-
Is the disease infectious? (EC01)
Module: EPM301 Epidemiology of Communicable Diseases
Course: PG Diploma/ MSc Epidemiology
This document contains a copy of the study material located
within the computer assisted learning (CAL) session. The first
three columns designate which page, card and screen position the
text refers to. If you have any questions regarding this document
or your course, please contact DLsupport via [email protected].
Important note: this document does not replace the CAL material
found on your module CDROM. When studying this session, please
ensure you work through the CDROM material first. This document can
then be used for revision purposes to refer back to specific
sessions. These study materials have been prepared by the London
School of Hygiene & Tropical Medicine as part of the PG
Diploma/MSc Epidemiology distance learning course. This material is
not licensed either for resale or further copying.
London School of Hygiene & Tropical Medicine September 2013
v2.0
-
Section 1: EC02 Is the disease infectious? Aim To determine
whether a disease is caused by an infectious agent or not.
Objectives By the end of this session, you should be able to:
identify the factors that suggest that a disease has an infectious
cause use these factors to determine whether a disease of unknown
cause is likely to
be infectious. This session should take you about 2 to 3 hours
to complete. Section 2: Introduction
How can you know whether a disease is caused by an infection or
by exposure to some other factor? You may remember that Koch's
postulates list a number of criteria that can be used to decide
whether an infectious agent is responsible for a disease. Review
these criteria and think about what their limitations are. (You
might want to refer back to session FE13 to fully review the
subject of causality.)
Koch's postulates (1892) The agent must be consistently
demonstrable in diseased individuals The agent must be isolated
from a diseased individual and grown in pure culture Inoculation of
the agent must induce the disease experimentally
2.1: Introduction
Often it is not possible to fulfil Koch's postulates because no
infectious agent has yet been identified, or it is not possible to
culture a suspected agent, or there are no experimental hosts. In
these cases, there are a number of epidemiological characteristics
that might indicate an infectious aetiology.
-
In this session you will learn about some common features of
infectious diseases that can help you assess whether a disease is
infectious or not.
When might this be necessary?
Interaction: Hyperlink: aetiology: output (appears in new
window) Aetiology
Aetiology is often used in epidemiology to refer to "cause".
The search for an infectious cause may begin when: a new disease
syndrome emerges
Example button new knowledge about old diseases raises questions
about possible infectious causes
Example button new technology opens-up new areas of
investigation
Example button
Interaction: Button: Example: output (appears in new window) The
search for the causative agent of AIDS began when the syndrome was
recognised and became increasingly prevalent in the early
1980s.
Interaction: Button: Example: output (appears in new window) A
variety of infectious agents have now been associated with diseases
that were previously considered to be chronic conditions: human
papillomavirus and cervical cancer see
http://www.who.int/vaccine_research/diseases/viral_cancers/en/index3.html)
Helicobacter pylori and gastritis
parvovirus (see
http://www.cdc.gov/parvovirusB19/about-parvovirus.html) and
arthritis.
Interaction: Button: Example: output (appears in new window)
Example 1: Since the advent of molecular biological tools such as
the Polymerase Chain Reaction (PCR), it is possible to identify
pathogens from smaller quantities of tissue. New techniques like
Genome Wide Sequencing, Single Nucleotide
-
Polymorphisms (SNP) genotyping are now available for further
studies of genetic variability in various pathogens.
Example 2: New agents are continually being recognised through
advanced virological techniques (e.g. hepatitis C virus, hepatitis
E virus).
Section 3: Epidemiological characteristics of infectious
disease
What characteristics might suggest that a disease has an
infectious cause? We can consider the characteristics of infectious
aetiology under the following headings: Time when the disease
occurs Place where the disease is distributed b Person who develops
the disease On the next few pages each of these categories are
considered in turn.
Section 4: Time when the disease occurs How can time be used to
indicate an infectious aetiology? There are 4 main patterns, which
are described on this page:
1. Do disease cases occur close together in time? (Temporal
clustering) 2. Does disease occurrence change with the climate
during the year?
(Seasonality)
3. Are there periodic changes in disease occurrence? (Cyclical
patterns) 4. Is there a change in disease occurrence over long time
periods (more than one year)? (Long-term trends)
4.1: Time when the disease occurs 1. Temporal Clustering: Do
disease cases occur close together in time?
-
Temporal clustering is more easily recognised for rare diseases
or when there are few sources of infection. It is especially
important in identifying the causes of epidemics. Epidemiologists
have developed a 'moving window' test to distinguish between random
temporal clustering of cases and clustering due to an exposure.
This assesses the statistical probability of whether more cases
occur within a specified time interval than would be expected by
chance. By 'moving' the window continuously across the period of
observation you can sum the number of cases occurring during the
interval at different times.
4.2: Time when the disease occurs
1. Temporal Clustering: Do disease cases occur close together in
time? The 'moving window' test By sliding the 28-day 'window' along
the time-line opposite, you can see that the number of cases in the
window changes.
Use the buttons beneath the graph to move the window left and
right.
(buttons moves green line left or right) 4.3: Time when the
disease occurs 1. Temporal Clustering:
-
Do disease cases occur close together in time? Knox and
Lancashire (1982) used this method to show that pityriasis rosea a
mild skin disease occurred in temporal clusters.
They found a maximum of 16 cases within a 28-day window. This
was significantly higher than the average of 4.8 cases expected in
any 28-day interval from the overall data. This supported the
hypothesis of an infectious aetiology.
4.4: Time when the disease occurs
2. Seasonality :
Does disease occurrence change with the climate during the
year?
The reasons for seasonality of cases are unclear. Seasonality
may be associated with changes in temperature and humidity that
determine the survival and transmissibility of the infectious
agent, or the prevalence of vectors.
Behaviour can also vary with season, affecting the amount of
contact with potential sources of infection.
Interaction: Hyperlink: vectors: output (appears in new window)
Vector
A living carrier, such as an insect, that transports an
infectious agent from an infected individual to a susceptible
individual. Examples
Interaction: Tabs: Meningitis : output For meningococcal
meningitis, in which season would more cases generally occur, and
why might this be? Choose one of the seasons below:
Interaction: Button: temperate summer/tropical wet season:
output (appears in new window) In fact, the incidence of droplet-
or aerosol-transmitted diseases such as meningococcal meningitis
tends to peak in the temperate winter or tropical dry season for a
number of reasons. In temperate climates, there tends to be more
crowding and less ventilation in winter. This leads to increased
contact, re-circulation of air and dry nasal mucosa, providing a
better environment for transmission. In the tropics, the free
bacterial particles in droplets survive better at lower
temperatures, and dry nasal mucosa provides a better environment
for transmission.
Click the 'graph' button below to see an example.
Interaction: Button: temperate winter/tropical dry season:
output (appears in new window) That's correct. The incidence of
droplet- or aerosol-transmitted diseases such as meningococcal
meningitis tends to peak in the temperate winter or tropical
dry
-
season for a number of reasons. In temperate climates, there
tends to be more crowding and less ventilation in winter. This
leads to increased contact, re-circulation of air and dry nasal
mucosa, providing a better environment for transmission. In the
tropics, the free bacterial particles in droplets survive better at
lower temperatures, and dry nasal mucosa provides a better
environment for transmission.
Click the 'graph' button below to see an example.
Interaction: Button: graph: output (appears in new window)
This graph shows that the epidemic of meningococcal meningitis
started towards the end of the dry season when the weather was hot,
with the dry and dusty Harmattan wind blowing from the Sahara. The
incidence of the disease declined as the absolute humidity rose and
the epidemic stopped shortly after the wind stopped and the rains
began.
Interaction: Tabs: Malaria : output For malaria, in which season
would more cases generally occur, and why might this be? Choose one
of the seasons below: Interaction: Button: temperate
summer/tropical wet season: output (appears in new window) That's
correct, the incidence of vector-borne diseases such as malaria
tends to peak in the temperate summer/tropical wet season because
the mosquito vector abundance is dependent on warm temperatures and
water pools for breeding.
