysis by the mandatory surveillance team using data from the Hospital Episodes
Statistics estimates that this captures over 80% of all hospital interactions.
Rates of CDI are highest among older age groups; those ≥ 75 years of age. Recent
research has shown that over time, elderly individuals are getting frailer and experiencing
polypharmacy (Melzer et al. 2015). A frailer population, receiving greater levels of
medication would suggest that greater levels of healthcare interaction were being
experienced by this age group. Despite this, rates of CDI have decreased considerably
among the oldest age group. The reason for this difference is likely to be due to the care
bundle introduced by the DHSC which recommended the use of personal protective
equipment, cohort nursing and environmental decontamination which were geared towards
preventing HCAI in acute care settings and not community settings. (Department of Health
2007).
The decline in rates has done little to affect the age- and sex-specific rates of CDI. Those
patients 85 years or older are the most frequently affected and there is little difference in
the rates between the sexes. This marks a difference in epidemiology between CDI and
bacteraemia in which rates are higher among males than females in almost all age groups.
The shift in seasonality is intriguing and currently lacks an explanation. Between 2010/11
and 2013/14, most hospital-onset and allsi community cases occurred in the April-to-June
quarter of the financial year. From 2016/17 onwards, the quarter with the greatest number
of cases was the July-to-September quarter.
In contrast to the geographic distribution of both MSSA and E. coli bacteraemia, which
both show a weak north-south divide in rates, there is a clear north-south split in the
distribution of CDI with higher rates in the north of England than in the south. Previous
work by this group has indicated an association between higher levels of deprivation and
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
65
higher rates of CDI, the same is true for E. coli and MSSA, although the strongest
association was seen in E. coli.
The decline in rates of CDI is difficult to explain fully, in terms of which interventions had
the largest effect, in part because the interventions were not applied in isolation. A number
of the interventions aimed at reducing MRSA rates were also aimed at reducing CDI (the
‘CleanYourHands’ and Saving Lives campaigns). It is perhaps not surprising, then, that the
trends in rates of CDI mirror those of MRSA. In addition, targets to reduce CDI cases were
introduced in 2008. These aimed to reduce the number of cases reported annually to 30%
of the 2007/08 count by 2010/11 (Department of Health 2008). Failure to achieve targets
was associated with financial penalties. Financial sanctions against acute trusts for
exceeding expected numbers of CDI were reduced in 2014/15 from £50,000 to £10,000
per case (NHS England 2014a), in addition, only those cases deemed to be associated
with a lapse in care are now subject to the financial sanctions. This was introduced in
recognition of the fact that some cases can occur even if best practice is followed.In
addition to the objectives and the ‘CleanYourHands’ and Saving Lives campaigns,
guidance was issued aiming to reduce clindamycin, cephalosporin or fluoroquinolone
prescribing, which had been shown to promote the spread of epidemic strains of C. difficile
(Department of Health and Health Protection Agency 2008). The resulting reduction in
prescribing of fluoroquinolones and cephalosporins has been associated with a significant
decline in the incidence of CDI, in regional analysis (Dingle et al. 2017).
The epidemiology of CDI is changing from a mostly hospital-onset infection to a
community-onset one, along with this change of setting the distribution of strains causing
CDI in England has changed over time. A C. difficile ribotype (ribotype 027) associated
with the high numbers of CDI cases in the early 2000s is no longer as prevalent.(Wilcox et
al. 2012) This particular ribotype is associated with poorer patient outcomes, reductions in
fataity rate from 26.3% of cases in 2007/08 to 15.2% in 2017/18, may be in part due to the
reduction in prevalence seen in ribotype 027 (Inns et al. 2013).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
66
Appendix
Background to the mandatory surveillance of MRSA and MSSA bacteraemia
During the 1990s, voluntary surveillance of antibiotic resistance by microbiology
laboratories across England, Wales and Scotland saw increasing reports of MRSA
(Johnson, Pearson, and Duckworth 2005). In response to the concerns about the rising
number of reports of MRSA, the Department of Health and Social Care (DHSC) made
surveillance of MRSA bacteraemia mandatory from April 2001 (Department of Health
2003a).
Public Health England (PHE) has managed the mandatory surveillance on behalf of the
DHSC since the inception of the surveillance. The surveillance initially captured aggregate
counts of bacteraemia due to S. aureus and the number of those that were MRSA. The
results were reported every six months.
The frequency of data collection was increased to every quarter in 2004 and then monthly
in 2005. Enhanced surveillance of MRSA bacteraemia was introduced in October 2005,
allowing reporting of individual cases, rather than aggregated counts for a time period.
This had the advantage of providing information about the date of onset of bacteraemia,
relative to the date of admission and the department or specialty in which the patient was
being treated (Department of Health 2005).
In 2011, the surveillance was further expanded to include enhanced surveillance of MSSA
bacteraemia which had not shown the same decline in incidence that MRSA had.
Between 2013 and 2014, the outcome of a PIR could only assign a case to either a clinical
commissioning group (CCG) or an NHS acute trust. However, it was acknowledged that
this did not sufficiently recognise the complexity of some MRSA cases and from 1 April
2014, cases could be assigned to a third party. Third party cases include patients who may
be resident in England but have received care from some third party (i.e. neither the
reporting organisation nor the commissioning CCG), cases in patients not resident in
England or intractable cases (NHS England 2014b). The PIR process was halted in April
2018 when it became a local-only process (NHSI 2018).
