[MUSIC PLAYING] Good day and welcome to the
introduction to epidemiology. My name is Dr. Kate Glyn and
I'm the Associate Director for Science at CDC's Division
of Scientific Education and Professional Development. This course is a basic
overview of epidemiology. In today's session,
we'll define epidemiology and explain the role
of the epidemiologist in public health. We'll learn how epidemiologists
characterize public health problems and the steps
an epidemiologist undertakes when investigating
a disease outbreak. After this session, you'll be
able to define epidemiology, describe basic terminology
and concepts of epidemiology, identify types of data sources
and basic methods of data collection and interpretation,
to describe a public health problem in terms of
time, place, and person, and identify the key components
of a descriptive epidemiology outbreak investigation. Before we talk
about epidemiology, let's learn a bit about
the public health approach and how it relates to the
public health core sciences. Let's think about public
health in a broader context. Public health
problems are diverse and can involve infectious
disease, chronic diseases, emergencies, injuries,
environmental health problems, and all kinds of
other health threats. Regardless of the
topic, however, we take the same general approach
to a public health problem by following four general steps. First, we ask, what
is the problem? In public health, we
identify the problem by using surveillance systems to monitor
health events and behaviours occurring among a population. After we've identified
the problem, we next ask, what is the cause? For example, there
are risk factors that might make certain
populations more susceptible to a particular
disease or condition, something in the environment, perhaps,
or certain behaviors that people are practicing. After we've addressed
what is the cause and identified risk
factors, we then ask ourselves, what intervention
works to address this problem? We think about
what interventions have worked in the
past and whether those are applicable to the
particular population that we're investigating. In the last step,
we ask, how can we implement the intervention? Given the resources
we have and what we know about the
affected population, will this intervention
really work? As we go through
the course, you'll see different examples about
how this four step public health approach is applied. But to implement this
public health approach, practitioners use and
apply scientific methods. These methods come from
a series of core sciences that provide the foundation
for public health. These sciences include
public health surveillance, which we use to monitor a
public health situation. Epidemiology enables
us to determine where diseases originate, how
they move through populations, and why they're
moving, and understand how we can prevent them. We're going to learn more about
epidemiology in today's course. Public health laboratories
support public health by allowing us to find a
diagnosis for a condition that we're investigating. Laboratories also
support public health by conducting
research and testing. As we continue to move from
paper to electronic health records, the science of
public health informatics allows us to do it that the
most effectively possible. Informatics deals
with the methods of collecting, compiling,
and effectively using electronic data to solve
public health problems. Prevention effectiveness
is closely linked to public health policy. Prevention effectiveness
studies provide important economic
information to decision makers and allow them to choose
from among the options the best option possible. Public health is
better able to respond by using all of the information
these sciences can provide. So take, for example, the public
health problem of influenza. Public health
surveillance can monitor where and when
cases of influenza are occurring each year. Professionals can use the
science of epidemiology to understand why some
populations choose to become vaccinated and some do not. They can use the
science of informatics to get clinical information
from electronic health records from doctors offices
and hospitals. Public health practitioners can
benefit from laboratory science because the laboratories
can help diagnose whether this disease
caused by causing fever and a cough is in fact
influenza or something else. These laboratories
can also tell us what particular
strain of influenza is predominant in a given year. And they can use a
prevention effectiveness to assess whether in fact
an influenza vaccination campaign, that might
cost say $200,000, might ultimately result in
a savings of over a million dollars because of savings
in medical care costs, loss of wages, and other costs. So with this background, let's
learn about epidemiology, how it aligns with the
scientific approach, and its purpose in
public health practice. Epidemiology is defined as
the study of the distribution and determinants of
health related states among specified populations
and the application of that study to the