Interaction: Button: temperate winter/tropical dry season:
output (appears in new window)
-
In fact, incidence of vector-borne diseases such as malaria
tends to peak in the temperate summer/tropical wet season because
the mosquito vector abundance is dependent on warm temperatures and
water pools for breeding.
(back to main card) Interaction: Tabs: Shigella : output For
shigella (bacterial dysentery), in which season would more cases
generally occur, and why might this be? Choose one of the seasons
below: Interaction: Button: tropical wet season: output (appears in
new window) That's correct. The incidence of faecal-oral diseases
such as shigella tends to peak in the tropical wet season for a
number of reasons. Higher temperatures lead to increased
multiplication of bacteria in contaminated food, and there are
likely to be more flies contaminating food. There is an increased
possibility of water contamination, due to flooding, for example.
Interaction: Button: tropical dry season: output (appears in new
window) In fact, the incidence of faecal-oral diseases such as
shigella tends to peak in the tropical wet season for a number of
reasons. Higher temperatures lead to increased multiplication of
bacteria in contaminated food, and there are likely to be more
flies contaminating food. There is an increased possibility of
water contamination, due to flooding, for example. 4.5: Time when
the disease occurs
2. Seasonality :
Does disease occurrence change with the climate during the
year?
Some infections do not strictly follow these patterns. Rotavirus
is transmitted by the faecal-oral route, yet it is frequent in the
winter in temperate climates, but is detected all year round in the
tropics. This characteristic might support the hypothesis that it
is also spread by aerosol transmission (Cook et al 1990).
Interaction: Hyperlink: Rotavirus: output (appears in new
window) Rotavirus For more information about this infection, see
http://www.cdc.gov/vaccines/pubs/pinkbook/rota.html However, it is
likely that factors other than mode of transmission are also
involved in determining seasonality. Click below to see graphs that
show how different strains of the same virus can have different
seasonal patterns. The incidence of parainfluenza virus type 3
peaks in summer, while types 1 and 2 peak in winter (Noah
1989).
Interaction: Hyperlink: parainfluenza virus: output (appears in
new window) Parainfluenza virus
-
For more information about this infection see
http://www.cdc.gov/vaccines/pubs/pinkbook/rota.html.
Interaction: Button: Graph: output
4.6: Time when the disease occurs 2. Seasonality :
Does disease occurrence change with the climate during the year?
When looking for seasonality it is important to take into account
changes in population size. For example, seasonal population
migration or seasonal variation in birth rates (for a disease of
infants) could falsely indicate a seasonally-transmitted disease
because the number of cases might increase as the population size
increases. Even if seasonality of the disease exists, there are a
number of alternative aetiological explanations. Can you think of
non-infectious exposures that might also vary seasonally?
Interaction: Button: Cloud: output (appears in new window)
Nutritional status varies seasonally, especially in poorer
countries or communities (e.g. Kigutha et al 1995). Exposure to
toxins has also been shown to be associated with season, for
example aflatoxin in West Africa (e.g. Wild et al 2000). You may
have thought of other examples.
Interaction: Hyperlink: aflatoxin: output (appears in new
window) Aflatoxin
-
A metabolic product of fungi of the Aspergillus spp. that grow
on cereals and grains. It has been identified as a risk factor for
liver cancer. 4.7: Time when the disease occurs 3. Cyclical
patterns:
Are there periodic changes in disease occurrence? What other
periodic events might influence the temporal pattern of a
disease?
Interaction: Tabs: Example 1 : output Periodic visits to the
market to sell goods may increase the potential for contact with
infected individuals at particular times.
Seasonal movements from the highlands to the malaria-endemic
lowlands for farming can expose non-immune individuals to malaria
resulting in a peak of malaria cases at harvest time.
Interaction: Hyperlink: malaria: output (appears in new window)
Malaria
For more information about this disease see
http://www.cdc.gov/malaria/. (back to main card)
Interaction: Tabs: Example 2 : output Annual increases in
measles transmission have been shown to coincide with the start of
school terms, which provide an ideal opportunity for contact
between infected and susceptible children (e.g. Fine & Clarkson
1982).
Interaction: Hyperlink: measles: output (appears in new window)
Measles
For more information about this disease, see
http://www.cdc.gov/measles/index.html. 4.8: Time when the disease
occurs 3. Cyclical patterns:
Are there periodic changes in disease occurrence? In addition to
the annual cycles, larger peaks of measles cases occurred at
regular 2-year intervals in England and Wales before the national
immunisation programme was introduced.
-
Click the 'plot' button below to see a plot that illustrates
this.
Interaction: Button: Plot: output (appears in new window)
Can you think of a reason for this 2-year cycle? Bear in mind
that measles infection confers lifelong immunity.
Interaction: Button: Cloud: output For infections that confer
lifelong immunity, it may take a number of years until there are
sufficient numbers of susceptible individuals born into the
population to enable the infection to circulate widely.
This "critical threshold of susceptibles" will be discussed in
session EC06. 4.9: Time when the disease occurs 3. Cyclical
patterns:
Are there periodic changes in disease occurrence? Changes in
climate may occur over longer periods of time. These can also cause
variations in disease occurrence. The El Nio_ Southern Oscillation
affects weather patterns in certain parts of the world at
approximate intervals of 4 to 6 years.
-
It shows a close association with the between-year variation in
malaria incidence in South America and the Indian subcontinent.
This association is due to periods of intense drought or rainfall
affecting the density of malaria mosquito vectors
Interaction: Hyperlink: El Nio: output (appears in new window)
El Nio
The El Nio Southern Oscillation is a periodic climatic
phenomenon associated with a warmer than average sea surface
temperature in the eastern equatorial Pacific Ocean. 4.10: Time
when the disease occurs
4. Long-term trends Is there a change in disease occurrence over
long time periods (more than one year)? Long-term trends in a
disease are useful if they can be associated with changes in the
proposed mode of transmission. For example, the trend in mortality
from cervical cancer in England and Wales was correlated with the
trend in gonorrhoea incidence at the age of 20 for the same birth
cohorts of women. Click below to see a graph of this trend.
This provided strong evidence that cervical cancer was a
sexually transmitted disease. It has subsequently been associated
with human papilloma virus infection (Beral 1974).
Interaction: Hyperlink: gonorrhoea: output (appears in new
window) Gonorrhoea
For more information about this disease, see
http://www.cdc.gov/std/Gonorrhea/
Interaction: Hyperlink: birth cohorts: output (appears in new
window) Birth cohorts
Cases were grouped by year of birth, not year of onset.
-
Section 5: Place where the disease occurs
What spatial distributions can be characteristic of infectious
diseases? 1. Do cases occur close together within a specified area?
(Spatial Clustering) 2. Does the disease tend to occur only in
specific geographical locations?
(Geographical Restriction) 5.1: Place where the disease
occurs
1. Spatial Clustering:
Do cases occur close together within a specified area? Spatial
clustering is useful to identify a common source of infection.
Can you remember an example of this from the epidemiological
work of John Snow?
Interaction: Button: cloud: output (appears in new window)
-
John Snow plotted the residence of fatal cases of cholera
http://www.cdc.gov/cholera/index.html on a map to identify the
water pump responsible for the cholera epidemic in Soho, London in
1854. Review this from FE01. (appears on RHS) More recently, the
cases of new variant Creutzfeld-Jacob disease were analysed to
assess whether more cases lived near to meat rendering plants than
would be expected by chance. There was no evidence that the
distance from residence to the nearest rendering plant was a risk
factor for the disease (Cousens et al 1999).
Click below to see the map from this study. Show Button
Interaction: Hyperlink: new variant: output (appears in new
window) New variant Creutzfeld-Jacob disease For more information
about this disease see
http://www.hpa.org.uk/Topics/InfectiousDiseases/InfectionsAZ/CreutzfeldtJakobDisease/
Interaction: Hyperlink: Creutzfeld-Jacob disease: output
(appears in new window) New variant Creutzfeld-Jacob disease
For more information about this disease see
http://www.hpa.org.uk/Topics/InfectiousDiseases/InfectionsAZ/CreutzfeldtJakobDisease
Interaction: Hyperlink: rendering plants: output (appears in new
window) Rendering plants are factories involved in the production
of meat and bone meal from animal carcasses.
