PEOPLE'S EPIDEMIOLOGY LIBRARY
PEL ESSAYS ON EPIDEMIOLOGY
S.D. Walter, McMaster University, Hamilton, Canada
Essay # 1:
Introduction to Epidemiology
This essay provides a brief summary of what epidemiology is, what it is used for and the kind of people and skills required to do it. We will touch on the history of the subject, and identify some examples of important findings from epidemiology studies.
Definition and uses of epidemiology
Epidemiology is the study of health and disease in populations. [Porta, 2008] Epidemiology studies are used to identify which people might be more or less likely to get certain diseases, whether the rates of disease are changing over time, and whether some geographic areas of a community might be at higher risk of disease than other areas.
In the next essays, it will become clear how epidemiology uses comparisons between groups of people, and thereby thinks in terms of populations, to arrive at its conclusions. The results of epidemiologic studies are very important for the administration of health care systems. Public health authorities frequently use this information to monitor the state of health of a population. They may also base public health programs for the early detection or prevention of disease on epidemiology findings.
The epidemiologic approach is equally important for clinical medicine. Then, the populations studied are patients, their diagnosis, prognosis and therapy. It provides much of the input needed in evidence-based medicine.[Sackett et al., 1996] An evidence-based framework is often used to develop reliable guidelines and provides advice to doctors about which are likely to be the best diagnostic tests and treatments for their patients.
Epidemiologists are usually involved in the design of research studies, in data collection for those studies, and in the analysis and interpretation of the results. They are also increasingly involved in knowledge transfer activities, which are communicating the results of their studies to assist health care policy makers, clinicians, and the general public.
Epidemiologic evidence is fundamental in the development of recommendations to the community about healthy lifestyles, and there can also be commercial implications, such as in developing the health messages and labels that may be attached to foods and other consumer products such as tobacco. Likewise, clinical epidemiologic evidence is the basis of developing guidelines for clinical practice.
Skills required to do epidemiology
The discipline of epidemiology involves many different aspects. Typically, epidemiologic studies require the expertise of doctors, biologists, biochemists or other scientists to inform the biological aspects of the disease or clinical question being studied.
Many epidemiologic studies are highly quantitative, and they draw on the skills of biostatisticians and information technologists to design appropriate ways to capture and store the data that is needed, and to carry out statistical analysis of the results. Additionally, qualitative approaches are often used in epidemiology, for instance, involving social scientists who determine important themes in the perception of a particular disease concern, and how the community might respond.
Because many scientific and other disciplines are required in epidemiology, the background and training of epidemiologists varies considerably. Some epidemiologists begin their careers in a laboratory science (for instance, chemistry or microbiology), but later move into doing epidemiologic studies in the general community.
Similarly, many doctors, nurses and other health professionals begin working in clinical practice, but then go on to work in epidemiology and public health, treating the whole population as their patient, and focussing on patient populations. Still others, such as biostatisticians, may begin their training in mathematics and statistics, and become involved in epidemiology through the development of appropriate study designs and methods of data analysis for epidemiologic data.
Epidemiology in history
Epidemiology has both a long and a short history. It might be argued that Hippocrates (a variant of this famous oath is taken by new doctors almost all over the world) was an epidemiologist, because he had a hypothesis that some diseases were caused by environmental factors that act upon human populations.
On the other hand, epidemiology as we know it today has a short history since there were no real attempts at comparisons between groups to consider illnesses in populations before the 17th century. In the 17th century John Graunt carried out a series of analyses on the newly implemented “Bills of Mortality” in London, which for the first time systematically determined which were the leading causes of death.[Graunt, 1662] Lists of numbers of deaths from various causes, as they were understood in those times, including examples such as “Frighted”, “Griping in the Guts” and “Worms”, were published every week as a kind of early warning signal for the onset of plague. More importantly, from our present-day view, he showed that overall mortality evolved smoothly from year to year while mortality from plague demonstrated peaks, suggesting that it had an environmental cause that came and went.