control of health problems. The purpose of epidemiology
in public health practice is to discover the agent, host,
and environmental factors that affect health. To determine the
relative importance of causes of illness,
disability, and death. To identify those segments
of the population that have the greatest risk from
specific causes of ill health. And to evaluate
the effectiveness of health programs and services
in improving population health. To solve health
problems, epidemiologists use the public health
approach that we discussed. And specifically, they do
this by collecting data, by conducting an assessment,
by doing hypothesis testing, and by taking action. The boxes on the far
right of your screen we'll go through one by
one in greater detail to show how an
epidemiologist actually implements these steps. First, data are collected
about health problems occurring among the population
through, as we've discussed, public health surveillance. The data collected
include information about when and where the
population was affected, as well as who was actually
affected by the condition under surveillance. That is time, place, and person. This is known as
descriptive epidemiology and we're going to talk more
about that later in the course. Next, the epidemiology
establishes inferences on the basis of
these collected data and draws initial conclusions. From there, he or she
uses the information to generate hypotheses
about what might be causing this public health problem. Then, the how and
why of the condition is determined by
conducting tests or studies to test
the hypotheses that you've developed. This determination
of how and why is known as analytic
epidemiology. Again, we'll cover
a little bit more about this later in the course. And finally, the
epidemiologist takes action. In public health, this action is
often known as an intervention. We take action to intervene
to either prevent the disease or condition from spreading
or continuing to occur, or to actually promote healthy
behaviors in a population. The epidemiologist
recommends implementing some form of action at
the population level. For example, a
community intervention. One of my earliest acts
as an epidemiologist, was to investigate a cruise ship
outbreak of diarrheal illness. I went in and I
conducted data collection from all of the passengers
and crew members on the ship by administering a
standard questionnaire. I asked them where they were on
the ship, when they became ill, if they did, and who
they actually were, those that became ill
and those that didn't. Using this data, I
generated a hypothesis that, in fact, the ship had
taken on contaminated water at one of its foreign ports. I was able to test this
hypothesis by comparing persons who were ill with the
diarrheal illness, to persons who were not. I looked back and said, I notice
that the persons who were ill were more likely
to have consumed tap water from the ship. Person's who did not get
ill were more likely to have consumed only bottled water. So based on this assessment,
I made the recommendation that the ship should
hyper chlorinate its water supply, which
is one of the most effective treatments
against the particular agent we believe was causing
this outbreak, Norovirus. So now, let's review
what we've learned about epidemiology so far
through a few knowledge check questions. Here's the first one. All of the following illustrate
the purpose of epidemiology except which of the following? Correct answer is C. It's
not a purpose of epidemiology to provide treatment for
patients in clinical settings. Epidemiologists use a model
for studying infectious disease and its spread. That involves the
microbe that causes the disease, the organism
that harbors the disease, and the external factors
that cause or allow disease transmission. This is also known as
which of the following? The correct answer is C,
host, agent, and environment. So now that we know
what epidemiology is, let's review some key
terms commonly associated with the study. You'll see the terms on this
side throughout the course. The first is
epidemic or outbreak. This is a disease occurrence
among a population that is in excess
of what is expected in a given time and place. A cluster is a group of cases in
a specific time and place that might be more than expected. The third term is endemic. This describes a
disease or condition that's present among a
population at all times. This is in contrast
with a pandemic, which is a disease
or condition that's causing an epidemic
that actually spreads across regions. And the last term is a rate. A rate is a number
of cases occurring during a specific period
in a specific population. A rate is always dependent
on the size of the population during that period. So now, I'd like you to pick the
appropriate term that matches with the following statements. Malaria is present in
Africa at all times because of the presence
of infected mosquitoes. Malaria is what in Africa? The correct answer is A,
malaria is endemic in Africa. Number two, the Ebola
virus in parts of Africa, is in excess of what is
