Statistics For Epidemiology Pdf

[Mark Tracy]

Asa field epidemiologist, you will collect and assess data from field investigations, surveillance systems, vital statistics, or other sources. This task, called descriptive epidemiology, answers the following questions about disease, injury, or environmental hazard occurrence:

Similarly, times of suspected exposures vary in their precision. With acute infections, poisonings, and injuries, you will often have precise exposure times to different suspected agents. Contrast this with chronic diseases that can have exposures lasting for decades before development of overt disease. Other relevant events supplementing a chronologic framework of a health problem include underlying environmental conditions, changes in health policy, and application of control and prevention measures.

Relating disease with these events in time can support calculation of key characteristics of the disease or health event. If you know both time of onset and time of the presumed exposure, you can estimate the incubation or latency period. When the agent is unknown, the time interval between presumed exposures and onset of symptoms helps in hypothesizing the etiology. For example, the consistent time interval between rotavirus vaccination and onset of intussusception (Table 6.1) helped build the hypothesis that the vaccine precipitated the disease (1). Similarly, when the incubation period is known, you can estimate a time window of exposure and identify exposures to potential causative agents during that window.

Graphs are most frequently used for displaying time associations and patterns in epidemiologic data. These graphs can include line graphs, histograms (epidemic curves), and scatter diagrams (see Box 6.4 for general guidelines in construction of epidemiologic graphs).

Contact diagrams are versatile tools for revealing relationships between individual cases in time. In contact diagrams (Figure 6.2, panel A) (5), which are commonly used for visualizing person-to-person transmission, different markers are used to indicate the different groups exposed or at risk.

Epidemic Curves

Epidemic curves (Box 6.5) are histograms of frequency distributions of incident cases of disease or other health events displayed by time intervals. Epidemic curves often have patterns that reveal likely transmission modes. The following sections describe certain kinds of epidemic situations that can be diagnosed by plotting cases on epidemic curves.

To approximate the time of exposure, count backward to the average incubation period before the peak, the minimum incubation period from the initial cases, and the maximum incubation period from the last cases. These three points should bracket the exposure period. If a rapidly acting intervention was taken early enough to prevent cases, discount the contribution of the last cases to this estimation.

Outbreaks can arise from common sources that continue over time. The continuing common source epidemic curve will increase sharply, similar to a point source. Rather than increase to a peak, however, this type of epidemic curve has a plateau. The downslope can be precipitous if the common source is removed or gradual if it exhausts itself. The rapid increase, plateau, and precipitous downslope all appeared with a salmonellosis outbreak from cheese distributed to multiple restaurants and then recalled (Figure 6.5).

A propagated pattern arises with agents that are communicable between persons, usually directly but sometimes through an intermediate vehicle. This propagated pattern has four principal characteristics (Box 6.6).

The epidemic curve accompanying the severe acute respiratory syndrome (SARS) contact diagram (Figure 6.2, panel B) illustrates these features, including waves with an approximate 1-week periodicity. Certain behaviors (e.g., drug addiction or mass sociogenic illness) might propagate from person to person, but the epidemic curve will not necessarily reflect generation times. Epidemic curves for large geographic areas might not reveal the early periodicity or the characteristic increase and decrease of a propagated outbreak. For these larger areas, stratifying the epidemic curves by smaller subunits can reveal the underlying periodicity.

An outbreak of dengue arising from a single imported case in a South China town reveals several of these features (Figure 6.6) (8). After the initial case, 15 days elapsed until the peak of the first generation of new cases. Control measures targeting the larva and adults of the mosquito vectors Aedes aegypti and A. albopictus began late in the first generation. The line indicates the rapid decrease in Aedes-infested houses (house index). A rapid decrease in dengue cases follows this decrease in vector density.

The epidemic curve for a zoonotic disease among humans typically mirrors the variations in prevalence among the reservoir animal population. This will be modified by the variability of contact between humans and the reservoir animal and, for vectorborne zoonoses, contact with the arthropod vector.