Current surveillance of MRSA and MSSA bacteraemia allows reporting of cases through a
web-based data capture system (DCS) which provides the means to capture clinical
information on the patient and the infection, the likely source of infection and previous
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
67
healthcare interactions. The DCS also provides dynamic, on-demand reporting to acute
trusts, CCGs and other organisations allowing those involved in the care of patients to
investigate trends at a local level.
Mandatory surveillance of C. difficile infection
Voluntary surveillance of CDI by microbiology laboratories identified a steady increase in
the number of reports between 1990 and 2001. From 2002, this increase accelerated
rapidly (Department of Health 2007). In 2004, quarterly surveillance of CDI in patients
aged 65 years or above was made mandatory (Department of Health and Health
Protection Agency, 2005). This involved reporting aggregate numbers of cases to PHE on
the behalf of the DHSC. From 2007, the surveillance was expanded to include all patients
two or more years old and was extended to enhanced surveillance to capture patient-level
information (Department of Health 2018). Since 2007, rates of CDI have declined
dramatically.
In April 2017, prior healthcare exposure questions were introduced to the CDI mandatory
surveillance programme, to determine onset status of cases. This involved further
classification of cases in to the following onset statuses, hospital-onset healthcare
associated, community-onset healthcare associated, community-onset indeterminate
associated and community-onset community associated. Thus allowing the CDI
surveillance to be more comparable with international definitions.
Mandatory surveillance of Gram-negative bacteraemia
E. coli bacteraemia
Since the mid-2000s onwards, E. coli has been the most common pathogen causing
bacteraemia in England and has seen year-on-year increases. Given this sustained
increasing trend in the number of E. coli bacteraemia reports made through the voluntary
service, the DoH made reporting of E. coli bacteraemia mandatory in June 2011
(Department of Health 2011).
Klebsiella and P. aeruginosa bacteraemia
From April 2017, it became mandatory for acute trusts to report bacteraemias caused by
any species of Klebsiella bacteria and P. aeruginosa. This was introduced to support the
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
68
UK Government’s ambition to reduce Healthcare-associated Gram-negative bacteraemias
by 50% by 2023/24.
A note on terminology
In financial year 2018/19, mandatory surveillance stopped using the term Trust-
apportioned and started using the terms hospital-onset and community-onset instead. This
change was introduced due to the need to increase awareness of cases that occur in the
community. The algorithm for the separation of cases into hospital-onset/trust-apportioned
and those that are not has not changed. Clostridium difficile taxonomy classification has
been changed to Clostridiodes difficile.
Use of mandatory surveillance statistics
The data presented in this commentary, and in the accompanying data files and
infographics serve several purposes. First, they provide information to clinicians in trusts
about rates of bacteraemia and CDI in their organisation, helping to improve care and
infection control at their trust. Second, they provide information on the epidemiology of
these infections to clinicians, healthcare researchers and other interested parties,
identifying the likely sources of infection. Third, the information at CCG level allows
commissioners of care to understand healthcare-associated infection rates in their
community. Fourth, the national picture provides information to the DHSC, NHS England
and NHS Improvement regarding the infection rates across the country, and how these are
changing over time, while also providing information about where new interventions could
be targeted. Fifth, they are utilised by NHS Choices to assist in providing information to the
general public about healthcare-associated infection rates in their area and in facilities
where they might receive care.
Further information on data purpose, relevance and associated user-need can be found in
the Mandatory Health Care Associated Infection Surveillance Data Quality Statement.
Data included in the 2018/19 Annual Epidemiological Commentary (AEC)
Counts and rates of MRSA, MSSA, Gram-negative bacteraemias and CDI included in this
report are those with dates of specimens found to be positive within the period 1 April 2007
to 31 March 2019. This report includes data, extracted from the Healthcare Associated
Infections (HCAI) DCS on 17 April 2019, from 148 NHS acute trusts, 196 clinical
commissioning groups (CCGs), 14 specialised commissioning hubs and mapped to 14
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
69
NHS local offices. Data is published in line with organisational arrangements on the HCAI
DCS as of 31 March 2019. Data by acute trust and CCG is presented on an annual and
quarterly basis. Epidemiological commentaries are presented on an annual basis. Data
tables included in the Annual Epidemiological Commentary can also be found in
OpenDocument Spreadsheets (.ods) format on the Annual Epidemiological Commentary
web page.
This is the sixth annual publication to include MRSA bacteraemia data by PIR assignment
- FY 2013/14 to FY 2017/18. Since the PIR process became a local process at the start of
FY 2018/19, future reports will not contain the PIR data.
This publication forms part of the range of National Statistics outputs routinely published
by PHE. Epidemiological analyses included in this report are on an annual (financial year)
basis. Further epidemiological analyses by quarter can be found in our quarterly
epidemiological commentaries.
Data is also reported monthly for MRSA, MSSA, E. coli, Klebsiella spp., P. aeruginosa
bacteraemia and CDI.
We are always striving to ensure that routine outputs meet user need as much as possible.
If you have any suggestions for changes and/or additions please email
mandatory.surveillance@phe.gov.uk. Hyperlinks are highlighted in red.
E. coli bacteraemia
Table 1: Financial year counts and rates of E. coli bacteraemia aggregated cases and by
onset status (2012/13 to 2018/19).