expected for this region. Ebola is a what in Africa? The Ebola virus is
currently causing an epidemic in parts of Africa. Number three, HIV/AIDS is one
of the worst global diseases in history. HIV/AIDS is a what? Correct? Number B, or Letter B,
HIV/AIDS is a pandemic. And for our final one, in March
1981, an outbreak of measles occurred among employees at
factory x in Fort Worth, Texas. This group of cases, in this
specific time and place, can be described as which of the
terms on the top of the slide? Well done. The correct answer
is B, cluster. So we've defined the term rate. Now, let's look a little
bit more in depth at rates and how you actually
calculate them. In epidemiology, we cannot stop
simply at looking at the number of cases. We also need to compare rates. Rates help us compare problems
among different populations that include two
or more groups who differ by a selected
characteristic or characteristics. For example, we
might compare persons who ate a certain meal
at a certain restaurant at a certain time with
people who did not eat that meal at that
restaurant at that same time, and look for cases of
food borne illness. Or we might compare
men with women, for example, looking
for a risky behavior, like whether or not
they drive intoxicated. By comparing population
characteristics, we can observe more clearly
what factors might be associated with the health event. Such as, in these cases,
what might have been causing the food borne illness or
who really is more commonly driving intoxicated. We can then determine
what actions to take. Rates also help us
determine unusual activity, allowing us to compare
a baseline level to a current level. So for example,
for influenza, we could calculate the rate
of influenza of this year and by comparing it to
baseline rates in other years, determine whether this year
the rate is higher than usual or not. To calculate a
rate, we first need to determine the
frequency of disease, which includes these three
components, the number of cases of the illness or the
condition we're counting, the size of the
population at risk, and the period during which
we're calculating the rate. The formula shown on the slide
provides the number of cases, as a percentage of the
population, for a given period. So let's say, for
example, that you want to find out how
common it is in your city whether people don't wear seat
belts while they're driving. You could actually observe
a given intersection, watch all of the drivers that
pass through this intersection, and then count among
all of those drivers the number that actually
are not wearing seat belts. So let's say in
this weekend, you observed that 10 drivers did
not wear their seat belts and there were 100
total drivers passing through that intersection. So your rate of not wearing
seat belts would be 10 over 100, times 100, or 10%
of the drivers. But now let's look at
an example of how rates were calculated and
used in an actual case of unexplained pneumonia. Members of the American
Legion gathered for the annual American Legion Convention
held July 21 through 24 , 1976, in Philadelphia. Soon after the convention
began, a substantial number of attendees were admitted to
hospital emergency departments or seen it doctors offices with
sudden onset of fever, chills, headache, malaise, dry
cough, and muscle pain. More troublesome than that, is
that during July 26 to August 1, a total of 18
conventioneers died, reportedly from pneumonia. On the morning of
August 2, a nurse at a veteran's hospital
in Philadelphia called CDC to report cases
of severe respiratory illness among the convention attendees. Subsequent conversations
that same day with public health officials
uncovered an additional 71 cases among persons who'd
attended the conference. The goal of the
investigation was to find out why these
conventioneers were becoming ill and in some cases, dying. These cases of
unexplained pneumonia were investigated
and subsequently given the name Legionnaires
Disease because of their association
with attendance at the American Legion
Convention in July of 1976. CDC investigators focused
on a particular hotel as the possible
source of the outbreak because it was a common
factor among persons who actually were ill with this
unusual pneumonia-like illness. The investigators
wanted to find out if any trends
existed by age group among hotel guests
who became ill. So they looked at the
different age groups and that's shown in
the rows of this table. In the first column, you
have the disease frequency, or the number of conventioneers
that became sick, by age group. In the second column,
you have the unit size of the population, or the
total number of conventioneers, in each age group. And the third
element is indicated by the arrow, the time period
that you are looking at to calculate this rate. So we can calculate the
rate using the formula that we've already discussed
for looking at each age group and the rate at which
they became ill after staying at, or attending a meeting, at
hotel A during the convention. And the rates themselves
are shown in the column to the far right. So looking at this
