Deaths Involving Clostridium difficile, England and Wales: 2012

• In 2012 there were 1,646 deaths involving Clostridium difficile (C. difficile) infection in England and Wales, 407 fewer than in 2011 (2,053 deaths)

• Mortality rates for deaths involving C. difficile fell for the fifth consecutive year and decreased significantly from 19.6 to 15.3 deaths per million population between 2011 and 2012

• The number of deaths involving C. difficile increases with age. During 2010–12 the mortality rate for people aged over 85 years was 818 and 810 per million for males and females respectively in England and Wales

• Deaths involving C. difficile accounted for 0.8 % of all hospital deaths in England and Wales during 2010–12 compared with 2.0 % during 2007–09

This bulletin presents the latest figures for deaths where Clostridium difficile (C. difficile) infection was mentioned on the death certificate by area, sex, age group and place of death, in England and Wales.

Figures are presented for 2012, with previously released figures for 1999 and 2001 to 2011 presented for comparison purposes. Information is also given about the context and use of the statistics, and the method used to produce them. The commentary will focus on England and Wales combined, however, figures for each country are provided separately in the reference tables.

Figures are presented for deaths registered rather than occurring in each calendar year. The median registration delay for deaths involving C. difficile is four days. Deaths certified by a medical practitioner are required to be registered within five days, therefore registration delays do not impact the figures in this bulletin (see background note 5).

Mortality rates for 2002–11 have been recalculated using revised mid-year population estimates which take account of the 2011 Census. Mortality rates may therefore differ from previously published figures (see Background note 7).

The number of death certificates in England and Wales mentioning C. difficile fell for the fifth consecutive year in 2012, from 2,053 (19.6 per million population) in 2011 to 1,646 (15.3 per million population) in 2012 (Figure 1).

Of the deaths where C. difficile was mentioned on the death certificate in England and Wales in 2012, 41% were deaths where C. difficile was the underlying cause (Table 2). This has decreased by one percentage point since 2011.

Age-standardised mortality rates for deaths involving C. difficile are similar for both sexes. In England and Wales rates were highest in 2007, at 85.9 and 80.9 per million for males and females respectively (Figure 3). Rates have since fallen each year and in 2012 they decreased to 15.9 per million population for males and 14.7 per million population for females. Most of the deaths involving C. difficile occur among older people. For both the 2007–09 and 2010–12 periods, mortality rates were highest in those aged 85 years and over at 2,461 and 812 per million population respectively in England and Wales (Table 4).

During the period 2010–12 deaths involving C. difficile accounted for 0.4% of all deaths in England and Wales, a fall from 1.2% of all deaths during 2007–09 (Reference table 5).

Between 2007–09 and 2010–12, the proportion of deaths involving C. difficile occurring in NHS hospitals in England and Wales decreased from 92.4 to 89.7% (Table 5).

Clostridium difficile (C. difficile) is a spore forming anaerobic bacterium that was first described in the 1930s (Hall and O’Toole, 1935). According to the Public Health England (2012), it is present in the gut of up to 3% of healthy adults and 66% of infants. However, C. difficile rarely causes problems in children or healthy adults, as it is kept in check by the normal bacterial population of the intestine. When certain antibiotics disturb the balance of bacteria in the gut, C. difficile can multiply rapidly and produce toxins which cause illness.

C. difficile infection ranges from mild to severe diarrhoea to, more unusually, severe inflammation of the bowel (pseudomembranous colitis). People who have been treated with broad spectrum antibiotics (those that affect a wide range of bacteria), people with serious underlying illnesses and the elderly are at greatest risk. Over 80% of C. difficile infections are in people aged over 65 years.

C. difficile is often referred to as a healthcare-associated infection (HCAI). HCAIs can develop either as a direct result of healthcare interventions such as medical or surgical treatment or from being in contact with a health or social care setting (including healthcare delivered in the community). However, HCAIS may be contracted outside a healthcare setting and brought in by patients, staff or visitors then transmitted to others (National Institute for Health and Care excellence, 2013).

C. difficile infection is usually spread on the hands of healthcare staff and other people who come into contact with infected patients, or with environmental surfaces contaminated with the bacteria or its spores (for example floors, bedpans and toilets). C. difficile produces spores when the bacteria encounter unfavorable conditions, such as being outside the body. These spores are very hardy and can survive on clothes and environmental surfaces for long periods.

Actions to reduce levels of healthcare associated infections have been detailed in various reports (Department of Health 2003a, Department of Health 2003b; Department of Health and Public Health England 2009; Department of Health 2013b; Public Health Wales 2004; Public Health Wales and Welsh Government 2009) and have been implemented in healthcare settings. These include reducing the infection risk from medical instruments, better antibiotic prescribing, isolating infected patients, environmental cleaning and disinfection, and improved hand hygiene.