Interaction: Button: Show: output (appears in new window)
-
5.2: Place where the disease occurs
1. Spatial Clustering:
Do cases occur close together within a specified area?
It is important to remember that spatial clustering of cases can
also indicate an environmental (non-infectious) cause. A
point-source exposure, such as a toxic contaminated water source,
would also produce a geographical cluster of cases.
5.3: Place where the disease occurs 1. Spatial Clustering:
Do cases occur close together within a specified area? Household
clustering of cases can be a good indicator of infectious
aetiology. Think about which non-infectious causes could also
explain household clustering.
Interaction: Button: cloud: output (appears in new window)
Individuals from the same family and household are likely to be
exposed to the same genetic and environmental factors, such as
socio-economic conditions, diet, and air pollution. Any of these
could provide an alternative, non-infectious explanation for the
disease.
-
Measures of household spread of infection will be discussed in
EC03. 5.4: Place where the disease occurs
2. Geographical restriction: Does the disease tend to occur only
in specific geographical locations? Climatic and ecological
conditions are responsible for the geographical restriction of
infectious diseases. Certain conditions favour the development of
the infective agent, transmission of the infection, and breeding of
vectors.
Examples
Interaction: Tabs: 1 : output Schistosomiasis is distributed in
tropical and sub-tropical areas, because optimal survival of the
parasite in water, infectivity to the intermediate snail host, and
development in the snail occurs at 25_. Click below to see a map of
the world distribution of this disease.
http://www.cdc.gov/parasites/schistosomiasis/. Map Button
Interaction: Hyperlink: Schistosomiasis: output (appears in new
window) Schistosomiasis
For more information about this disease see . (Back to main
card)
Interaction: Button: Map: output (appears in new window)
-
Interaction: Tabs: 2 : output Sandflies, and therefore the
Leishmania parasitic infection that they transmit, are distributed
in the tropics and subtropics, and are extending into southern
European climates.
Click below to see a map of the world distribution of this
disease. Map Button
Interaction: Hyperlink: Leishmania: output (appears in new
window) Leishmania
For more information about this infection see
http://www.cdc.gov/parasites/leishmaniasis/index.html (Back to main
card)
Interaction: Button: Map: output (appears in new window)
-
Interaction: Tabs: 3 : output Burkitt's lymphoma is
geographically restricted to the lowlands of Africa, despite a
wider distribution of Epstein-Barr virus, which is implicated in
the disease. The association with areas of high rainfall and warm
temperatures suggests that malaria might play a role in the
aetiology of the disease (de Th 1979). Map Button
Interaction: Hyperlink: Leishmania: output (appears in new
window) Burkitt's lymphoma
For more information about this disease.
Interaction: Button: Map: output (appears in new window)
Burkitt's lymphoma and climate in Africa
-
The red overlay on this map shows the distribution of Burkitt's
lymphoma. The unshaded areas of the map beneath represent regions
of high rainfall and high mean temperature (in the coolest month).
5.5: Place where the disease occurs
2. Geographical restriction: Does the disease tend to occur only
in specific geographical locations?
There are a number of characteristics that can help to associate
a disease with a particular place. They are listed opposite.
1. The disease is equally common in all ethnic groups that live
in the area; 2. the disease is much less common among people from
similar groups that live elsewhere;
-
3. individuals migrating into the area become ill with the same
or a greater frequency; 4. individuals migrating out of the area do
not become ill with the same
frequency.
Section 6: Person who develops the disease How might particular
groups of individuals be at greater risk of infection and disease?
Some characteristics of particular groups may increase their
exposure to infection or facilitate the onset of the disease.
Groups to consider are:
1. Occupational
2. Behavioural
3. Socio-economic
4. Immunological status
6.1: Person who develops the disease 1. Occupational:
A number of infections are associated with occupational
exposure. Occupation often acts to bring a specific group of
individuals in contact with infection as a direct result of the
activity. For example, butchers are more likely to develop warts.
(Keefe et al 1994).
Interaction: Hyperlink: warts: output (appears in new window)
Warts
For more information about this disease see www.cdc.gov/hpv. In
the same way, it might be expected that cattle farmers in the UK
should be more likely to develop new-variant Creutzfeld-Jakob
disease because of their increased contact with cattle that have
Bovine Spongiform Encephalitis. So far there have been too few
cases to assess this risk. Interaction: Hyperlink: warts: output
(appears in new window) New-variant Creutzfeld-Jakob disease
For more information about this disease see
http://www.hpa.org.uk/Topics/InfectiousDiseases/InfectionsAZ/CreutzfeldtJakobDisease/.
-
1. Occupational: Occupational groups can also be at greater risk
of disease because of associated behaviours. In the tropics,
illegal night-time fishing and logging activities increase outdoor
exposure to mosquito bites and therefore increase the risk of
malaria infection.
Long-distance truck drivers tend to have a more promiscuous
sexual behaviour, leading to an increased risk of HIV infection
(Bwayo et al 1994).
6.2: Person who develops the disease 2. Behavioural:
Certain lifestyle or cultural behaviours may increase the risk
of an infectious disease by increasing the likelihood or amount of
exposure to the infectious agent. This aspect may be less easy to
investigate, for reasons of cultural sensitivity or privacy, for
example sexual behaviours.
Examples 1. Pneumocystis pneumonia and Kaposi sarcoma in
homosexuals
2. Kuru and cannibalism
Interaction: Hyperlink: Pneumocystis pneumonia: output (appears
in new window) Pneumocystis pneumonia
For more information about this disease see
http://www.cdc.gov/fungal/pneumocystis-pneumonia/. Interaction:
Hyperlink: Kaposi sarcoma: output (appears in new window) Kaposi
sarcoma
For more information about this disease see
http://www.cancer.gov/cancertopics/pdq/treatment/kaposis/HealthProfessional
Interaction: Hyperlink: Kuru: output (appears in new window)
Kuru
For more information about this disease see Collinge et at
(2006) http://dx.doi.org/10.1016/S0140-6736(06)68930-7
-
6.3: Person who develops the disease 3. Socio-economic: Many
infectious diseases are related to socio-economic status. This is
often related to living-conditions that can affect transmission,
such as levels of crowding, poor-sanitation, dampness and
house-structure.
For example, tuberculosis is frequently associated with
overcrowded and low-income households (Mangtani et al 1995).
Interaction: Hyperlink: tuberculosis: output (appears in new
window) Tuberculosis
For more information about this disease see
http://www.cdc.gov/tb/. Can you think of non-infectious causes of
disease that might also be associated with socio-economic
status?
Interaction: Button: cloud: output (appears in new window)
Malnutrition and toxic exposure could both cause disease and are
also likely to be related to socio-economic status. You might have
thought of other exposures. 6.4: Person who develops the disease 4.
Immunological status: High incidence of a disease among people
lacking immunity or with an impaired immunological system is highly
suggestive of an infectious cause.
Individuals undergoing therapy with immunosuppressive drugs or
suffering from an immunodeficiency are more susceptible to
infectious agents.
Diseases among these groups are likely to be of an infectious
aetiology, e.g. the increased incidence of lymphomas in transplant
patients. 6.5: Person who develops the disease 4. Immunological
status: Migrants moving from an area of low to high endemicity will
be at greater risk of being infected and developing the disease, as
they have not developed a protective immunological response.
Example
-
Studies of migrants have shown that there is a critical age at
exposure for multiple sclerosis, after which an individual is less
likely to develop the disease (for example, see Gale & Martin
1995).
Interaction: Hyperlink: multiple sclerosis: output (appears in
new window) Multiple sclerosis
Multiple sclerosis (also called disseminated sclerosis) is a
chronic, often disabling disease of the central nervous system,
characterised by impairment of transmission of nerve impulses,
particularly those involved with vision, sensation, and the use of
limbs. There appear to be multiple causes, possibly including
viruses and environmental, genetic, and immune system factors.