table, which age group had the largest number of
conventioneers that became ill? It's actually the age
group of 50 through 59, who had 27 persons who became ill. But you'll notice
that this is also the largest proportion of all
of the convention attendees. So now I'll ask you, which
group had the highest rate of persons becoming ill? So the highest rate is actually
among those individuals 70 years or older, and
this is why epidemiologists not only look at the raw numbers
but also must look at the rate, or they might actually be
slightly misled by their data. So let's pause for
a knowledge check. On day one of a technology
conference in San Diego, 15 presenters who were
setting up for their sessions in annex x, became ill
with flu like symptoms. During the course
of the conference, 20 participants who attended
sessions, also in annex x, became ill with
the same symptoms. To begin calculating the
rate of this outbreak, investigators should
first determine which of the following? The correct answer is B, the
number of cases of illness. So we're going to come back
to this Legionnaires Disease investigation a
little later, but what I'd like to talk about
now are approaches to epidemiology
studies, specifically, experimental and
observational epidemiology. In an experimental
study, the investigators can control certain
factors within the study from the beginning. An example of this kind of study
is a vaccine efficacy trial that might be conducted by
investigators from the National Institutes of Health. They could take a group
of study participants and administer to
some, randomly, a new experimental
vaccine, and the balance would receive the
standard routine vaccine that's already being given. The investigators would
observe the outcome of this study, looking
at whether the health event that the vaccine is
protecting against actually occurs or not, and
take the decision about which vaccine
should ultimately be implemented more widely. In an observational study,
in contrast, the epidemiology does not control
the circumstances but can only observe
what is happening or what has already happened. Observational studies
can be further subdivided into descriptive and
analytic studies. Descriptive epidemiology is the
more basic of these categories and the stalwart of
epidemiology practice. In a descriptive study,
the epidemiologist collects information to
characterize and summarise the public health
event or problem. In the analytic study,
the epidemiologist relies on comparisons
between groups to determine the role of
specific causative conditions or risk factors. So hopefully, you're already
learning that time, place, and person is the mantra
of the epidemiologist. So another way of comparing
descriptive and analytic epidemiology is to say that
during the descriptive process, we're concerned with when
the population was affected, where the population
was affected, and who was actually affected. From these observations,
the epidemiologists can generate a hypothesis about
why things really happened. And then to test
that hypothesis, epidemiologists must
use an analytic process in which they ask how and why
the population was affected. So let's look at an example. In 1982, an epidemiologist
in the Georgia Department of Public Health
became interested in the number of deaths
associated with farm tractors. He determined he could
actually examine this issue by using information that
had already been collected, by using death certificates
that were part of a previously existing surveillance system. So he obtained the
death certificates for all of the deaths
from 1971 through 1981 that were associated with
farm tractor incidents. After collecting the data,
he described the problem and then actually
used the information to generate hypotheses. Now, we'll talk
about the hypotheses in a second but let's
first look at how he described the problem. This graph describes the
when for 166 of the farm tractor associated deaths. Let's examine the data by
looking at the time of day when the deaths occurred. What inferences can we
make from this graph? Well, you can see that
there are two real peaks in the number of deaths. One is happening
between 11 and 12, right before lunch, and
the second, the highest peak, between 4:00 and 5:00
PM, towards the end of the day. Also when children
might be home. In addition, you can see
that one of the lowest events is between these two
peaks, between the hours of 12:00 PM to 1:00 PM. So looking at these
data, you might infer that increased
number of deaths occur when farmers are
more tired, perhaps right before lunch or right
at the end of the day, and that in fact, fewer
deaths happen between 12 and 1 because people are
eating lunch and less likely to be out on tractors and
potentially hurting themselves. Also, you might think
about that later peak as when children
are home from school and maybe somehow children might
be contributing to the fact that these events are happening. So this graph describes the who. It indicates the number
of deaths by age group. What else can you actually