In 2009, the Government brought in a new regulation regarding cleanliness and infection control to ensure that patients and staff are protected from the known risks of acquiring an HCAI. Trusts must meet this regulation to be registered with the Care Quality Commission, the independent regulator of health and adult social care in England (Care Quality Commission 2009).

Figures on the number of deaths from C. difficile in England and Wales are used by various organisations, including the Department of Health (DH), Public Health England (PHE), and Public Health Wales, for monitoring and evaluation purposes. They are also used by Clinical Commissioning Groups (CCGs), local health boards (LHBs) and individual healthcare establishments.

The Operating Framework for the NHS in England 2012/13 (Department of Health, 2011) states: `Protecting the safety of our patients is of paramount importance. The zero tolerance approach to all avoidable healthcare associated infections will continue. All NHS commissioners and providers should identify and agree plans for reducing MRSA bloodstream and Clostridium difficile infections in line with the national objectives’.

The NHS Outcomes Framework 2013/14 (Department of Health, 2013a) contains an indicator of C. difficile incidence. CCGs in England are required to reduce the number of C. difficile infections with organisations with higher baseline rates required to deliver larger reductions.

Since April 2007, it has been mandatory for all NHS trusts to report all cases of C. difficile infection in persons aged two years and over to Public Health England (in April 2013, the key functions of the Health Protection Agency were transferred into PHE). The latest figures show that a total of 14,684 cases of C. difficile occurring in patients aged two years and over were reported in England between April 2012 and March 2013 (Public Health England, 2013). This represents a reduction of 19% on the 18,022 cases of C. difficile reported in 2011/12 and a 32% reduction on the 21,707 cases reported in 2010/11.

In Wales, surveillance of C. difficile is managed by the Welsh Healthcare Associated Infection Programme (WHAIP), which is part of Public Health Wales. The latest figures for October 2011 to September 2012 (2011/12) show that there were 1,418 cases of C. difficile reported in hospital inpatients in Wales (Public Health Wales, 2013). This represents a reduction of 25% on the 1,900 cases reported over the same period in 2010/11.

Statistics for deaths involving C. difficile have been produced by the Office for National Statistics (ONS) for deaths registered in 1999 and from 2001 onwards (see ‘Methods’ below for more information). Figures for recent years show a large decrease in the number and rate of deaths, where C. difficile was the underlying cause of death or was mentioned anywhere on the death certificate among both males and females. This finding is consistent with the incidence figures reported by Public Health England and Public Health Wales. The decreases may be due to the actions taken to reduce healthcare associated infections described above.

The number of death certificates in England and Wales mentioning C. difficile fell for the fifth consecutive year in 2012, from 2,053 (19.6 deaths per million population) in 2011 to 1,646 (15.3 per million) in 2012 (see Table 2). This represents a reduction of 20% in one year.

Figure 1 shows that the number of deaths involving C. difficile increased between 2002 and 2007. The number of deaths rose rapidly between 2004 and 2007 increasing from 2,238 (23.3 per million population) in 2004 to 8,324 (82.9 per million population) in 2007, an increase in the mortality rate of over 250%. Since its peak in 2007, mortality rate from C. difficile has fallen by over 80%, and is now similar to the rate observed in 2002.

In England, the number of deaths where C. difficile was mentioned on the death certificate was highest in 2007 at 7,916 (83.9 per million population). This figure fell to 1,487 (14.7 per million population) in 2012, a fall in mortality rate of 82%. In Wales, the number of deaths involving C. difficile peaked a year later than in England at 461 (77.9 deaths per million) in 2008. This figure fell to 155 (23.5 per million) in 2012, a fall in the mortality rate of 70%.

Figure 2 shows that there was a steep decline in deaths involving C. difficile in England from its peak in 2007 to 2010. Wales has also exhibited a steep decline, albeit a more staggered one, from its peak in 2008 to 2012. Since 2010 the decline in mortality rates from C. difficile has begun to slow down in England, but not in Wales. There was a 40% decline in the mortality rate for England between 2010 and 2012 compared with a 60% decline in the mortality rate for Wales in the same period. This suggests that the rate for England may be stabilising whereas the rate for Wales is continuing to fall.

This decline in the number of deaths involving C. difficile since 2007 may have been caused by a reduction in the number of infections that is attributed to improved hygiene in hospitals. A recent study (Stone et al. 2012) found an association between the amount of soap and alcohol hand rub purchased by hospitals and the number of C. difficile infections.

In 2008 Public Health Wales set up a strategy, called 1000 lives, which aimed to improve hospital conditions and save 1000 unnecessary deaths by 2010. One outcome of this strategy, reported in ‘Two years to make a difference in Welsh healthcare’, was that there was a 30% increase in hand hygiene compliance, meaning that compliance with the hand hygiene regulation reached 90% across Welsh hospitals (Public Health Wales, 2010). This may be partly responsible for the reduction in C. difficile deaths in Wales since 2008.