Section 7: Interactions
Often, identifying the cause of a disease is not
straightforward. Interactions can occur between any of the factors
already mentioned.
The main areas to consider are: Are cases clustered in both
space and time? (Spatial-temporal clustering) Does the incidence
vary by geographical region? (Regional variation) Are there other
factors that are necessary or seem to pre-dispose an individual to
disease? (Co-factors)
7.1: Interactions 1. Spatial-temporal clustering:
Are cases clustered in both space and time? Spatial-temporal
clustering occurs because of person-to-person transmission, and can
be investigated in a number of ways.
One way is to pair each individual with every other individual
who has the disease. Distance in time and geographical distance
between each possible pair can be plotted against each other to
look for any correlation. By considering particular intervals of
time and distance, a chi-squared test can be used to test for any
association (Messenger et al 1982).
This method has been adapted to study diseases with long latent
periods. Another adaptation involves recording all significant
contacts between cases of a disease, and between appropriate
controls - a difficult task! (Pike & Smith 1974)
-
7.2: Interactions
2. Regional variation: Does the incidence vary by geographical
region?
Regional variations in infectious diseases are likely to be due
to an interaction between geographical, socio-economic,
immunological, dietary and cultural factors. Genetics and other
environmental exposures also vary by place and it is important to
exclude these as possible causes before specifying an infectious
aetiology. Some infections are limited to particular countries. In
the case of political boundaries, these variations in disease
incidence are usually due to socio-economic factors such as the
degree of financial investment in infrastructure (e.g. sanitation
systems) and public health measures (e.g. immunisation
programmes).
7.3: Interactions
3. Co-factors: Are there other factors that are necessary or
seem to pre-dispose an individual to disease?
In the case of a number of malignant diseases, the infective
agent may not be a sufficient cause of disease. There might be a
sequence of epidemiological, immunological and molecular events
that are additionally required to cause the disease.
Often, external (environmental) and/or internal (genetic and
physiological) factors also play a necessary role. The involvement
of these 'co-factors' makes the search for an infectious aetiology
more difficult.
Interaction: Hyperlink: sufficient cause: output (appears in new
window) Sufficient cause refers to a particular exposure providing
the full explanation for an outcome. In this case, infection alone
is not enough to cause disease.
Interaction: Tabs: Example 1 : output It is suspected that
Hodgkin's disease occurs in only a proportion of individuals
infected with Epstein-Barr virus because of some genetic or
environmental co-factor (Stiller 1998).
Interaction: Hyperlink: Hodgkin's disease: output (appears in
new window) Hodgkin's disease
For more information about this disease see
http://www.who.int/vaccine_research/diseases/viral_cancers/en/index1.html
-
(Back to main page)
Interaction: Tabs: Example 2 : output Burkitt's lymphoma (BL) is
a tumour whose epidemiological characteristics strongly suggested
Epstein-Barr virus (EBV) as the cause. However, EBV is widespread
and the restricted geographical distribution of BL suggested that
malaria might be involved. A higher incidence of BL was found among
migrants with lower immunity to malaria. This was consistent with
malaria being the major co-factor necessary for the disease
(de-Th_1979). Interaction: Hyperlink: Burkitt's lymphoma: output
(appears in new window) Burkitt's lymphoma
For more information about this disease see
http://www.who.int/vaccine_research/diseases/viral_cancers/en/index1.html
7.4: Interactions
Over the next few pages you will complete an exercise to
determine whether or not a disease of unknown aetiology is
infectious. You may like to take a break at this point, before
going on to complete the exercise.
Section 8: Exercise: A disease of unknown aetiology
This exercise demonstrates how the epidemiological pattern of a
disease can be defined by morbidity surveys. Using the information
you have covered earlier in this session, interpret the data
provided to decide whether the disease under investigation is
infectious or not. Clinical information is not given, to show you
that the answer can be reached by epidemiological assessment alone.
On the following pages, you will be shown a number of tables.
Examine each table and summarise the epidemiological features that
you think are most important. Consider the information from each
table separately, and then evaluate all the information together to
determine the aetiology of this disease.
8.1: Exercise: A disease of unknown aetiology
Background The disease was recognised in the early 1900s and
could be diagnosed with reasonable accuracy by experienced
physicians. It was thought to be associated with poor and
unsanitary living conditions.
-
Cases appeared to be localised in their geographical
distribution. Multiple cases were frequently observed in the same
family and recurrent attacks were common. A number of hypotheses
were proposed for the aetiology of the disease. These included
infection, genetic susceptibility, toxicity, etc.
8.2: Exercise: A disease of unknown aetiology To collect data on
this disease in a defined population, a responsible agency
undertook an extended survey. This was done in an area in the
Northern Hemisphere where there had been a high incidence of the
disease over a period of years. The data presented in this exercise
were collected from 24 villages during a one-year period. The name,
age, sex and marital status of each family member were recorded for
all individuals in the villages surveyed. Each household was
visited every two weeks. Reported symptoms and a clinical
examination were used to determine the occurrence of 'the disease'.
Questionable cases were referred to a more experienced physician
who was directing the study. Note: The number of cases is the same
in the first three tables but the population size is inconsistent,
leading to different estimates of disease risk. The reason for this
is unclear, but may be due to missing data on age and sex.
Section 9: Exercise: Table 1
Table 1 opposite shows the total number of cases and risk of
'the disease' in each village during the survey year. Summarise the
data in Table 1: The village population size ranges from Calc 1 to
Calc 2. The minimum number of cases in one village is Calc 3. The
maximum number of cases in one village is Calc 4 The lowest risk of
disease is Calc 5 per 1000, and the highest risk of disease is Calc
6 per 1000.
Interaction: Calculation: Calc 1 Correct response: Correct
That's right, the smallest population size is 284 individuals in
village 'Fn'.
-
Incorrect response: Sorry, that's not right. The smallest
population size is 284 individuals in village 'Fn'.
Interaction: Calculation: Calc 2 Correct response: Correct
That's right, the largest population size is 1569 individuals in
village 'Ola'. Incorrect response: Sorry, that's not right. The
largest population size is 1569 individuals in village 'Ola'.
Interaction: Calculation: Calc 3 Correct response: Correct
That's right, the minimum number of cases is 13, in village
'Gr'.
Incorrect response: Sorry, that's not right. The minimum number
of cases is 13, in village 'Gr'.
Interaction: Calculation: Calc 4 Correct response: That's right,
the maximum number of cases is 119, in village 'Rc'. in village
'Rc'. Incorrect response: Sorry, that's not right. The maximum
number of cases is 119, in village 'Rc'.
Interaction: Calculation: Calc 5 Correct response: Correct
That's right, the lowest risk of disease is 19.6 per 1000, in
village 'Gr'. Incorrect response: Sorry, that's not right. The
lowest risk of disease is 19.6 per 1000, in village 'Gr'.
Interaction: Calculation: Calc 6 Correct response: Correct
The highest risk of disease is 100.8 per 1000, in village 'In'.
Incorrect response: Sorry, that's not right. The highest risk of
disease is 100.8 per 1000, in village 'In'.
-
9.1: Exercise: Table 1
-
RHS remains static (chart) Which epidemiological feature of 'the
disease' can we assess from these data? Click on the answer from
the options below:
Interaction: Hotspot: Clustering in ethnic groups (appears in
new window)
No, the table does not give us any information on whether there
is any ethnic variation by village - there is insufficient data
presented to assess this feature.
Interaction: Hotspot: Geographical variations (appears in new
window)
Yes, we can assess whether the cases are distributed randomly
between the villages or whether they are clustered in particular
villages
Interaction: Hotspot: Temporal trends (appears in new
window)
No, there is no information in this table about how 'the
disease' was distributed over time - the table shows the total
number of cases diagnosed in a one-year period. 9.2: Exercise:
Table 1 (RHS remains static) Is there spatial clustering of cases?