Klebsiella spp. bacteraemia
Table 2: Financial year counts and rates of Klebsiella spp. bacteraemia aggregated cases
and by onset status (2017/18 to 2018/19).
P. aeruginosa bacteraemia
Table 3: Financial year counts and rates of P. aeruginosa bacteraemia aggregated cases
and by onset status (2017/18 to 2018/19).
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
70
MRSA bacteraemia
Table 4: Financial year counts and rates of MRSA bacteraemia aggregated cases
(2007/08 to 2018/19) and by onset status (2008/09 to 2018/19).
MSSA bacteraemia
Table 5: Financial year counts and rates of MSSA bacteraemia aggregated cases
(2011/12 to 2018/19) and by onset status (2011/12 to 2018/19).
C. difficile infections
Table 6: Financial year counts and rates of C. difficile infections bacteraemia aggregated
cases, by onset status (2007/08 to 2018/19), and by prior healthcare exposure (2017/18 to
2018/19).
Commentary
This document contains national and regional level (NHS England local office)
epidemiological commentaries for Gram-negative bacteraemias, MRSA and MSSA
bacteraemias and C. difficile infections. Mandatory surveillance data series included in this
report start from FY 2007/08 or the earliest full quarter of data collection.
For local office analyses presented in this report, CCGs and their attributed cases are
linked to the local offices of which they are part.1
However, CCGs only came into existence on 1 April 2013. In 2015, NHS regions were
reorganised from nine regions and 27 area teams into four regions and 14 local offices.
Analyses looking at time trends use retrospective attribution of cases to CCGs and local
offices. These may become less accurate the older the data is and; therefore, these have
only been performed back to FY 2009/10 reports. The temporal data before 2013/14
contained in this report with regards to CCGs and local offices has only been provided as
an indication of the trend over time for a given CCG/local office and thus, should be
treated with caution.
______________________ 1 Local offices can also be mapped via NHS acute trusts, based upon the local offices within which an NHS trust is
located, which is how the local offices included in tables 1-6 (to which this document is a supplementary commentary)
are derived.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
71
Alternative presentations of the annual data
Data included in this report is also available on https://fingertips.phe.org.uk/ ; this is a user-
friendly application that enables access to local data. It is ideal for both healthcare
professional and the general public alike. Fingertips enables access to these data in a
succinct format. Much of the mandatory surveillance data is available in graphical format,
facilitating easy understanding of key trends and geographical differences.
Please note that there is a one month delay between the publication of our statistics on
gov.uk and the publication of the same data on fingertips. For example, data published on
gov.uk at the start of April 2019 would be available on Fingertips in May 2019.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
72
Glossary
Average: Scientifically speaking, this is a measure of location. It is a way of describing
data and helps to distribute any inequalities in the data across the whole series. There are
three main mathematical measures which can be used to calculate an “average” value; the
mean, mode and median. Each of these methods has their own strengths and
weaknesses.
Bacteraemia: The presence of bacteria in blood
Bias: Is the systematic deviation of either results and/or inferences from the real situation.
Confidence interval (CI): Confidence intervals indicate the likely range in which an
estimated parameter (such as a mean or rate) is likely to fall. For most scientific studies, it
is impractical or impossible to measure every single member of a population and therefore
the true population mean cannot be determined. Instead, a representative sample is taken
and the sample mean is used as an estimate of the population mean. Although the sample
is intended to be representative, a different sample from the same population may provide
a different result simply by chance. A confidence interval, over unlimited repetitions of the
sample, should contain the true value of a parameter (such as the true population mean)
no less than its confidence interval. It is usual to calculate the 95% confidence interval.
That means that if we were to draw several independent, random samples from the same
population and calculate 95% confidence intervals from each of them, then 95% of such
confidence intervals would contain the true population mean. If we took 20 samples from
the same population and calculated 95% confidence intervals then 19 of 20 (95%) of these
95% confidence intervals would contain the true population meanwhile 1 of 20 (5%) will
not.
Denominator: The lower portion of a rate or ratio. This should reflect the population at risk
of developing a disease.
Epidemiology: Study of the occurrence and distribution of events (mostly health-related)
in a population.
Gram-negative: Class of bacteria that do not retain crystal violet stain as used as part of a
differential staining technique (called the Gram stain). The Gram stain is used as a way of
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
73
identifying bacteria and the difference in staining results are due to differences in the
bacterial cell wall, which has important implications for antimicrobial usage.
Incidence and incidence rate: New cases of a disease occurring in a study population.
An incidence rate is then the number of new cases that occur in a defined population in a
defined period of time.
NHS Local Office: An administrative unit of the NHS. NHS England has four
administrative regions; North of England, Midlands and East of England, London and
South of England. Below these four regions are 13 administrative geographies referred to
as Local offices.
Mean: The arithmetic mean is often what people think of when they say “average value”.