Epidemic curves from environmentally spread diseases reflect complex interactions between the agent and the environment and the factors that lead to exposure of humans to the environmental source. Outbreaks that arise from environmental sources usually encompass multiple generations or incubation periods for the agent. You should include on the epidemic curve a representation of the suspected environmental factor (e.g., rainfall connected with leptospirosis in Figure 6.7 [9]). In this example, nearly every peak of rainfall precedes a peak in leptospirosis, supporting the hypothesis regarding the importance of water and mud in transmission.

As an alternative to plotting onset by calendar time, plotting the time between suspected exposures and onset can help you understand the epidemiologic situation. For example, a plot of the days between contact with a SARS patient and onset of SARS in the person having contact indicates an approximation of the incubation period (Figure 6.8) (5).

To reveal distinctive internal patterns (e.g., by exposure, method of case detection, place, or personal characteristics) in time distributions, epidemic curves should be stratified (Figure 6.9). This puts each stratum on a flat baseline, enabling undistorted comparisons. Stacking different strata atop one another (as in Figure 6.7, which is not recommended) defeats attempts to compare the time patterns by group.

Temporal disease rates are usually illustrated by using a line graph (Box 6.4). The x-axis represents a period of interest. The y-axis represents the rate of the health event. For most conditions, when the rates vary over one or two orders of magnitude, an arithmetic scale is recommended. For rates that vary more widely, a logarithmic scale for the y-axis is recommended for epidemiologic purposes (Figure 6.10) (10). You should also use a logarithmic scale for comparing two or more population groups. Equal rates of change in time (e.g., a 10% decrease/year) will yield misleading, divergent lines on an arithmetic plot; a logarithmic scale will yield parallel lines.

For certain conditions, a description by season, month, day of the week, or even time of day can be revealing. Seasonal patterns might be summarized in a seasonal curve (Box 6.8). Stratifying seasonal curves can further expose key differences by place, person, or other features (Figure 6.12) (12).

The site is secure.

The ensures that you are connecting to the official website and that anyinformation you provide is encrypted and transmitted securely.

About 800 openings for epidemiologists are projected each year, on average, over the decade. Many of those openings are expected to result from the need to replace workers who transfer to different occupations or exit the labor force, such as to retire.

Epidemiologists are public health workers who investigate patterns and causes of disease and injury. They seek to reduce the risk and occurrence of negative health outcomes through research, community education and health policy.

Epidemiologists collect and analyze data to investigate health issues. For example, an epidemiologist might study demographic data to determine groups at high risk for a particular disease. They also may research trends in populations of survivors of certain diseases, such as cancer, to identify effective treatments.

Epidemiologists typically work in applied public health or in research. Applied epidemiologists work for state and local governments, often addressing public health problems through education outreach and survey efforts in communities. Research epidemiologists typically work for universities or in affiliation with federal agencies, such as the Centers for Disease Control and Prevention (CDC) or the National Institutes of Health (NIH).

Epidemiologists who work in private industry may conduct research for health insurance providers or pharmaceutical companies. Those in nonprofit companies often focus on public health advocacy instead of research, which is expected to be unbiased.

For more information on occupations that concentrate on the biology or effects of disease, see the profiles for biochemists and biophysicists, medical scientists, microbiologists, and physicians and surgeons.

Work environments vary because of the diverse nature of epidemiological specializations. Epidemiologists typically work in offices and laboratories to study data and prepare reports. They also may work in clinical settings or the field, supporting emergency actions.

Epidemiologists working in the field may need to be active in the community, including traveling to support education efforts or to administer studies and surveys. Because modern science has reduced the prevalence of infectious disease in developed countries, infectious disease epidemiologists often travel to remote areas and developing nations in order to carry out their studies.

Epidemiologists who work full time and typically have a standard schedule. Occasionally, epidemiologists may have to work irregular schedules in order to complete fieldwork or attend to duties during public health emergencies.

3a8082e126