Regional mortality rates

Regional mortality rates from Clostridium difficile tend to follow a similar trend to England as a whole. Mortality rates in most regions peaked in 2007, with the exception of the North East which peaked a year later in 2008 and the East Midlands which peaked a year earlier in 2006.

Between 2011 and 2012, the mortality rate per million population decreased for every region. The smallest decrease occurred in the South West with a fall of half a percentage point from 14.9 to 14.3; and the largest decrease occurred in the West Midlands with a fall of 8.2 percentage points from 27.5 to 19.2.

In 2012 the North East had the highest mortality rate out of all regions with 23.9 deaths per million population and the South East had the lowest rate with 10.7 deaths per million. Even though all regions have decreased since 2011, mortality rates in the North East, North West and West Midlands are all significantly above the national average for 2012.

The North East and North West have remained significantly higher than the rate for England over the last five years, and the West Midlands remained significantly higher over the last 10 years. Conversely, the East of England and the South East have remained significantly below the rate of England over the last five years.

Regional mortality rates (with 95% confidence limits) by sex are available for 2001 to 2012 in reference table 4.

Table 2 shows that, of the deaths where C. difficile was mentioned on the death certificate in England and Wales in 2012, 40.9% were deaths where C. difficile was the underlying cause, a fall of 0.9 percentage points since 2011, and the lowest proportion in the time series. The proportion of deaths where C. difficile was the underlying cause increased between 2002 (53.2%) until its peak in 2004/2005 (both 54.9%), this proportion has been generally decreasing since 2005.

The numbers of death certificates where C. difficile was mentioned, either as the underlying cause of death or as a contributing factor, are available for 2001 to 2012 in reference table 1.

Of the five most common underlying causes in deaths mentioning C. difficile, three appear in the top five leading causes of death of the general population: ischaemic heart disease, chronic lower respiratory diseases and cerebrovascular diseases. A high proportion of people with conditions such as ischaemic heart disease enter the healthcare system for treatment. In addition, the majority of deaths from these conditions occur in those aged over 65. These factors mean they are vulnerable to health-care associated infections like C. difficile.

Deaths with an underlying cause of certain cancers, for example malignant neoplasm of lymphoid, haematopoietic and related tissue, are more likely to be related to C. difficile infection compared to deaths in the general population. This is possibly due to a patient being at higher risk of becoming infected with the disease if they are already suffering from a disease or receiving treatments that compromise their immune response.

Diseases of the urinary system was the third most common underlying cause in deaths where C. difficile was mentioned, accounting for 3.3% of all C difficile deaths. Where as diseases of the urinary system only accounted for 1.9% of all deaths in England and Wales in 2012. This cause may be related to C. difficile due to the use of broad spectrum antibiotics to treat conditions such as recurrent catheter-associated urinary tract infections.

Rates have tended to by slightly higher in males than females in most years, with the largest difference occurring in 2008 where males were 6.4 percentage points higher than females with 62.7 compared to 56.3 deaths per million population.

Since 2009 there have been no significant differences in the rates between males and females. In England and Wales age-standardised mortality rates for deaths involving C. difficile have shown similar time trends for males and females. Rates peaked in 2007 for both sexes and have since decreased by 81 % from 85.9 to 15.9 per million population for males, and decreased by 82% from 80.9 to 14.7 per million population for females.

Most of the deaths involving C. difficile occur among older people. Table 4 shows that the mortality rate per million population increases with age. For both the 2007–09 and 2010–12 periods, mortality rates were highest in those aged 85 years and over (2,461 and 812 deaths per million population respectively in England and Wales). Older people tend to be more susceptible to C. difficile infection as their immune response is likely to be weaker. They are also more likely to be hospitalised, exposed to long-term care facilities or prescribed antibiotics, thereby increasing their chance of infection (Owens et al., 2008).

In comparison with those aged 85 years and over, the rates for persons aged between 75 and 84 years in England and Wales during both 2007–09 and 2010–12 were much lower, at 715 and 235 per million population respectively. Rates were lowest in the under 55 age group with just two deaths per million in 2007-09 and one death per million population in 2010-12 for both sexes in England and Wales.

Between 2007–09 and 2010–12 the mortality rate fell for all age groups and sexes. The mortality rates for persons, male and females aged 65 and over decreased by around two thirds each.

During the period 2010–12 deaths involving C. difficile accounted for 0.4% of all deaths in England and Wales, a fall from 1.2% of all deaths during 2007–09. Death certificates rarely specify the place where an infection was acquired, but the place of death is routinely recorded.