Look at the data and write down your answer without doing any
formal tests.
Interaction: Button: cloud: output (appears in new window) A
number of villages have a risk of disease considerably higher or
lower than the overall risk of 50.6 per 1000 (see final row of
Table 1). This indicates that there might be clustering of cases.
(appears on main card below original text) How can you tell if
there is significant geographical variation? Think carefully about
what statistical test you could use and then click on the box
below.
Interaction: Button: cloud: output (appears in new window) A
chi-squared test of the observed population size and number of
cases would show whether the cases are distributed randomly between
all the villages. The results of a chi-squared test are:
_ = 186.8, 23 degrees of freedom, P ~< 0.001. 9.3: Exercise:
Table 1 (RHS remains static) How are cases distributed?
Interaction: Hotspot: Randomly (appears in new window)
-
In fact, the significant chi squared value indicates that cases
are *not* randomly distributed among villages.
Interaction: Hotspot: Randomly (appears in new window) That's
right. The chi-squared value indicates that there is a significant
association between villages and cases. Some villages have more
cases and some villages have fewer cases than would be expected if
cases were randomly distributed. 9.4: Exercise: Table 1 Another
approach to the distribution of cases is to consider population
size as a risk factor for 'the disease'.
The best way to visualise this is to plot a graph of the data.
Drag and drop the variables in the boxes below into the blue bays
next to the appropriate axes opposite. Reset Button Interaction:
Drag and Drop: Risk of disease Correct response: (drag to top box
on RHS) That's right, the number of cases observed will be a
function of the number of individuals in the village. We need to
use risk of disease as our outcome variable to adjust for
differences in the denominator population being observed. Incorrect
response: (drag to bottom box on RHS) In fact it is more
conventional to put the dependant or outcome variable on the
Y-axis. Interaction: Drag and Drop: Number of cases Incorrect
response: (drag to either box on RHS) Thats not quite right. The
number of cases observed will be a function of the population size.
For villages with the same risk, those with more individuals will
have more cases. We need to use risk of disease to adjust for
differences in the denominator population. Interaction: Drag and
Drop: Population sizes Correct response: (drag to bottom box on
RHS) That's correct, it is better to put the independent or
'exposure' variable on the x-axis. Incorrect response: (drag to top
box on RHS) Actually it is more conventional to put the independent
or exposure variable on the x-axis.
-
In fact it is more conventional to put the dependent or outcome
variable on the y-axis. 9.5: Exercise: Table 1 Is there an
association between population size and risk of 'the disease'?
Interaction: Hotspot: yes In fact the data are scattered
randomly, implying that there is *no* association between
population size and risk of 'the disease'.
Interaction: Hotspot: no That's correct - if there had been an
association between population size and risk of 'the disease', the
points would have formed more of an ellipse. (back to main text)
Does this mean that over-crowding is a risk factor for 'the
disease'?
Interaction: Hotspot: yes
In fact, population size is not the same as population density
(number of people in a specified area), so we have no information
on whether crowding might be a risk factor.
Interaction: Hotspot: no That's correct, population size is not
the same as population density (number of people in a specified
area), so we have no information on whether crowding might be a
risk factor.
-
9.6: Exercise: Table 1
Exercise: Table 2 Table 2 opposite shows the risk of 'the
disease' by age and gender. The data for males are shown first;
click 'swap' to see the data for females. Notice that the age
groups that have been used are not even. Complete Table 2 by
calculating the missing risk value for males to one decimal place.
Next click 'swap' and complete the corresponding table for
females.
Then, look at the data in Table 2 and try to identify the main
epidemiological features of 'the disease'. Think about how this
data can be better visualised.
Interaction: Button: cloud: output (appears in new window) By
plotting a graph of risk of 'the disease' by age group for each
gender, the distribution of 'the disease' by age and the
comparative trends for each gender will be clearer.
-
Interaction: Calculation: Calc 1 Correct response: That's right,
the risk for males aged 5 to 9 years is (193 / 1574) x 1000 =
122.6.
Remember to also calculate the empty cells in the table for
females. Incorrect response: Sorry, that's not correct. The risk
for males aged 5 to 9 years is given by the number of cases divided
by the population in that age group: Risk per 1000 = (193 / 1574) x
1000 = 122.6.
-
Try again with the empty cells in the table for females.
(Back to main page)
Interaction: Button: Swap: output (changes table on RHS)
Interaction: Calculation: Calc 1 Correct response: That's right,
the risk for females aged 2 years is (16 / 365) x 1000 = 43.8.
-
Incorrect response: Sorry, that's not correct. The risk for
females aged 2 years is given by the number of cases divided by the
population in that age group:
Risk per 1000 = (16 / 365) x 1000 = 43.8
Interaction: Calculation: Calc 2 Correct response: That's right,
the risk for females aged 25 to 29 years is (75 / 997) x 1000 =
75.2. Incorrect response: Sorry, that's not correct. The risk for
females aged 25 to 29 years is given by the number of cases divided
by the population in that age group:
Risk per 1000 = (75 / 997) x 1000 = 75.2. 9.7: Exercise: Table 1
(RHS remains static - chart)
The button below will show a graph of the data in Table 2. Look
at the graph and identify the main characteristics of 'the
disease'. Think about epidemiological inferences that are
consistent with each of the characteristics.
Write down your answer before continuing.
Interaction: Button: graph: output (appears in new window)
-
9.8: Exercise: Table 1
The points listed opposite summarise the data shown in the
graph. We can potentially make inferences from some of these
points.
Click below if you would like to see the graph again.
Interaction: Button: graph: output (appears in new window)
1. There are no cases during the first year of life.
Interaction: Button: Inferences: output Inferences This feature
is consistent with a number of potential inferences: The disease is
caused by an infection with a long incubation period. Infants are
protected from a nutritional deficiency by a breast milk diet.
Infants are protected from an infection by passive immunity from
the mother Infants are not exposed to the causal factor because of
some behavioural difference. (back to main text) 2. There is a
rapid rise in risk to a peak at age 5-9 that is similar for both
genders.
-
Interaction: Button: Inferences: output Inferences A peak of
disease among school-aged children is consistent with transmission
of an aerosol infection, such as measles, chickenpox, etc. (back to
main text) 3. The risk falls to a very low level among adolescents
(15-19). 4. In adulthood there is a difference between the genders
with males having a lower risk from 20-50 years, and then having
excess risk above 60
years.
Interaction: Button: Inferences: output Inferences Women aged 20
50 years old are of childbearing age and may have a lower immunity
to infection or a greater nutritional requirement.
Men above 60 years could have some behavioural difference in
exposure but this is difficult to assess without sociological
information. Section 10: Exercise: Table 3
Table 3 opposite shows the distribution of cases by month of
onset, adjusted to a 31-day month.
How would you describe the distribution of the cases? How do you
interpret this?
Interaction: Button: cloud: output (appears in new window) There
is a dramatic peak of cases in May and June. The majority of cases
occur between April and July. Over 50% of cases occur in May and
June alone. This disease exhibits marked seasonality, which is
likely to relate to warmer temperatures and increased sunlight
hours in the Northern Hemisphere.
-
Section 11: Exercise: Table 4
Table 4 opposite shows the relationship between risk of 'the
disease' with sanitary rating at the village level.
These ratings are developed by sanitary engineers and are based
on composite indices of general cleanliness, excreta disposal and
water supply. The ratings are given on a scale of 000 where less
than 40 is to be poor sanitation, and 70 and above is good
sanitation. Relationship between risk of the disease and sanitary
ratings
Unknown disease
-
11.1: Exercise: Table 4
Describe the data in Table 4 by selecting the correct words from
the drop-down menus below: The highest risk of 'the disease' is in
villages with the pulldown 1 sanitation. The lowest risk of 'the
disease' is in villages with the pulldown 2 sanitary rating. The
risk of 'the disease' pulldown 3
decrease with increasing sanitary rating.