The mean is calculated by summing all of the values in a series (𝑎1 + 𝑎1 + ⋯ + 𝑎𝑛) and
then dividing by the number of values included in the series (𝑛). Mathematically, this is
described by the following formula:
mean =𝑎1 + 𝑎1 + ⋯ + 𝑎𝑛
𝑛
A real-world example would be if you wanted to calculate the mean amount spent on food
shopping over a four-week period (ie the average amount per week) having spent £51 in
week one, £59 in week two, £67 in week three and £52 in week four:
mean cost of food per week =£51 + £59 + £67 + £52
4= £57.25
Median: The median of a series of numbers is the mid-point of that series. This provides a
measure of an average value that is not overly affected by a few extreme values. The
median of the following set of numbers [1, 2, 3] is 2, while the median of the set of
numbers [1, 1, 1, 2, 10, 15, 16, 20, 100, 105, 110] is 15. To calculate the median value,
the set of numbers needs to be arranged in order of magnitude, the median is the number
that is exactly in the middle. If there is an even number of values in a set, then the median
value is the arithmetic mean of the two central values.
Mode: Is the most frequent value in a set of data (numbers or text values), for example, in
the following set of numbers [1, 1, 1, 2, 10, 15, 16, 20, 100, 105, 110] the mode is 1 as it
was included in the set three times, while the other numbers were only included once.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
74
Rate ratio: Is the ratio between two rates. For example, if the rate of MRSA bacteraemia
was 2 per 100,000 population in a year among men, and 4 per 100,000 population in a
year among women, the rate ratio would be 2.0. The rate would be two times higher
among women than men.
Secular trends: Changes over long periods of time.
Methods
Inclusion criteria for reporting to the surveillance system
MRSA bacteraemia
The following positive blood cultures must be reported to PHE, for the mandatory MRSA
surveillance: all cases of bacteraemia caused by S. aureus resistant to meticillin, oxacillin,
cefoxitin or flucloxacillin.
MSSA bacteraemia
The following positive blood cultures must be reported to PHE, for the mandatory MSSA
surveillance: all cases of bacteraemia caused by S. aureus which are susceptible to
meticillin, oxacillin, cefoxitin, or flucloxacillin ie not subject to MRSA reporting.
E. coli bacteraemia
The following E. coli positive blood cultures must be reported to PHE: all laboratory
confirmed cases of E. coli bacteraemia.
C. difficile infection
Any of the following defines a C. difficile infection in patients aged 2 years and above and
must be reported to the PHE:
• diarrhoea stools (Bristol Stool types 5-7) where the specimen is C. Difficile toxin
positive1
• toxic megacolon or ileostomy where the specimen is C. difficile toxin positive
1 DH/APRHAI have released guidance which incorporates C. difficile testing recommendations:
http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_133016.pdf
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
75
• pseudomembranous colitis revealed by lower gastro-intestinal endoscopy or Computed
Tomography
• colonic histopathology characteristic of C. difficile infection (with or without diarrhoea or
toxin detection) on a specimen obtained during endoscopy or colectomy
• faecal specimens collected post-mortem where the specimen is C. difficile toxin
positive or tissue specimens collected post-mortem where pseudomembranous colitis
is revealed or colonic histopathology is characteristic of C. difficile infection
Methods of reporting data on the HCAI Data Capture System (DCS)
The HCAI DCS is a web portal designed by PHE to facilitate the collection of the enhanced
data set.
Trusts using the website have access to all the data they have entered, which enables
them to assess their burden of these HCAIs. This can be compared to a regional and
national aggregate also available to trusts from the website. Clinical Commissioning
Groups (CCGs), Local Authorities, NHS Local Offices, and Directors of Public Health
(DPHs) are also able to register users, allowing them to access data specific to their
patients.
The dataset to be collected is described in the mandatory HCAI surveillance protocol
available on the DCS(pdf) and in the case capture user guide available on the same site
here(pdf) Case unlocks can be requested by reporting organisations using the process
described in the Unlock Requests User Guide. Revisions to data are covered by a data
specific revisions and correction policy(pdf).
An R package for working with data downloaded from the DCS can be found on github.
Deadline for entering data
All cases reported by the NHS with specimen dates during the previous month must be
entered onto the website by the 15 of the following month. The previous month’s data must
then be signed off by the trust’s Chief Executive Officer (CEO) by the 15th of every month.
For example, data concerning specimens collected in October must be entered and signed
off by 15 November.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
76
CCG attribution process
All cases of bacteraemia and C. difficile infection are attributed to a CCG, regardless of
their trust apportioning status or PIR assignment.
PHE’s HCAI DCS does not currently request NHS organisations to record patient CCG
details for any of bacteraemia or C. difficile infection cases. To obtain this data an extract,
comprising patient NHS number and date of birth are submitted to NHS Digital, via
Demographics Batch Services (DBS), on a daily basis to identify patient GP registration
details and patient residential postcode.
Overview of CCG attribution
The CCG for each case is attributed, in the following order:
• if patient’s GP practice code is available (and is based in England), the case will be
attributed to the CCG at which the patient’s GP is listed
• if the patient’s GP practice code is unavailable but the patient is known to reside in
England, the case is attributed to the CCG catchment area in which the patient resides
• if both the patient’s GP practice code and patient post code are unavailable or if a
patient has been identified as residing outside England, then the case is attributed to a
CCG based upon the postcode of the HQ of the acute trust that reported the case
Note that the retrospective attribution of cases to a CCG may become less accurate the
older the data are. Therefore, the data contained in this report for time periods prior to
2013/14 should be treated with caution and only used as an indication of the trend over
time for a given CCG.
