Epidemiology and surveillance of fungal infections: an overview

1
Epidemiology and surveillance of fungal
infections an overview

• David W. Warnock
• Centers for Disease Control and Prevention
• Atlanta, Georgia

2
Public Health
Healthcare

• Focus on
• individual
• diagnosis
• treatment

• Focus on
• - population
• prevention

3
The cycle of disease prevention
4
The cycle of disease prevention
5
The cycle of disease prevention
6
The cycle of disease prevention
7
Public health surveillance

• The ongoing systematic collection, analysis and
  interpretation of information about a disease1
• The reason for collecting, analyzing and
  disseminating information on a disease is to
  control that disease 2
• Collection and analysis should not be allowed to
  consume resources if action does not follow 2

1 Langmuir N Engl J Med 1963268182-92 2
Foege Int J Epidemiol 1976529-37
8
The spectrum of public health surveillance
Population
Case definition
Data collection
Sentinel surveillance
Passive surveillance
Syndromic surveillance
Population-based surveillance
Laboratory-based surveillance
Active surveillance
9
Comparison of surveillance systems

• Active surveillance
• Investigator-initiated
• Dedicated staff needed
• Extensive case finding performed
• Extensive clinical and laboratory information
• Audits performed
• High cost

• Passive surveillance
• Provider-initiated
• No dedicated staff needed
• Limited case finding performed
• Limited clinical and laboratory information
• No audits performed
• Lower cost

10
Number of reported cases of coccidioidomycosis
United States, 2007
2
77
24
3
12
3
72
68
2991
9
4832
23
3
0
Morbid Mortal Wkly Rep 2007 56 (no 53)
(published July 9, 2009)
11
Mycotic diseases passive surveillance

• Healthcare providers feel no need to report
  fungal infections since no immediate public
  health action is required
• Limitations of current diagnostic tests hinder
  the development of case definitions
• As a result these infections are under-diagnosed
  and under-reported

In 2007, CSTE adopted a modified case definition
for coccidioidomycosis a single positive
serologic test for IgG is now adequate for
definition of a case
12
Mycotic diseases surveillance case definitions

• A standardized case definition is needed to
  perform reliable surveillance for a disease
• In some diseases, a positive culture is
  indicative of colonization rather than infection
• Consensus case definitions for clinical trial
  enrollment of immunocompromised patients are too
  complicated for surveillance and not
  generalizable to other patient groups

13
Population-based surveillance

• Provides the most representative information on a
  disease in the entire population of a defined
  geographic location, and specific groups within
  that population
• All cases of the disease in the catchment area
  are identified, but only cases among residents
  are counted
• Incidence is calculated as the number of new
  cases occurring in a defined time period divided
  by the total population
• Active population-based surveillance to determine
  trends in disease incidence is expensive to
  conduct, and difficult to sustain for long periods

14
Incidence of Candida bloodstream infections(per
100,000 population)
7.1 1998-2000
7.1 1992-1993
6.0 1998-2001
24.0 1998-2000
25.0 2008-2009
8.7 1992-1993
14.0 2008-2009
Diekema et al. 2002 Hajjeh et al. 2004 Kao et
al. 1999
15
Incidence of Cryptococcus gattii infection
British Columbia, Canada
Average incidence 19992006 Vancouver Island
2.8 cases per 100,000 Mainland 0.65 cases
per 100,000
Cases per 100,000 population
Source BC Centre for Disease Control 2007
16
Estimated population-based incidenceof Candida
bloodstream, by race
Atlanta, GA, and San Francisco, CA 1992-1993
Cases per 100,000 population
Connecticut, and Baltimore, MD 1998-2000
Kao et al. 1999 Hajjeh et al. 2004
17
Population-based surveillanceCohort studies

• In cohort studies, the population is defined as a
  particular group of individuals (e.g. persons
  with AIDS or transplant recipients)
• Adequate follow-up is essential to determine the
  presence or absence of infection and therefore
  inclusion as a case of disease or non-case
• These studies are most useful when only a subset
  of the population is at risk for a particular
  disease
• Information is more broadly representative than
  reports from single centers

18
Incidence of invasive fungal infections after
stem cell transplant, 2001-2006 (TransNet)
12-month cumulative incidence ()
Kontoyiannis et al. 2009 submitted for
publication
19
Incidence of invasive aspergillosis after
allogeneic stem cell transplant, 2001-2005
(TransNet)
Overall incidence 1.6
12-month cumulative incidence ()
Transplant center
20
Sentinel surveillance

• Conducted at selected medical centers or sites,
  rather than in the entire population of a
  geographic location
• Total burden of disease in the general population
  cannot be estimated, but useful for diseases
  where the at-risk population is captured
• Less expensive and easier to perform than
  population-based surveillance

21
Sentinel surveillance

• Hospital are good sites for sentinel surveillance
  of invasive fungal infections because good
  denominators are available
• Useful for monitoring trends in incidence of
  particular pathogens and infections, species
  distribution, and antifungal drug resistance
• Site selection can be biased, so information may
  not be representative of the general hospital
  population