infer from this graph? It's clear that
there's an increase in the number of deaths
among older persons, which again is part of the
descriptive analysis. So let me ask you, by
reviewing these data, which age group is at greatest risk for
death from tractor related incidents among this population? In fact, you cannot tell that
from this graph because this only has raw data, not rates. So maybe, in fact, there
are more tractor drivers among the 60 through
69-year-old age category and that's why
there's more deaths. Conversely, maybe it's
because older tractor drivers are more likely to be associated
with dangerous tractor events, or more likely to die
if they get injured. But you don't really
know because you don't know the population who's
actually driving the tractors. And what about these children? The deaths occurring in
zero to nine years of age? What can you say about them? Well, you could
hypothesize that when they come home from school maybe
they take over on the tractors. Maybe they play and are more
likely to cause accidents. But again, you can't actually
answer these questions. But these data are valuable
for generating hypotheses for further investigation. And this map describes
the where of the tractor associated deaths, with
the numbers showing the number of deaths
and where they occurred in the state of Georgia. You can see that more
deaths are actually occurring in northern Georgia. In fact, this part of
Georgia is more mountainous and has more rugged terrain. Fewer deaths are happening in
southern and central Georgia, where the land is
actually flatter. So again, these
data can't tell you why the accidents are happening,
but you could hypothesize. Perhaps it's more dangerous
to drive a tractor in the more rugged terrain
of northern Georgia. Or perhaps there
are fewer deaths in the south because
there's more tractor driving and the drivers are more
experienced at driving their tractors, so less
likely to get into an accident and die. So these are the
kinds of hypotheses that this investigator
would have gone on then to further try and
test in his analytic process. So time for another
knowledge check. Choose the correct answer from
these following three choices. An epidemiologist
is doing a study on the sleep patterns
of college students but does not provide
any intervention. What type of study is this? The correct answer is C, this
is an observational study. A study of heart disease
comparing a group who eats healthy foods
and exercises regularly with one who does not, in an
effort to test association. What kind of study is this? This is B, an analytic study. And number two, a study to
describe the eating habits of adolescents, aged 13 to
18 years, in community x, is what kind of study? This is A, a descriptive study. So we've covered definitions,
key terms, and approaches to epidemiology. Let's turn now to the
sources epidemiologists use, the methods they
use to collect data, and the three common
types of study designs. Where do all of the data come
from that epidemiologists use? And how are they collected? Here are some
methods and examples for each of the main
sources of data. But please note that this
list is far from exhaustive. We often collect
data from individuals by using questionnaires
and surveys, for example. Environmental data are collected
in multiple different ways. For example, an epidemiologist
might collect samples from a river to test for the
presence of certain toxins. Health care providers collect
considerable amounts of data and record these in
clinical records, which can be used by an epidemiologist. And data from non-health
related sources, such as financial
and legal records, can also be gathered
and reviewed. Looking at sales, court
proceedings, arrests, and so on. Examples of non-health
related sources that can provide key
data to public health include information
on cigarette sales, or records of intoxicated
driver arrests. After data are collected,
as we've mentioned, hypotheses are tested,
trends are evaluated, and factors between or
among groups are compared. By conducting different
studies, epidemiologists are attempting in
different ways to discover if a causal association exists
between an exposure, or risk factor, and a health condition
by making comparisons of factors between groups. To test hypotheses, three study
designs are commonly used. These are cross-sectional,
cohort, and case control. The first of these, a
cross-sectional study, is similar to a
survey in that it provides a snapshot of the
population at a point in time. Using this study, epidemiologist
defines the target population, then collects data from
the population or a subset of the population, at one
specific point in time, and participants are included
regardless of their exposure or disease status. An example would be to
conduct a random telephone survey at a university and
ask a subset of students how frequently they'd
exercised in the last week, and whether they were
considered overweight, normal weight, or obese. You could get an idea
of exercise habits, and the prevalence of obesity