Algorithms for apportioning cases
Please note that the algorithm applied for the apportioning of bacteraemia versus C.
difficile infection uses a different number of days between specimen collection and
admission to apportion cases; the principle is the same however. All cases of bacteraemia
and C. difficile infection are either trust-apportioned or non-trust-apportioned based on the
algorithms below (see also Appendix Figure A1). Previously, E. coli bacteraemia were not
trust-apportioned, however this changed in 2017 and cases were retrospectively
apportioned.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
77
It is not possible for PHE to change the apportionment of a case, as apportionment is
based on the data entered by the acute trust and the algorithm is applied to the entire
dataset not on a case by case basis; a case may only change from one category to
another if the relevant case details are incorrect and require amendment by the trust.
In addition to apportioning, all cases are also attributed to a CCG (see above). Thus, all
trust-apportioned and non-trust-apportioned cases will be attributed to a CCG. The
apportioning algorithm changed slightly with the launch of the new DCS. Prior to the
launch on 26/10/2015, there was no category for ‘Unknown’ patient location; the field was
simply left blank. An ‘Unknown’ patient location was introduced with the launch and new
cases from patients whose location was given as ‘Unknown’ at time of specimen were
trust-apportioned if the met the other criteria for trust-apportioned cases.
Bacteraemia
Hospital-onset:
Any NHS patient specimens taken on the third day of admission onwards (eg day 3 when
day 1 equals day of admission) at an acute trust (including cases with unspecified
specimen location) for Inpatients, Day-patients, Emergency Assessment, or unspecified
patient category.
Records with a missing admission date (where the specimen location is acute trust or
missing and the patient category is Inpatient, Day-patient, Emergency Assessment, or
unspecified) are also included.
Community-onset:
Any NHS patient specimens not apportioned to the above. This will typically include the
following groups: * any acute trust specimens taken on either the day of admission or the
subsequent day (eg days 1 or 2, where day 1 equals day of admission) * any specimens
from patients attending an acute trust who are not Inpatients, Day patients or under
Emergency Assessment (ie non admitted patients) * any specimens from patients
attending an identifiable healthcare location except an acute trust. This will typically
include GP, nursing home, non-acute NHS hospital and private patients
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
78
C. difficile
Hospital-onset:
Any NHS patient specimens taken on the third day of admission onwards (e.g. day 3 when
day 1 equals day of admission) at an acute trust (including cases with unspecified
specimen location) for Inpatients, Day-patients, Emergency Assessment, or unspecified
patient category. Records with a missing admission date (where the specimen location is
acute trust or missing and the patient category is Inpatient, Day-patient, Emergency
Assessment, or unspecified) are also included.
Community-onset:
Any NHS patient specimens not apportioned to the above. This will typically include the
following groups:
• any acute trust specimens taken on either the day of admission or the subsequent
day (e.g. days 1, 2 where day 1 equals day of admission)
• any specimens from patients attending an acute trust who are not Inpatient, Day-
patient or under Emergency Assessment (e.g. non-admitted patients)
• any specimens from patients attending an identifiable healthcare location except an
acute trust. This will typically include GP, nursing home, non-acute NHS hospital
and private patients.
Assignment of MRSA cases through the Post Infection Review
All NHS organisations reporting positive cases of MRSA bacteraemia from the 1 April
2013, were required to complete a Post Infection Review (PIR) until 1 April 2018.
A PIR was undertaken on all MRSA bacteraemias with the purpose of identifying how a
case occurred, to identify actions which will prevent a reoccurrence and to identify the
organisation best placed to ensure improvements are made (this is known as “assigning” a
case to an organisation).
Between 01 April 2013 and 31 March 2014 this was limited to either the NHS acute trust
who reported the case or the Clinical Commissioning Group with responsibility for
commissioning care for the patient; however, on 1 April 2014 an additional category was
included in the PIR process allowing for assignment to a Third Party. This provision was
made to acknowledge the increasingly complex nature of MRSA bacteraemias being
reported.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
79
Example of Third Party Assignment to a “Third Party” occurred through the arbitration
process for MRSA bacteraemias and has been available for any cases with an MRSA
positive blood culture post 1st April 2014. This category was designed to capture instances
where the MRSA case could not legitimately be assigned to either the trust or the CCG.
Therefore, for the purposes of the published data on MRSA cases, these Third-Party
cases were not be assigned to either the trust or the CCG. An example of “Third Party”
assignment is below.
Third Party Provider (England) Patient “A”
Background
A CCG in Berkshire commissions specialist services for Patient “A” from a London
specialist provider. After a few days, the patient returns to Berkshire Trust and is found to
test positive for MRSA bacteraemia. Since the sample was taken on the day of admission,
the Post Infection Review is led by the CCG. During the Post Infection Review process it
has been established that the patient had received no clinical care in Berkshire in the
immediate period prior to admission and that it is most likely that the bacteraemia
developed in trust A. In view of these facts, the CCG feels that the case should not be
assigned to them on the Data Capture System.
The matter is, therefore, referred to the arbitration panel led by the Regional Medical
Director and the Regional Director of Nursing.
Outcome
The arbitration panel agrees with the CCG and recommends that the case is assigned on
the Data Capture System to a Third Party. In the interests of patient safety, the CCG in
Berkshire (which commissioned the service) should inform the London provider to support
clinical learning and minimise the risk of a reoccurrence.
Further examples can be found in the PIR toolkit.