Interaction: Pulldown: pulldown 1 (appears in new window)
Correct response: best Yes, the risk is highest in the village with
the highest sanitary rating. Incorrect response: intermediate No,
the village with the intermediate sanitation (a rating of 40 69)
does not have the highest risk of disease. Try again. Incorrect
response: worst No, the village with the worst sanitation (a rating
of less than 40) does not have the highest risk of disease. Try
again.
Interaction: Pulldown: pulldown 2 (appears in new window)
Correct response: intermediate Yes, the risk is lowest in the
village with the intermediate sanitary rating (40 69). Incorrect
response: best No, the village with the highest sanitary rating in
fact has the highest risk of disease. Try again. Incorrect
response: worst No, the village with the worst sanitation (a rating
of less than 40) does not have the lowest risk of disease. Try
again.
-
Interaction: Pulldown: pulldown 3 (appears in new window)
Correct response: does not That's right, there is no evidence here
that the risk of 'the disease' decreases with increasing sanitary
rating. Incorrect response: does In fact the risk does not decrease
with increasing sanitary rating, since the risk is higher in the
village with the best sanitation than it is in the village with the
worst sanitation.
11.2: Exercise: Table 4 (RHS remains static chart) Sanitation is
a risk factor for the spread of a faecal-oral transmitted disease.
Is the sanitary rating used here a good measure of the quality of
sanitation? Think about how the investigators might have validated
this rating score during their study.
Interaction: Button: cloud: output (appears in new window) To
validate whether the rating used here is useful, you could compare
it with a disease known to be spread by faecal-oral transmission.
Information was collected on the risk of typhoid fever, which is an
intestinal infection transmitted by the faecal-oral route. (appears
below main text on LHS card) By looking at the data you can see
that the risk of typhoid fever decreases as the quality of
sanitation improves. This suggests that this sanitary rating is a
good indicator of the quality of sanitation, which is related to
the spread of faecal-oral transmitted infections. (appears on RHS
card)
Typhoid fever
11.3: Exercise: Table 4 (RHS remains static)
-
Now think about how you would interpret the data from Table
4.
Interaction: Button: cloud: output (appears below main text on
LHS) There does not appear to be any association between sanitation
and risk of 'the disease' of unknown aetiology. This is surprising
as sanitation levels are usually correlated with standards of
living, hygiene, nutrition, etc., which are linked to many kinds of
diseases. However, these ratings are allocated at the village level
and may hide variations at the household level.
Section 12: Exercise: Table 5
Table 5 opposite shows the risk of disease by economic status.
The measure used is weekly family income per Adult Male Index
(AMMAIN). By studying a large number of families in the survey
area, the consumption of all types of goods, services and food was
estimated for each family member. Males aged 25 years were found to
consume the most. Adult females, children and older adults consumed
less. The amount of consumption of a 25-year-old male was
arbitrarily chosen as one unit, and other individuals were weighted
according to this. In this way, family income was adjusted for
family consumption.
Interaction: Hyperlink: AMMAIN: output (appears in new window)
AMMAIN measures variation in the gross demand for articles of
consumption among individuals of different ages and gender.
-
12.1: Exercise: Table 5 (RHS remains static - chart)
-
Is there an association between adjusted family income and risk
of 'the disease'?
Think about how you would describe the epidemiological
interpretation of these data. Write down your answer before
continuing.
Interaction: Hyperlink: cloud: output (appears in new window)
There is a clear and dramatic inverse association between risk of
'the disease' and adjusted weekly income. Risk declines as family
income per AMMAIN increases.
(appears on LHS card below main text) Is this contradictory to
the relationship with sanitary rating?
Interaction: Hyperlink: cloud: output (appears in new window) At
first glance it seems that this is contradictory, because
sanitation is likely to be positively associated with income.
However, the sanitation ratings used here were at the
village-level, while the income analysis was at the family
level.
The mistaken assumption that differences between groups reflect
differences between individuals is sometimes called the 'ecological
fallacy'.
-
Section 13: Exercise: Overall evaluation Consider all the data
together and decide what you think the aetiology of 'the disease'
is. Write it down, together with any supporting evidence from the
information given so far. Now review what each data table tells you
about the epidemiology of 'the disease'.
13.1: Exercise: Overall evaluation
For each data source listed on the next page, select whether
there is evidence in favour of or against the hypotheses listed on
the chart opposite. For each data source listed below, select
whether there is evidence for or against the hypotheses listed
opposite. Note that for some hypotheses there is no evidence either
for or against.
Interaction: Button: Geographic village distribution: output
(appears on RHS) Data source: Geographic-village distribution For
Against Insufficien
t evidence Hereditary ? (hotspot
1) ? (hotspot 7)
? (hotspot 13)
Infectious : aerosol
? (hotspot 2)
? (hotspot 8)
? (hotspot 14)
Infectious : faecal oral
? (hotspot 3)
? (hotspot 9)
? (hotspot 15)
Infectious : vector borne
? (hotspot 4)
? (hotspot 10)
? (hotspot 16)
Infectious : contact
? (hotspot 5)
? (hotspot 11)
? (hotspot 17)
Nutritional ? (hotspot 6)
? (hotspot 12)
? (hotspot 18)
Interaction: Hotspot: ? (hotspot 1) (from above table) Incorrect
response: (appears in new window) In fact, a potentially fatal
disease is unlikely to have such a high annual risk if it is
hereditary.
Interaction: Hotspot: ? (hotspot 2) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
-
Interaction: Hotspot: ? (hotspot 3) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 4) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 5) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 6) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 7) (from above table) Correct
response: (appears in new window) That's right, a potentially fatal
disease is unlikely to have such a high annual risk if it is
hereditary.
Interaction: Hotspot: ? (hotspot 8) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 9) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 10) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 11) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 12) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 13) (from above table)
Incorrect response: (appears in new window) In fact there is
sufficient evidence in this case. Please try again.
-
Interaction: Hotspot: ? (hotspot 14) (from above table)
Incorrect response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 15) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 16) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 17) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 18) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this. Table is as
follows after all hotspots have been clicked on:
Interaction: Button: Age-sex distribution: output
(table changes on RHS)
-
Data source: Age-sex distribution For Against Insufficien
t evidence Hereditary ? (hotspot
1) ? (hotspot 7)
? (hotspot 13)
Infectious : aerosol
? (hotspot 2)
? (hotspot 8)
? (hotspot 14)
Infectious : faecal oral
? (hotspot 3)
? (hotspot 9)
? (hotspot 15)
Infectious : vector borne
? (hotspot 4)
? (hotspot 10)
? (hotspot 16)
Infectious : contact
? (hotspot 5)
? (hotspot 11)
? (hotspot 17)
Nutritional ? (hotspot 6)
? (hotspot 12)
? (hotspot 18)
Interaction: Hotspot: ? (hotspot 1) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 2) (from above table) Correct
response: (appears in new window) That's right. The peak risk among
school-aged children is consistent with an infection spread by
aerosol-contact, such as measles, chickenpox, etc.
Interaction: Hotspot: ? (hotspot 3) (from above table) Incorrect
response: (appears in new window) In this case the evidence is
inconsistent.
The decline in risk among older children is consistent with
reduced transmission of faecal-oral transmitted diseases in the
older age groups. Younger children tend to have less hygienic
behaviours and therefore have a higher risk of faecal-oral
transmitted diseases. However the increase with age in adulthood is
not consistent with risk of faecal-oral transmitted diseases.
Interaction: Hotspot: ? (hotspot 4) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 5) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 6) (from above table) Correct
response: (appears in new window) That's right. The high risk among
women of child-bearing age is consistent with a nutritional factor
as they have greater nutritional requirements during pregnancy
-
and lactation. The absence of cases among infants could be
because they are provided with sufficient nutrition from
breast-milk. In this case the peak risk among young children could
be explained by patterns of food distribution in the household or a
higher nutritional requirement in this age group.
Interaction: Hotspot: ? (hotspot 7) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 8) (from above table) Incorrect
response: (appears in new window) In fact, the evidence of peak
risk among school-aged children is consistent with an infection
spread by aerosol-contact, such as measles, chickenpox, etc.