among this population, but you also need to consider
when you take that survey. For example, was the previous
week you were asking about, was that finals week? Was that spring break? You really have to
think that this is just a single snapshot in time and
that selection of that time is really important. In the cohort study,
the epidemiologist selects a population,
then categorizes everyone in that population by
whether he or she was exposed to one or more
risk factors of interest. Participants are
followed over time to determine whether
a particular health outcome actually develops. So here's an example
of a cohort study. Let's say you want to
find out whether taking a optional healthy eating
course at a high school has an impact on the food
choices that students make, and maybe even on their outcome,
such as obesity or diabetes. A cohort study could follow
a class, an the entire year of students, and you'd
separate those students into those who opted to take
the healthy eating course and those that did not. You then follow those students
over the following year and observe how their eating
patterns actually play out. And in fact, maybe even issues
related to obesity or diabetes. So this is an example
of a cohort study. And finally, the third study
type is the case control study. This kind of study
compares one group who has a disease or condition
with another group who does not. The first group,
the ill persons, we refer to as case patients. Those without the conditions,
we refer to as control subjects. The epidemiologist,
in this kind of study, then works backwards
from those who had illness who
didn't looking back to see if they had exposure
to certain risk factors or practiced certain
kinds of health behaviors. So for this example,
let me go back to the cruise ship investigation
that I discussed earlier. I conducted a case
control study and I looked at people who had the
diarrheal illness and those who did not. And I then looked
back and looked at their exposure, whether
they consumed the tap water or whether they consumed
only bottled water. So let's check what you've
learned during this section. Which of the
following are examples of a health related
source of data collection? The correct answers are B and
C. Intoxicated drivers arrests, and medical board actions
against the physicians, are non-health related
sources of data collection. Examine the terms on the
slide and match each study with correct case definition. Study one, subjects
with diabetes are compared with
subjects without diabetes. This is C, a case control study. A study of women,
aged 50 to 60 years, in a community located close
to a nuclear power facility, is what kind of study? This is A, a
cross-sectional study. And finally, subjects who have
received nutritional counseling and who have
exercised twice a week are compared with
subjects who have not. What kind of study is this? This is a cohort study. So for the last
portion of the course, we're going to review the steps
to investigate an outbreak. Epidemiologists do not
practice their craft in a vacuum or only
sitting at a desk. In fact, epidemiologist are
out in the field every day applying and putting
their expertise to use. A classic way that
epidemiologists apply their knowledge, is by
using the 10 steps involved in investigating an outbreak. These steps include establishing
the existence of an outbreak, preparing for field work,
verifying the diagnosis, defining and identifying cases,
using descriptive epidemiology, developing, evaluating,
and then refining hypotheses, implementing
control and prevention measures, and finally,
communicating findings. Now, only in theory, in
fact, do these activities of investigation roll out in
this nice, coherent order. In fact, it's often
more fluid and you need to be more flexible
based on the actual situation at hand. Now, let's go through
these 10 steps, going back to our Legionnaires
Disease investigation that we discussed earlier. So as a reminder, after such a
substantial number of American Legion conference attendees,
with similar symptoms, were admitted to hospital
emergency departments, or examined in doctors' offices,
an initial investigation began. During step one,
epidemiologists were able to establish the existence
of the Legionnaires Disease outbreak by reviewing records
and data from multiple sources that confirmed the
disease cases were indeed higher than normal from the
previous weeks or months. Step two involved
preparing for field work by researching the
outbreak, gathering supplies and equipment,
and preparing to travel. The epidemiologist also
consulted with other entities. For example, health care
providers, and members of the local health
department that they would be collaborating
with on this investigation, and finding out who
they should touch bases with when they arrived. During step three,
the epidemiologists ensured that the
problem was accurately diagnosed by speaking
with patients and by reviewing laboratory
findings and clinical test results. After the diagnosis
was confirmed, a standard set of criteria,