Prior trust exposure algorithm
In April 2017, the mandatory surveillance captured information on prior trust exposure, and
introduced specific prior healthcare categories, detailed below:
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
80
Hospital-onset, healthcare associated (HOHA) - Cases where the date of onset is two
or more days after the date of admission, where date of admission is day zero and the
patient was admitted to an NHS acute trust at time of specimen.
Community-onset, healthcare associated (COHA) - Cases which are not HOHA but
have previously been discharged from the reporting organisation within the 28 days prior
to specimen date.
Community-onset, indeterminate association (COIA) - Cases which are not HOHA but
have previously been discharged from the reporting organisation but have been
discharged from the reporting organization between 28 and 84 days prior to specimen
date.
Community-onset, community associated (COCA) - Cases which are not HOHA and
have not previously been discharged from the reporting organisation within the past 84
days.
Analysis of data
Time to onset calculations
To describe time to onset of an episode (bacteraemia or C. difficile) among inpatients, the
number of days between the date of admission to an NHS acute trust and the date of
positive specimen were calculated. This was only performed for patients who were admitted
to an acute trust (defined as either an inpatient, day patient or emergency assessment, ie
patients who should have an admission date to an acute trust) and for those whose
specimen was taken on or after the date of admission also at an NHS acute trust.
The number of days between the date of admission and the date of specimen can then be
described in two different ways; by grouping the number of days into meaningful
categories or by describing the “average”. Both have been provided in this report. As
mentioned in the glossary, there are three metrics which can be used to describe the
average value; the mean, median and mode. In this report, the median was used,
providing us with the central value of the number of days between date of admission and
date of positive specimen. The median was selected rather than the mean, because the
latter can provide spurious ‘average’ values if the data are skewed, ie if a few inpatients
had very long hospital stays before they had a C. difficile infection, then the mean value
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
81
would become much greater as it would be largely influenced by the value of the numbers
in the range.
Denominator data
Trust denominators NHS acute trust-level population data does not currently exist in
England as NHS acute trusts do not treat patients within defined geographical boundaries.
Therefore, a suitable proxy for population is required in order to calculate trust
apportioned/assigned rates. The occupied overnight beds (from the national KH03
dataset) provides the daily average overnight bed occupation for a specific time period; full
financial years for 2007/08 to 2009/10 and by quarter for financial years 2010/11 to
2017=8/19. This dataset is an open access return published by NHS England and
provides a measure of clinical activity in each trust, which is used as a proxy measure of
the patient population.
KH03 data can be found at NHS England’s website here.
Where data for trusts were missing, data for the same quarter in the preceding year were
used. These included the following trusts and periods:
Moorfields Eye Hospital NHS Foundation Trust (RP6) – KH03 data was missing for FY 2007/08. Last recorded data from 2006/07 were used as a proxy.
• Moorfields Eye Hospital NHS Foundation Trust (RP6) – KH03 data was missing for FY
2007/08. Last recorded data from 2006/07 were used as a proxy.
• The Rotherham NHS Foundation Trust (RFR) – KH03 data were missing for FY
2009/10, Q1 2010/11 to Q1 of 2011/12. Last recorded data from 2008/09 were used
as a proxy
• Sheffield Teaching Hospitals NHS Foundation Trust (RHQ) – KH03 data were missing
for Q1 2010/11 to Q1 2011/12. Data from 2009/10 were used as a proxy
• The Princess Alexandra Hospital NHS Trust (RQW) – Data were missing from Q1
2014/15. Data from Q1 2013/14 were used as a proxy. Data were missing from Q
2014/15 and Q3 2013/14 data were used as a proxy
KH03 data are published 6-8 weeks after the end of the relevant period. When this report
was produced, the final quarter of 2018/19 (i.e. January-March 2019) had not yet been
published. The KH03 data used in both the Annual Epidemiological Commentary and the
Annual Tables for 2018/19 has had this quarter replaced with the same quarter from the
previous financial year (i.e. KH03 data for 2018/19 includes data for April-June 2018, July-
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
82
September 2018, October-December 2018 and January-March 2018 in place of January-
March 2019). This is in line with our standard approach. In Annual Epidemiological
Commentaries published prior to 7 July 2016, April-June 2014 to October-December 2014
quarterly KH03 figures for one acute trust (RWD) had a percentage change of more than
20% compared to the previous quarter and the same quarter in the previous year. As a
result it was replaced with the KH03 data of the same quarter in the previous year (April-
June 2013 to October-December 2013).
However, PHE has reviewed its policy for processing KH03 data. All data irregularities
identified are now flagged with colleagues at NHS England (data owners of the KH03
dataset). Until we receive confirmation that any identified change in the occupied overnight
bed-days for an acute trust is anomalous, PHE will use the data as published in the KH03
dataset. This affects all reports published since 1 December 2015. In order for the KH03
data used to calculate rates included in this report to be consistent over the full-time
period, previously amended KH03 data for trust RWD for FY 2014/2015 has been altered
to reflect that published in the KH03 dataset. Please note that this could lead to slight
differences in trust-apportioned/assigned rates when compared with publications prior to 1
December 2015.