22
Incidence of Candida bloodstream infections (per
10,000 hospital admissions)
Sweden 3.2
Canada 4.0
France 2.0
United States 8.0
Spain 5.3
Italy 3.8
Brazil 24.9
Australia 2.1
Almirante et al. 2005 Colombo et al. 2005 Chen
et al. 2006 Hajjeh et al. 2004 Laupland et al.
2005 Tortorano et al. 2004
23
Administrative data as sources of information

• Hospital discharge and death records, based on
  ICD codes, are widely available, and permit
  application of common definitions to similar data
  from different institutions,
• Use of these data minimizes ascertainment bias
  when investigators use diverse methods for case
  finding
• Helpful for investigating long-term trends in
  disease incidence rates
• Limitations include diagnostic errors,
  inconsistent disease coding, and undetected
  duplicate reporting of cases

24
Comparison of the use of administrative data with
an active system for surveillance of invasive
aspergillosis in a single hospital, 2001-2005

• 64 of 1736 transplant recipients had ICD-9 codes
  consistent with IA, triggering medical record
  review 3 cases detected by other methods
• 48 of 67 patients reviewed had other or no
  infections, or had insufficient evidence to be
  classed as proven or probable IA
• 14 of 19 patients reviewed and classed as having
  IA identified by both methods 3 identified by
  active surveillance only 2 identified by ICD-9
  code only

Chang et al. Infect Control Hosp Epidemiology
2008 29 25-30
25
Incidence of zygomycosis in France based on
analysis of hospital records,1997-2006
Cases per 100,000 population
Bitar et al. Emerg Infect Dis 2009 15 1395-1401
26
Disease registries as sources of information

• Provide useful information about clinical
  details of rare fungal infections, or infections
  occurring in special hosts, such as transplant
  recipients
• Limited value for public health surveillance
  because meaningful and appropriate denominator
  data are not available
• Subject to ascertainment bias variable
  participation or case finding leads to
  unrepresentative results
• Should not be interpreted as being representative
  of broader populations

27
FoodNet Foodborne Diseases Active Surveillance
Network

• Established in 1996 as a collaboration among CDC,
  USDA, FDA, and state health departments
• Conducts population-based, active surveillance
  for laboratory-confirmed infections caused by 9
  pathogens commonly transmitted through food
• Campylobacter spp., Listeria monocytogenes,
  Salmonella spp., Shigella spp., STEC O157, Vibrio
  spp., Yersinia enterocolitica, Cryptosporidium
  spp., Cyclospora spp.

2009 10 sites, 45 million persons, 15 of
population
28
FoodNet trends E. coli O157 infections
Incidence per 100,000 population
1.30
National objective
0.90
Healthy People 2010 Objective 1.0 illness per
100,000 persons
29
Estimating the burden of disease

• Most surveillance systems do not capture the
  total burden of disease in a population
• This is because the reporting of a case depends
  on a number of steps the patient must visit a
  doctor the doctor must collect a sample the
  sample must be tested the pathogen must be
  identified and the test result must be notified
• The proportion of cases that are detected and
  reported differs from disease to disease

30
Food-related illness and death in the United
States

• Foodborne diseases cause 76 million illnesses,
  325,000 hospitalizations, and 5000 deaths in the
  United States each year

This article had been cited on 2350 occasions
through 9.21.09
Mead et al. Emerg Infect Dis 19995607-25
31
Estimating the global burden of HIV-associated
cryptococcosis

• Literature search for studies reporting estimates
  of incidence among HIV populations since 1996
• Median incidence for each UNAIDS geographic
  region multiplied by the HIV population to
  estimate number of cases
• To estimate number of deaths, assumed a 3-month
  case fatality rate
• 9 in high-income regions
• 55 in low- and middle-income regions
• 70 in Sub-Saharan Africa

Park et al. AIDS 2009 23 525-30
32
Estimated annual cases of HIV-associated
cryptococcosis
Eastern Europe Central Asia 27,200
Western Central Europe 500
North America 7,800
East Asia 13,600
North Africa Middle East 6,500
Caribbean 7,800
South South-East Asia 120,000
Sub-Saharan Africa 720,000
Latin America 54,400
Oceania 100
Global total 957,900 cases (range 371,700
1,544,000)
Park et al. AIDS 2009 23 525-30
33
Estimated annual deaths from HIV-associated
cryptococcosis
Eastern Europe Central Asia 15,000
Western Central Europe 45
North America 700
East Asia 1,200
North Africa Middle East 3,600
Caribbean 4,300
South South-East Asia 66,000
Sub-Saharan Africa 504,000
Latin America 29,900
Oceania 10
Global total 624,700 deaths (range 125,000
1,124,900)
Park et al. AIDS 2009 23 525-30
34
Estimated deaths in Sub-Saharan Africa from
cryptococcosis and other infectious diseases
WHO estimates
Excluding HIV AIDS
35
In conclusion

• Similar burden of disease estimates need to be
  developed for other fungal infections
• These estimates would allow for comparison with
  other diseases
• These estimates would help set public health and
  healthcare priorities, and determine resource
  allocations

36
Thank you for your attention
The findings and conclusions in this presentation
are those of the author and do not necessarily
represent the views of the Centers for Disease
Control and Prevention