also known as a case definition, was established as step
four of the investigation. To determine whether a
person should be categorized as having the specific illness
they were investigating, or something else. Four components are
typically included in a case investigation-- uh,
excuse me, in a case definition, including,
one, clinical information about the disease. What signs and symptoms
have been observed? The second component
are characteristics about the persons
who are affected. Do you see any
commonalities among those who have become ill? The third, is information
about the location or place. Where are the affected
persons locating? Or where have they been? The fourth aspect are
specifications of time during which the
illness onset occurred. What date or time did
persons first become ill? And what was the
duration of the symptoms? During the outbreak, a case
definition typically evolves. Often, in the beginning,
it's pretty broad to try and make sure you're
finding all possible cases. But as you get more information,
the case definition generally gets more specific to be able
to accurately identify cases that are part of your outbreak. In this instance, to
assist epidemiologists in identifying early
Legionnaires Disease cases, public health nurses made rounds
of local Philadelphia hospitals to gather data about
those who became ill and to verify the
diagnosis according to the case definition. Meanwhile, laboratory samples
and clinical examinations were tested and
reviewed, respectively. After the initial case
definition was established, health care facilities,
such as doctor's offices, clinical laboratories,
and hospitals, were contacted to request that
any observations of illness matching the case definition
be reported to public health authorities. Step five used
descriptive epidemiology to describe the
Legionnaires Disease and orient the data by
identifying what, who, where, and when. After these were identified,
the epidemiologist proceeded to study the dates,
times, places, and persons, hopefully those words are
sounding familiar to you, and ultimately develop
their hypotheses about how and why people had become ill. This graph indicates the number
of Legionnaires Disease cases among conventioneers
and non-conventioneers by day in July and
August of 1976. Now remember, these
are the absolute number of cases, not the disease rate. You can see that the number
of cases peaks from July 25 through July 27,
indicated by the arrow, soon after the
convention begins, which is indicated
by the bracket on the bottom of the slide. By the end of the convention,
the number of cases had reached its peak and
then subsequently started to decline. However, this
descriptive analysis only tells us what
can be read directly from this graph and data. And analytic processes
take us one step further in testing any hypothesis
we would generate. You'll remember
from the discussion before that the investigation
suggested that the illness was somehow associated with
hotel A. This table lists the rates of illnesses among
American Legion delegates by age and where
they were staying, hotel A or other hotels. So let's see what
the data tell us. The rows, again, indicate
data specific by age group for the conventioneers. The first three columns
show the data for hotel A. The number of ill persons, the
total number of conventioneers, and the calculated percent ill. We know from before that
the highest percentage of ill persons was those
age 70 or older in hotel A, and this is somewhat,
although not 100%, true for the other
hotels as well. But by combining the information
from all of the age groups, those who stayed in hotel A
have the highest percentage of illness. 9% in hotel A versus 5.4%
and 6.8% at the other hotels. So we can infer, therefore,
that a connection exists between staying in
hotel A and becoming ill. We can also infer that older
persons may somehow be more susceptible to the disease. During steps six,
seven, and eight, focused hypotheses,
that is theories that could be
tested for validity, about Legionnaires
Disease were developed regarding how and why
the outbreak occurred. They were then evaluated for
validity and refined as needed. One hypothesis stated that
the illness was associated with convention attendees
who were guests at or who had visited
the particular hotel during their time
at the convention. Epidemiologists on
this investigation tested this hypothesis
in a series of ways. They conducted a randomized
telephone survey of guests registered at four hotels
in the area from July 6 through August 7,
including hotels where the conventioneers
had stayed and those where
they'd not stayed. They reviewed hotel
data by interviewing hotel employees who'd become
ill during the convention. They reviewed hospital emergency
department admissions data and collected
environmental samples from selected
locations in hotel A. They reviewed
weather data to look for any correlation between
specific weather events and the onset of illness. They interviewed hotel
guests and workers about places they visited,
people they met with, and foods they'd eaten