Other affected organisations and time periods include:
• Shrewsbury and Telford Hospital NHS Trust (RXW) Financial year 2009/10
• Imperial College Healthcare NHS Trust (RYJ) April-June 2012
Rate calculations
CCG rates
All cases are attributed to a CCG (see above) and using this data we calculate rates per
100,000 population for each CCG. Data at a CCG level can be scaled up to both NHS
Commissioning Board Local Office and a national level; therefore, to calculate rates for
CCGs, local offices and nationally the following equation is applied:
Rate, per 100,000 population = (𝑛 new cases attributed to CCG
Financial year population) × 100,000
We use the ONS mid-year population estimated data for relevant time periods, adjusted as
described above. For instance, for financial year 2010/11 mandatory surveillance data we
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
83
have used mid-year 2010/11 population estimates. Rates for CCG assigned MRSA cases
for 2013/14 onwards have also been calculated using the following equation:
Rate, per 100,000 population = (𝑛 new cases assigned to CCG
Financial year population) × 100,000
Trust rates
We calculate acute trust rates using trust-apportioned (or trust-assigned cases). The total
occupied bed days (KH03) data are used as an indicator of the total activity in each trust
during the relevant time period(s). Since 2010/11 KH03 have been published quarterly,
prior to this they were published on a financial year basis. The average daily overnight bed
occupancy for all acute trusts has been multiplied by the number of days in the relevant
time period. The relevant rate per 100,000 bed days was calculated as follows:
Rate, per 100,000 bed days = (𝑛 new cases reported by trust
Average daily occupancy × 𝑛 days in period) × 100,000
Prior to trust apportioning, all-reports rates were calculated per acute trust. Therefore, for
historical purposes to retain the time series, we also calculate an all-reports rate per acute
trust. “All reported cases” refers to all bacteraemias or C. difficile infections that were
detected by the acute trust that processed the specimen. It is important to note that this
does not necessarily imply that the infection was acquired there.
Healthcare associated infections in Wales, Scotland and Northern Ireland
Surveillance for C. difficile infections and MRSA and MSSA bacteraemias is also
performed in Wales, Scotland and Northern Ireland. Links to the relevant web pages are
as follows:
Wales: http://www.wales.nhs.uk/sites3/page.cfm?orgid=457&pid=25224
Northern Ireland: http://www.publichealth.hscni.net/directorate-public-health/health-
protection/healthcare-associated-infections-antimicrobial-resistanc
Scotland: http://www.hps.scot.nhs.uk/haiic/news/details?Id=23489
Please note that there are several differences between the English mandatory surveillance
systems and the systems run by the devolved administration, including case
definitions/protocols for diagnosing the infections, definitions re: inpatient episode vs. trust
apportioned/assigned episodes and the way in which data are presented. Therefore, the
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
84
data provided in the published reports from Public Health Agency Northern Ireland, Public
Health Wales and Health protection Scotland are not directly comparable with those data
published by Public Health England, found in this report and annual tables.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
85
References
Abernethy, J, Sherland JAP, and Hope R. 2017. “How Do the Epidemiology of Paediatric Methicillin-Resistant Staphylococcus Aureus and Methicillin-Susceptible Staphylococcus Aureus Bacteraemia Differ?” Journal of Medical Microbiology 66 (6). Microbiology Society: 737–43. Abernethy J, Guy R, Sheridan EA, Hopkins S, Kiernan M, Wilcox MH, Johnson AP, Hope R; E. coli bacteraemia sentinel surveillance group. 2017. Epidemiology of Escherichia coli bacteraemia in England: results of an enhanced sentinel surveillance programme. Journal of Hospital Infections. 365-375. Department of Health. 2003a. “Letter from the Chief Medical Officer: Surveillance of Healthcare Associated Infections. Department of Healt. London. 2003.” Department of Health. 2003b. Winning Ways: Working Together to Reduce Healthcare Associated Infection in England. Department of Health. 2005. “Mandatory Surveillance of methicillin resistant Staphylococcus aureus (MRSA) bacteraemias/ 9-6-2005. London, Department of Health.” Department of Health. www.dh.gov/en/Publicationsandstatistics/Lettersandcirculars/Dearcolleagueletters/DH_123275. Department of Health. 2007. “Saving Lives: reducing infection, delivering clean and safe care High Impact Intervention No 7 Care bundle to reduce the risk from Clostridium difficile.” https://webarchive.nationalarchives.gov.uk/20130124060604/http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_078126.pdf. Wiley E*, Smith C, Thakerar J, Wilson P. 2019. Preventing Escherichia coli bacteraemia through improved community urinary tract infection (UTI) management: use of the TARGET Uncomplicated UTI audit tool in primary care. Journal of Hospital Infection. Chief Medical Officer. 2008. “National target to reduce Clostridium difficile infections: SHA envelopes.” London. http://webarchive.nationalarchives.gov.uk/20130107105354/http://www.dh.gov.uk/en/Publicationsandstatistics/Lettersandcirculars/Dearcolleagueletters/DH_082702?IdcService=GET_FILE&dID=159244&Rendition=Web. Department of Health. 2018. “Letter from the Chief Medical Officer and Chief Nursing Officer: Infection caused by Clostridium difficile. London: Department of Health; 2005.” Department of Health.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