Interaction: Hotspot: ? (hotspot 9) (from above table) Incorrect
response: (appears in new window) In this case the evidence is
inconsistent.
The decline in risk among older children is consistent with
reduced transmission of faecal-oral transmitted diseases in the
older age groups. Younger children tend to have less hygienic
behaviours and therefore have a higher risk of faecal-oral
transmitted diseases. However the increase with age in adulthood is
not consistent with risk of faecal-oral transmitted diseases.
Interaction: Hotspot: ? (hotspot 10) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 11) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 12) (from above table)
Incorrect response: (appears in new window) In fact, the high risk
among women of child-bearing age is consistent with a nutritional
factor as they have greater nutritional requirements during
pregnancy and lactation. The absence of cases among infants could
be because they are provided with sufficient nutrition from
breast-milk. In this case the peak risk among young children could
be explained by patterns of food distribution in the household or a
higher nutritional requirement in this age group.
Interaction: Hotspot: ? (hotspot 13) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 14) (from above table)
-
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again.
Interaction: Hotspot: ? (hotspot 15) (from above table)
Incorrect response: (appears in new window) In this case the
evidence is inconsistent.
The decline in risk among older children is consistent with
reduced transmission of faecal-oral transmitted diseases in the
older age groups. Younger children tend to have less hygienic
behaviours and therefore have a higher risk of faecal-oral
transmitted diseases. However the increase with age in adulthood is
not consistent with risk of faecal-oral transmitted diseases.
Interaction: Hotspot: ? (hotspot 16) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 17) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 18) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again. Table is as
follows after all hotspots have been clicked on:
-
Interaction: Button: Seasonal distribution: output
(table changes on RHS)
Data source: Seasonal distribution For Against Insufficien
t evidence Hereditary ? (hotspot
1) ? (hotspot 7)
? (hotspot 13)
Infectious : aerosol
? (hotspot 2)
? (hotspot 8)
? (hotspot 14)
Infectious : faecal oral
? (hotspot 3)
? (hotspot 9)
? (hotspot 15)
Infectious : vector borne
? (hotspot 4)
? (hotspot 10)
? (hotspot 16)
Infectious : contact
? (hotspot 5)
? (hotspot 11)
? (hotspot 17)
Nutritional ? (hotspot 6)
? (hotspot 12)
? (hotspot 18)
Interaction: Hotspot: ? (hotspot 1) (from above table) Incorrect
response: (appears in new window) In fact, seasonality of cases is
not consistent with a hereditary disease.
Interaction: Hotspot: ? (hotspot 2) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 3) (from above table) Correct
response: (appears in new window) Yes, the peak of cases in summer
is consistent with a faecal-oral transmitted infection.
Interaction: Hotspot: ? (hotspot 4) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 5) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 6) (from above table) Correct
response: (appears in new window) Yes, the peak of cases prior to
the harvesting of crops in July and August is consistent with a
nutritional deficiency.
-
Interaction: Hotspot: ? (hotspot 7) (from above table) Correct
response: (appears in new window) That's right, seasonality of
cases is inconsistent with a hereditary disease.
Interaction: Hotspot: ? (hotspot 8) (from above table) Correct
response: (appears in new window) Correct, the peak of cases in
summer is not consistent with an aerosol-transmitted infection.
Interaction: Hotspot: ? (hotspot 9) (from above table) Incorrect
response: (appears in new window) In fact, the peak of cases in
summer is consistent with a faecal-oral transmitted infection.
Interaction: Hotspot: ? (hotspot 10) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 11) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 12) (from above table)
Incorrect response: (appears in new window) In fact, the peak of
cases prior to the harvesting of crops in July and August is
consistent with a nutritional deficiency.
Interaction: Hotspot: ? (hotspot 13) (from above table) Correct
response: (appears in new window) In fact, there is insufficient
evidence in this case. Please try again
Interaction: Hotspot: ? (hotspot 14) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again.
Interaction: Hotspot: ? (hotspot 15) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again.
Interaction: Hotspot: ? (hotspot 16) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 17) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
-
Interaction: Hotspot: ? (hotspot 18) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again. Table is as
follows after all hotspots have been clicked on:
(back to LHS card)
Interaction: Button: Relationship with sanitary ratings:
output
(table changes on RHS)
Data source: Relationship with sanitary ratings For Against
Insufficien
t evidence Hereditary ? (hotspot
1) ? (hotspot 7)
? (hotspot 13)
Infectious : aerosol
? (hotspot 2)
? (hotspot 8)
? (hotspot 14)
Infectious : faecal oral
? (hotspot 3)
? (hotspot 9)
? (hotspot 15)
Infectious : vector borne
? (hotspot 4)
? (hotspot 10)
? (hotspot 16)
Infectious : contact
? (hotspot 5)
? (hotspot 11)
? (hotspot 17)
Nutritional ? (hotspot 6)
? (hotspot 12)
? (hotspot 18)
-
Interaction: Hotspot: ? (hotspot 1) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 2) (from above table) correct
response: (appears in new window) Thats right, there is sufficient
evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 3) (from above table) Incorrect
response: (appears in new window) In fact, the lack of any
association with village sanitary ratings suggests that this is not
a faecal-oral transmitted infection.
Interaction: Hotspot: ? (hotspot 4) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 5) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 6) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 7) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 8) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this. Interaction: Hotspot: ?
(hotspot 9) (from above table) Correct response: (appears in new
window) That's right, the lack of any association with village
sanitary ratings suggests that this is not a faecal-oral
transmitted infection.
Interaction: Hotspot: ? (hotspot 10) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 11) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
-
Interaction: Hotspot: ? (hotspot 12) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 13) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 14) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 15) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again.
Interaction: Hotspot: ? (hotspot 16) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 17) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 18) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this. Table is as
follows after all hotspots have been clicked on:
-
Interaction: Button: Relationship with sanitary ratings:
output
(table changes on RHS) Data source: Relationship with economic
status
For Against Insufficient evidence
Hereditary ? (hotspot 1)
? (hotspot 7)
? (hotspot 13)
Infectious : aerosol
? (hotspot 2)
? (hotspot 8)
? (hotspot 14)
Infectious : faecal oral
? (hotspot 3)
? (hotspot 9)
? (hotspot 15)
Infectious : vector borne
? (hotspot 4)
? (hotspot 10)
? (hotspot 16)
Infectious : contact
? (hotspot 5)
? (hotspot 11)
? (hotspot 17)
Nutritional ? (hotspot 6)
? (hotspot 12)
? (hotspot 18)
Interaction: Hotspot: ? (hotspot 1) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 2) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
-
Interaction: Hotspot: ? (hotspot 3) (from above table) Incorrect
response: (appears in new window)
An inverse correlation with socio-economic status could indicate
an inverse correlation with household-level sanitation. This would
be consistent with a faecal-oral transmitted infection. However
there is insufficient information from the economic data to support
this.
Interaction: Hotspot: ? (hotspot 4) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 5) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 6) (from above table) Correct
response: (appears in new window) The inverse correlation with
socio-economic status could indicate an inverse correlation with
consumption, and this would be consistent with a disease due to a
nutritional deficiency.
Interaction: Hotspot: ? (hotspot 7) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this.
Interaction: Hotspot: ? (hotspot 8) (from above table) Incorrect
response: (appears in new window) No, in fact there is insufficient
evidence or information to support this. Interaction: Hotspot: ?
(hotspot 9) (from above table) Incorrect response: (appears in new
window) An inverse correlation with socio-economic status would
indicate an inverse correlation with household-level sanitation,
and this would be consistent with a faecal-oral transmitted
infection. However there is insufficient information from the
economic data to support this. Interaction: Hotspot: ? (hotspot 10)
(from above table) Incorrect response: (appears in new window) No,
in fact there is insufficient evidence or information to support
this.
Interaction: Hotspot: ? (hotspot 11) (from above table)
Incorrect response: (appears in new window) No, in fact there is
insufficient evidence or information to support this.
