Introduction to Epidemiology

[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.