86
Department of Health, and Health Protection Agency. 2008. “Clostridium difficile infection: How to deal with the problem.” London: Department of Health, Health Protection Agency. https://www.gov.uk/government/publications/clostridium-difficile-infection-how-to-deal-with-the-problem. London: Department of Health. Department of Health. 2006. The Health Act 2006: Code of Practice for the Prevention and Control of Health Care Associated Infections. https://www.rdehospital.nhs.uk/docs/patients/services/housekeeping_services/Health%20Act%202006.pdf Dingle, K E., Didelot X, Phuong Quan T, Eyre DW, Stoesser N, Golubchik T, Harding RM, et al. 2017. “Effects of control interventions on Clostridium difficile infection in England: an observational study.” The Lancet Infectious Diseases 17 (4): 411–21. doi:10.1016/S1473-3099(16)30514-X. Harron, K, Parslow R, Mok Q, Tibby SM, Wade A, Muller-Pebody B, and Gilbert R. 2015. “Monitoring Quality of Care Through Linkage of Administrative Data: National Trends in Bloodstream Infection in Uk Picus 2003–2012.” Critical Care Medicine 43 (5). LWW: 1070–8. HM Government. 2019. Tackling antimicrobial resistance 2019-2024: The UK’’s five year national action plan. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/784894/UK_AMR_5_year_national_action_plan.pdf Inns, T., R.Gorton, A. Berrington, A. Sails, T. Lamagni, J. Collins, J. Perry, K. Hill, J. Magee, and K. Gould. 2013. “Effect of ribotype on all-cause mortality following Clostridium difficile infection.” Journal of Hospital Infection 84 (3): 235–41. doi:10.1016/j.jhin.2013.04.008. Johnson, A P, Pearson A, and Duckworth G. 2005. “Surveillance and Epidemiology of MRSA Bacteraemia in the UK.” Journal of Antimicrobial Chemotherapy 56 (3). Oxford University Press: 455–62. Department of Health. 2005. “The Delivery Programme to Reduce Healthcare Associated Infections (Hcai) Including Mrsa.” London: Department of Health. Reacher MH, Livermore DM, Shah A. 2000. Bacteraemia and Antibiotic Resistance of Its Pathogens Reported in England and Wales Between 1990 and 1998: Trend Analysis. BMJ. Vol. 320. 6. BMJ. Melzer, D., B. Tavakoly, R. E. Winder, J. A. Masoli, W. E. Henley, A. Ble, and S. H. Richards. 2015. “Much more medicine for the oldest old: trends in UK electronic clinical records.” Age Ageing 44 (1): 46–53.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
87
NHS England. 2014a. “Clostridium difficile infection objectives for NHS organisations in guidance on sanction implementation.” NHS England. http://www.england.nhs.uk/wp-content/uploads/2014/03/c-diff-obj-guidance.pdf. NHS England. 2014b. “Guidance on the reporting and monitoring arrangements and post infectionreview process for MRSA bloodstream infections from April 2014 (version 2). NHS England websit: NHS England; 2014.” NHS England. https://improvement.nhs.uk/documents/2513/MRSA_post-inf-guidance2.pdf. NHS England. 2013. “NHS England, Eveneryone Counts. Planning for patients 2013/14 (version 2). NHS England; 2013.” NHSE. NHSI. 2017. “Ambition to halve healthcare associated Gram-negative blood stream infections by March 2021. NHS Improvement 2017.” NHSIh. https://improvement.nhs.uk/documents/822/2017-03-15_Letter_-_GNBSI_edits.pdf. NHS Improvement. 2018. “NHS Improvement. Update on the reporting and monitoring arrangements and post infction review process foe MRSA bloodstream infection. 2018 (available at https://improvement.nhs.uk/documents/2512/MRSA_post_infection_review_2018_changes.pdf NHS Improvement. 2019. Preventing healthcare associated Gram-negative bloodstream infections (GNBSI). https://improvement.nhs.uk/resources/preventing-gram-negative-bloodstream-infections/ Organisation of Economic Co-operation and Development. 2018. “Length of Hospital Stay.” https://data.oecd.org/healthcare/length-of-hospital-stay.htm PHE. 2018. “30-day all-cause fatality subsequent to MRSA, MSSA and Gram-negative bacteraemia and C. difficile infections, 2017/18.” PHE. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/765014/HCAI_fatality_report_201718_v2.pdf. SepsisTrust. 2013. “Key Recommendations. 2013.” https://www.england.nhs.uk/wp-content/uploads/2015/08/Sepsis-Action-Plan-23.12.15-v1.pdf" Stone, Sheldon Paul, Christopher Fuller, Joan Savage, Barry Cookson, Andrew Hayward, Ben Cooper, Georgia Duckworth, et al. 2012. “Evaluation of the National Cleanyourhands Campaign to Reduce Staphylococcus Aureus Bacteraemia and Clostridium Difficile Infection in Hospitals in England and Wales by Improved Hand Hygiene: Four Year, Prospective, Ecological, Interrupted Time Series Study.” Bmj 344. British Medical Journal Publishing Group: e3005.
Annual epidemiological commentary. Gram-negative bacteraemia, MRSA bacteraemia, MSSA bacteraemia and C.
difficile infections up to and including financial year April 2018 to March 2019
88
Wilcox, M. H., N. Shetty, W. N. Fawley, M. Shemko, P. Coen, A. Birtles, M. Cairns, et al. 2012. “Changing epidemiology of Clostridium difficile infection following the introduction of a national ribotyping-based surveillance scheme in England.” Clinical Infectious Diseases 55 (8): 1056–63. doi:10.1093/cid/cis614
89