Around half of all deaths in England and Wales occur in hospital. It is therefore expected that most deaths involving C. difficile infection will occur in hospital. ONS does not routinely produce figures by individual hospital establishment.

During the period 2010–12, 89.7% of C. difficile deaths in England and Wales occurred in NHS hospitals. This represents 0.8% of all deaths that occurred in NHS hospitals. Between 2007–09 and 2010–12, the proportion of deaths involving C. difficile occurring in NHS hospitals in England and Wales decreased from 92.4 to 89.7% (Table 5).

In England and Wales, 6.3% of deaths involving C. difficile occurred in non-local authority care homes during 2010–12, this represents a 1.2 percentage point increase from 2007-09. Deaths from C. difficile account for 0.2% of all deaths which occurred in non-local authority care homes (see reference table 5).

More than 96% of all deaths involving C. difficile occur in care homes or NHS hospitals; other places such as at home, in a hospice or elsewhere only account for around 3-4% of these deaths.

Data for deaths involving C. difficile can be found in a Microsoft Excel workbook in the data section of this release.

The workbook contains the following results for England and Wales:

Number of death certificates with Clostridium difficile mentioned and as the underlying cause of death, England and Wales, deaths registered between 1999 and 2012

• Number of death certificates with Clostridium difficile mentioned and as the underlying cause of death, by annual registration quarter, England and Wales, deaths registered between 1999 and 2012

• Age-standardised mortality rates (with 95% confidence limits) for deaths where Clostridium difficile was mentioned on the death certificate, by sex, England and Wales, deaths registered between 1999 and 2012

• Age-standardised mortality rates (with 95% confidence limits) for deaths where Clostridium difficile was mentioned on the death certificate, by sex and region, England, deaths registered between 1999 and 2012

• Number of deaths where Clostridium difficile was mentioned on the death certificate by place of death, England and Wales, deaths registered in periods 2007–09 and 2010–12

• Number of deaths involving Clostridium difficile by sex and five-year age-group, England and Wales, deaths registered between 1999 and 2012

• Number of deaths involving Clostridium difficile by sex and five-year age-group, England, deaths registered between 1999 and 2012

• Number of deaths involving Clostridium difficile by sex and five-year age-group, Wales, deaths registered between 1999 and 2012

• Mortality rates (with 95% confidence limits) for deaths where Clostridium difficile was mentioned on the death certificate, by age and sex, England and Wales, deaths registered in periods 2007–09 and 2010–12

All deaths in England and Wales are coded by ONS according to the International Classification of Diseases (ICD) supplied by the World Health Organisation (WHO). In the Tenth Revision (ICD-10), used by ONS from 2001 onwards, there is a specific code (A04.7) for `Enterocolitis due to Clostridium difficile’. While this code identifies the vast majority of deaths involving C. difficile, a small number of C. difficile-related deaths are not captured by this code alone.

Since 1993 ONS has stored the text of death certificates on a database, in addition to all the ICD codes relating to causes identified on the death certificate. This means that it is possible to identify records where C. difficile is mentioned, but is not coded under the specific ICD-10 code.

In addition to extracting all deaths related to the specific A04.7 ICD-10 code, deaths mentioning a number of other ICD codes to which diseases including C. difficile could be coded were also extracted. The text of these death certificates was searched manually for mentions of Clostridium difficile, C. difficile or pseudomembranous colitis. The ICD-10 codes used to select deaths in order to search manually are shown in Box 1.

Deaths registered in 1999 were coded to both ICD–9 and ICD–10 as part of a special study to compare the two ICD revisions, and have therefore been used to give an additional year of data on deaths involving C. difficile.

Deaths with an underlying cause of C. difficile were identified by selecting those deaths with a mention of C. difficile that also had an underlying cause of one of the following ICD-10 codes: A04.7, A41.4 and A49.8. Death certificates that mention C. difficile and record the code A09 (Diarrhoea and gastroenteritis of presumed infectious origin) as the underlying cause of death, are also taken to indicate that C. difficile was the underlying cause of death.

Since 1986, ONS has used the internationally recommended death certificate for neonatal deaths (deaths of infants aged under 28 days). This means that these deaths cannot be assigned an underlying cause of death. However, as the data in this bulletin were based on deaths where C. difficile or pseudomembranous colitis were mentioned on the death certificate, neonates have been included. Deaths were extracted in the same way as described above for post-neonatal deaths (deaths of infants over 28 days but under one year old).

This statistical bulletin presents figures for C. difficile deaths in England and Wales. Clostridium difficile deaths in Scotland are published by National Records of Scotland and ‘Deaths registered with Clostridium difficile mentioned on the death certificate in Northern Ireland’ is published by the Northern Ireland Statistics and Research Agency (NISRA).