Interaction: Hotspot: ? (hotspot 12) (from above table)
-
Incorrect response: (appears in new window) In fact, the inverse
correlation with socio-economic status could indicate an inverse
correlation with consumption, and this would be consistent with a
disease due to a nutritional deficiency.
Interaction: Hotspot: ? (hotspot 13) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 14) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 15) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 16) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 17) (from above table) Correct
response: (appears in new window) That's right, there is
insufficient evidence to support or refute this.
Interaction: Hotspot: ? (hotspot 18) (from above table)
Incorrect response: (appears in new window) In fact, there is
insufficient evidence in this case. Please try again. Table is as
follows after all hotspots have been clicked on:
-
13.2: Exercise: Overall evaluation
What do you consider the aetiology of 'the disease' to be now?
Write down your answer, then look at the hints given below.
Interaction: Button: Hint 1: output (appears in new window) Hint
1
The disease is highly seasonal, and starts to decline at the
time of crop harvest in July and August.
Interaction: Button: Hint 2: output (appears in new window) Hint
2
Income is linked to consumption. There is a strong association
between income and risk of 'the disease'.
Interaction: Button: Hint 3: output (appears in new window) Hint
3
Men aged 25 years old have the highest consumption of goods and
food. They also have the lowest risk of 'the disease'.
Has your opinion changed after seeing the hints? Check whether
you were correct by clicking on the button below.
Interaction: Button: Show: output (appears on RHS)
-
Despite having many epidemiological characteristics of an
infectious disease, 'the disease' is in fact due to a nutritional
deficiency. It is called pellagra, from Italian pelle (skin) and
agro (rough) because of the distinctive photosensitive dermatitis.
The clinical presentations are:
diarrhoea dermatitis
dementia.
Interaction: Hyperlink: photosensitive dermatitis: output
(appears in new window)
On parts of the body exposed to sunlight, the skin develops a
rash that later develops into the characteristic pellagrous
lesions. This may be another reason why more cases were diagnosed
at the height of the summer in May and June. 13.3: Exercise:
Overall evaluation Pellagra is caused by a deficiency of tryptophan
and/or nicotinic acid.
Maize contains nicotinic acid but in an unavailable form, so
that a maize-rich diet can be associated with pellagra. Good
sources of nicotinic acid include meat, wheatgerm and yeast.
The data presented here are from surveys conducted in the
Southern USA under the supervision of J. Goldberger. See his essay
"Considerations on Pellagra", for information on the debate
surrounding the aetiology of the disease prior to these
studies.
Interaction: Hyperlink: Considerations on Pellagra: output
(appears in new window)
This is on pages 99-102 of one of the course books: "The
Challenge of Epidemiology: Issues and Selected Readings" Eds. Buck
et al. (1989). 13.4: Exercise: Overall evaluation This exercise
enabled you to use the information learned in the session to
investigate whether or not the case data supplied were due to an
infectious aetiology. In this case the disease was not infectious
despite sharing many epidemiological characteristics of infectious
diseases.
-
By contrast, Beral's comparison of mortality patterns for cancer
of the cervix, with trends in incidence of sexually transmitted
diseases, showed associations between the temporal, social class,
occupational, and geographic distributions for these diseases
(Beral, 1974). She proposed that the cause of cervical cancer was a
sexually transmitted infection, when many considered it to be a
non-infectious disease. Human papilloma virus has since been
associated with cases of the disease. See Beral (1974) in your
workbooks, for the evidence of an infectious aetiology for cervical
cancer. 13.5: Exercise: Overall evaluation There are a number of
diseases for which the aetiology is still not fully understood. In
many cases the evidence points to the need for a number of
co-factors to be present together before an individual develops the
disease. The cause of multiple sclerosis still remains unknown, but
it appears to involve a complex interaction of genetic and
environmental factors. There is also some evidence that a viral
infection may be involved in triggering the disease in genetically
susceptible individuals. For more information on this topic read
the first two pages of the review by Noseworthy (1999) in your
reader. 13.6: Exercise: Overall evaluation The aetiology of
childhood leukaemia is also still under debate. It has been
associated with exposure to X-rays in utero and to ionising
radiation. In a few cases a link has been found with genetic
abnormalities such as Down's syndrome. Over 80% of cases are still
unexplained, and an infectious aetiology has been proposed for
this. For further information on this topic read the review by
Kinlen (1996) on the evidence supporting the hypothesis of an
infectious aetiology of leukaemia in your reader. Section 14:
Summary This is the end of EC02. When you are happy with the
material covered here please move on to session EC03. The main
points of this session will appear below as you click on the
relevant title. Characteristics of infectious aetiology (link
returns RHS to relevant page)
-
During this session we identified a number of epidemiological
features that are associated with, but not exclusive to, infectious
diseases. These fall into three distinct categories: Time Place
Person
Time (link returns RHS to relevant page) The temporal
distribution of cases can be divided into the following patterns:
(i) Temporal clustering associated with a common source or sources
of infection. (ii) Seasonality regulated by climatic changes that
restrict transmission of the infection. (iii) Cyclical patterns
associated with recurrent social events, periodic climatic
phenomena, or the interval to reach the critical threshold of
susceptibles. (iv) Long-term trends that may indicate associations
with an exposure. Place (link returns RHS to relevant page) The
spatial distribution of cases can occur on two levels: (i) Spatial
clustering associated with a common source or sources of infection.
(ii) Geographical restriction regulated by ecological and climatic
conditions that restrict transmission of the infection. Person
(link returns RHS to relevant page) The disease may be exclusive
to, or more common among, a particular group of individuals: (i)
Occupational groups tend to have
-
special behaviours and known exposures that can be associated
with the infection. (ii) Behaviours may increase exposure to the
infection and can be an important area for focusing public health
interventions. (iii) Socio-economic status is associated with many
infectious diseases, as poor living conditions often provide
favourable environments for transmission of infection. (iv)
Immunological status can provide a clear indication of an
infectious aetiology. Final tips (link returns RHS to relevant
page)
Interaction: Tabs: 1 : output When faced with epidemiological
data about a disease, think about what each piece of information
might be telling you about the cause. Remember that some
epidemiological features may indicate different aetiologies.
Spatial clustering may be due to an infectious or environmental
cause, and at the family level it can even be genetic.
Interaction: Tabs: 2 : output Also consider that lack of
conclusive evidence for a proposed cause does not necessarily mean
that it is not involved. It could suggest that it is only one of a
number of factors required to cause the disease. Other potential
co-factors may need to be investigated to fully understand the
aetiology of the disease and the role, if any, of infectious
agents.
2.1: Introduction4.1: Time when the disease occurs4.2: Time when
the disease occurs4.3: Time when the disease occurs4.4: Time when
the disease occurs4.5: Time when the disease occurs4.6: Time when
the disease occurs4.7: Time when the disease occurs4.8: Time when
the disease occurs4.9: Time when the disease occurs4.10: Time when
the disease occurs5.1: Place where the disease occurs5.2: Place
where the disease occurs5.3: Place where the disease occurs5.4:
Place where the disease occurs5.5: Place where the disease
occurs6.1: Person who develops the disease6.2: Person who develops
the disease6.3: Person who develops the disease6.4: Person who
develops the disease6.5: Person who develops the disease7.1:
Interactions7.2: Interactions7.3: Interactions8.1: Exercise: A
disease of unknown aetiology8.2: Exercise: A disease of unknown
aetiology9.1: Exercise: Table 19.2: Exercise: Table 19.3: Exercise:
Table 19.4: Exercise: Table 19.5: Exercise: Table 19.6: Exercise:
Table 19.7: Exercise: Table 19.8: Exercise: Table 111.1: Exercise:
Table 411.2: Exercise: Table 411.3: Exercise: Table 412.1:
Exercise: Table 513.1: Exercise: Overall evaluation13.2: Exercise:
Overall evaluation13.3: Exercise: Overall evaluation13.4: Exercise:
Overall evaluation13.5: Exercise: Overall evaluation13.6: Exercise:
Overall evaluation