during the convention. And ultimately, they conducted
two case control studies. Although the
investigation suggested that exposure in the
outbreak apparently occurred over several
days, investigation results were actually initially unclear. Researchers were unable to
locate the bacteria that caused Legionnaires
Disease because of the historic difficulty of
growing bacteria like these in laboratory conditions. Five months after the
outbreak investigation of these cases of
Legionnaires Disease occurred, results finally indicated
that spending time in the lobby of hotel A was
the risk factor for illness. In December of 1976, a CDC
laboratorian successfully located the source
bacteria after continuing to test specimens that were
actually thought to be dead. A new bacteria that
had caused the disease was discovered in this process. After isolating the
bacteria, conditions at hotel were investigated to
determine their connection with the onset of the illness. Bacteria similar to this one
grow in warm waters in nature, such as hot springs,
and also had been associated with air
conditioning cooling systems. So investigators
concluded that those who had become ill
in Philadelphia had inhaled air from the
air conditioning unit at hotel A either at
guests or by walking past the lobby on the
sidewalk outside the hotel. They also noted that hotel
workers had not become ill and they believe that this
was because they actually had built up an
immunity over time to this particular bacteria,
which explained why the outbreak didn't
seem to affect them as much as the conventioneers. After the cause of the
outbreak was identified, investigators moved
on to step nine. To minimize the growth
of the bacteria that causes Legionnaires
Disease, controls were implemented,
such as chlorination of water and testing of
industrial air cooling systems. During step 10,
investigation findings were communicated to a variety
audiences, local health authorities, the medical
community, the general public, and lawmakers and other leaders. Usually, the findings are
reported in an oral briefing to health authorities and those
responsible for implementing control and prevention measures. During the briefing,
epidemiologists describe what they did and how
they did it, what they learned, and what they recommend should
be done about the illness. The findings are also written
up in reports that generally contain an introduction, a
background, methods, results, the discussion, and
finally, recommendations. And these reports serve
not only as a record of the investigation but also as
a tool for future investigators and as a contribution to the
scientific knowledge base. So let's answer a few
knowledge check questions that are based on what you've
learned about Legionnaires Disease and the 10 steps
involved in an outbreak investigation. In 1976, during an
American Legion Convention, 11 attendees had died of
apparent heart attacks by August 1. Dr. Campbell contacted the
Pennsylvania Department of Health after realizing
that he had treated three of those 11 attendees. What is the first step that
the Pennsylvania Department of Health should have followed? The first step would
be to establish the existence of an outbreak. CDC then launched
an investigation. However, no effective
communication existed between
scientists in the field interviewing patients and
those in the laboratory who were testing specimens. As a first step in
stopping this outbreak, what should the team have
done to identify persons who were part of the outbreak? The correct answer is B,
establish a case definition to identify cases. In speculating that the
cooling system might be the source of the
outbreak, what step was the epidemiologist
implementing? That was the process of A,
developing a hypothesis. In January 1977, the
Legionella bacterium was finally identified
and isolated, and was found to be
breeding in the cooling tower of the hotel's
air conditioning system. The bacteria then spread
through the building whenever the system was engaged. Which answer below best
describe what the investigation team should do regarding
their original hypothesis? The correct answer is
D, they should both A, evaluate it and B, refine it. The finding from this
outbreak investigation led to the development
of new regulations worldwide for climate
control systems. What step does this illustrate? This is B, implementing control
and prevention measures. Before we wrap up
the course, let's review what we've learned today. During this course, you
learned to define epidemiology, to describe basic terminology
and concepts of epidemiology, to identify types
of data sources, and basic methods of data
collection and interpretation. You learned to describe
a public health problem in terms of time, place,
and person, and to identify the key components of a
descriptive epidemiology outbreak investigation. I hope you've enjoyed
the introduction to epidemiology course. For additional information
about the topics we've covered, I've provided here resources
and additional reading. Thank you very much
for your attention.