Another well known pioneer or epidemiologic thinking was James Lind who compared six pairs of sick seamen with different treatments in each member of the pair and observed that oranges and lemons were effective treatment for scurvy.[Lind, 1853]
In London, some 200 years after Graunt, the epidemiologist John Snow’s studies compared the mortality from cholera between the clients of two companies providing water in London.[Snow, 1855] The study confirmed his hypothesis that getting cholera was somehow linked to contaminated drinking water. The specific inquiry of a cluster of cases which occurred in the vicinity of Broad Street famously led to the removal of the handle of the Broad Street water pump. This was a controversial move at the time, because several other theories about the cause of cholera existed, including the notion that living at a low altitude would increase the population’s exposure to dangerous vapours. We know nowadays that Snow was right, and the provision of a clean and secure drinking water supply was one of the main elements of a major movement towards improving public health during the 19th century and subsequently.[Vandenbroucke et al, 1991]
Other important examples of 19th century epidemiology include the study of Dr Pierre Louis on the efficacy of bloodletting in the treatment of pneumonia [Louis 1835] and the intervention of Ignaz Semmelweis demonstrating the effectiveness of hand washing in the prevention of puerperal fever [Carter, 2005][Semmelweis, 1861]
More recently, in the mid-20th century, a number of epidemiologic studies were central in the discovery that smoking could increase the risk of lung cancer. Some compared the frequency of smoking in cases of lung cancer and carefully chosen controls [Wynder and Graham, 1950][Doll and Hill, 1950][Levin et al., 1950] Other compared the occurrence of lung cancer in smokers and non smokers.[Link to [Doll and Hill, 1954, 1964] [Hammond & Horn, 1958]
Although this idea is now well accepted, at the time it too was controversial and much contested by scientists [Fisher, 1959;Stolley, 1991] and to Berkson [Berkson, 1946] and the tobacco industry.[US surgeon general report, 1964] Epidemiology studies over the years since then have confirmed the existence of numerous adverse health effects of smoking, including, heart disease and several other types of cancer. [Pettiti, 2001]
Epidemiology in the community
Use of the epidemiologic method is an important practical way to study health and disease, because epidemiologists work in the general community.[Holland et al., 2007] Epidemiologists can be found working in the general population, based in hospitals or local health units, or working for government or commercial enterprises such as the pharmaceutical industry. They all have in common that they are working on studying health events in various groups of people, making comparisons over time and between various places. Laboratory scientists and research doctors often use experimental designs for their studies, focusing on one particular risk-factor and one outcome, in such a way that all other risk-factors are suitably controlled, often by the use of randomisation: random numbers, i.e., pure chance determines who will receive what treatment.
In contrast, epidemiologists may be engaged in studying several diseases and several risk factors for those diseases simultaneously, and sometimes refined study designs are required to do so. Epidemiology is characterized by attempts to deal with the numerous determinants of disease in the community; epidemiologists are often concerned with trying to isolate the association between one risk factor and one particular health outcome while taking into account other important associations that exist in people’s daily lives. Epidemiologists do not limit themselves to factors that are measured at the individual level. They also measure factors at the community level, and integrate both type of information in their analysis.[Susser, 1998, McMichael, 1999 Diex-Roux,1998] They also take into account the fact that some factors may have different effects at different stages of life, and that the effects of exposures at a younger age may only become manifest at an older age.[Kuh & Shlomo, 2004]
While important and useful information can emerge from laboratory studies and randomised clinical trials of treatment, the findings from those studies may not be generalisable to the population at large, and here epidemiology has an advantage by directly studying what is actually occurring in the general population. Furthermore, many health-related questions of interest, for example the health effects of smoking, cannot be studied in humans using experimental randomisation, and we are therefore obliged to consider the problem by comparing groups of people, using the study designs that are explained in essay.
In some special circumstances, epidemiologists can use randomisation in the community. For instance, in order to evaluate the potential benefits and harms of screening for the early detection of cancer, they might randomise portions of the population or particular individuals to be offered screening or not, and then observe their subsequent experience of cancer detection and prognosis.[Shapiro et al., 1988] However, this type of study is unusual because of the large sample sizes and extended follow-up periods that are required to do them.
Instead, epidemiologists are more usually concerned with studying health effects in people as they live their daily lives but without the ability to randomise people to categories of risk exposure. Studying health effects in people who go about their usual daily lives is one major advantage over other scientific methods, but it is also one that presents epidemiology with some of its greatest challenges. In the community, epidemiology is instrumental in detecting and tracking epidemics of infectious diseases, like flu, HIV, measles of outbreaks like food poisoning.
Epidemiology in the clinic
One of the greater triumphs of epidemiology was the acceptance of its instruments, i.e., its study designs, as ’arbiters’ in clinical medicine: a great impact was made when the randomized controlled trial became accepted as the gold standard for deciding which therapy works best.[Streptomycin Trial, 1948] Similarly, adverse effects of treatments are nowadays commonly researched using epidemiologic studies, i.e., the same study designs as were used to assess how much smoking causes lung cancer.[James Lind Library]