epi final exam

Below is a table from a hypothetical cohort study of 3,000 cigarette smokers and 5,000 non-smokers following them for the development of coronary heart disease (CHD). Calculate the etiologic fraction for the contribution of smoking to CHD.

((84/3000) - (87/5000))/ (84/3000), then multiplied by 100 = 37.85%

Calculate the odds ratio using the following data:The following data is from a case-control study, examining if babies with heart defects were more likely to have mothers who used retinoic acid (to treat acne) during their pregnancy than babies born without heart defects.

(a*d)/(b*c) - (yes retinoic "heart defects"X no retiniocic"no hert defect)/(yes ret "no heart" Xno ret "heart defects")

Individuals are selected for inclusion into a case control study based on:

- Cases are selected for inclusion with the disease under study - Controls are selected who are comparable to cases but must not have the disease under study

Some of the problems or issues with using SMRs for research include:

- Everyone in a study population will eventually die, so the longer you follow it a group, the more likely the SMR will tend to equal 100% over time - As with directly adjusted rates, SMRs can also hide differences in subgroups - SMRs of working populations typically exhibit the healthy worker effect

Which factors may modify a toxic agent's effect on human health?

- Host factor such as age and genetics - Concentration and transport of the agent in the environment

The purpose of double-blinding in experimental studies is:

- reduce error that results from the way in which the outcome is assessed. - reduce error that results from subjects' knowledge of their assignment to study conditions.

A cohort study yields a Risk Ratio of 2.4 Fill in the blank by selecting the correct interpretation. "Subjects with exposure have a ____ greater risk of developing disease than subjects without exposure.

140%

Working backwards, if an exposure increases the risk of disease by 40%, the risk ratio must have been:

40% = 40/100 = 0.40 0.40 + 1 = (a risk ratio) of *1.40*

Which is NOT a method or source for selecting population based controls for a case control study?

A list of all friends and co-workers

The following sentence summarizes the main finding from a prospective study: "People who received the vaccine were 0.33 times as likely to contract the disease than those who were not vaccinated." This corresponds to:

For this question, since the RR is less than 1, this indicates a protective effect. Start by substracting the null effect from your estimate. (0.33 - 1= -0.67. ) The negative value confirms that this is a protective effect. Now take the absolute value and multiply it by 100 to get the %: 0.67 X 100 = 67% reduction in risk.

Which area of the human skin tends to be most absorbent with respect to substances?

Genitals

A reservoir may be:

Humans Other non-human vertebrates Soil and water

Below is a table from a hypothetical cohort study of 3,000 cigarette smokers and 5,000 non-smokers following them for the development of coronary heart disease (CHD).Calculate the risk ratio using the data presented above.

Incidence in exposed = (84/3000) =(developed/total) Incidence in the non exposed = (87/5000). (developed/total) So, the risk ratio = (84/3000)/(87/5000) = 1.61

Prospective cohort studies typically begin by selecting:

Individuals with and without the exposure of interest.

Based on the highest relative risk identified in the previous question, you developed a short list of possible pathogens that could be the culprit. Lab tests confirm that E.coli is to blame, and samples taken indicate that 130 of the shelter residents were infected with it.What is the infectivity of E.coli in this outbreak?

Infectivity = # infected / # susceptibles * 100 = 130/250 * 100= 52%

Which kind of a design is a case control study?

Observational and analytic

Cohort studies are good design for studying rare exposures (risk factors), but a bad design for studying rare diseases.

True

Compared with other common study designs used by epidemiologists, RCTs rank highest in internal validity but often suffer from low external validity.

True

Finding a statistically significant difference in a large study reveals a true difference, but the findings may not be clinically important.

True

Given the incubation period and typically lower gastrointestinal symptoms, the pathogen causing this outbreak is most likely an infection (as opposed to an intoxication).

True

Nativity refers to place of origin or birth of an individual.

True

Prophylactic trials are designed to prevent disease. Therapeutic trials are designed to treat or cure disease.

True

What is the virulence of E. coli in this outbreak?

Virulence = # with severe disease / # with clinical disease * 100 = 3/58 * 100 = 5.17%

Mold in home

allergic reactions

Arsenic in drinking water

bladder cancer

The best way to reduce observer bias and any impact of the subject's perception due to knowledge of their treatment status is through

double blinding (masking).

Increasing sample size and the number of measurements taken usually results in an increase of random error.

false

The healthy worker effect refers to the observation that:

healthy persons are more likely to gain employment or remain employed than unhealthy persons.

Inhalation of asbestos particles

mesothelioma

A studying reporting an SMR of 200% indicates that:

the study population has a death rate that is double that of the standard population.

The lowest dose of a toxic substance at which a particular response (e.g., symptom or disease) may occur:

threshold

Pesticides (a type of chemical agent) have been linked to breast cancer, sterility, and birth defects.

true

Experimental designs do not typically suffer from:

Recall bias

What is the null value for a risk difference?

0

When laying out a "2 x 2" table to correspond to our standard formulas, health outcome status (as opposed to exposure status), should always go:

Across the top

Which of the following may be represented by a tall, pointed population pyramid?

All of the above may be represented by a tall, pointed population pyramid

In a cohort study, the comparison population may come from:

An internal population An external population Population incidence rates

The last choice offers an interpretation for the etiologic fraction.

Approximately xx % of the incidence of CHD in subjects who smoke could be prevented or reduced if the subjects did not smoke

A pathogen (or agent) of infectious diseases may include:

Bacteria Parasite Viruses

The best way to prevent known confounders is through

Blocking

The third choice offers an interpretation for incidence rate data, not incidence proportion data.

CHD could be prevented or reduced in approximately xx per 1000 subjects who smoke if they did not smoke.

Examples of descriptive epidemiologic studies do not include:

Case Control Studies

WHICH STUDY DESIGN AM I? To determine what household factors pose the greatest risk for unintentional burns in children in rural Bangladesh, a hospital-based study was conducted in which 240 children with burn injuries were selected and compared with 240 children with non-burn injuries. To find out about their home environments, the children's parents were asked about the type of stove, heater, lamps, kitchen design (e.g., separate kitchen with a door), and other related factors present in their homes at the time of their child's injury. The children were then compared by exposure and burn/non-burn status to determine if children with burns were more likely to have lived in a home with a specific type of heating/cooking appliance and kitchen arrangement than children with non-burn injuries.

Case-control study

Which of the following satisfy Hill's criteria of temporality?

Case-control study Prospective cohort

Select the correct interpretation of an Odds Ratio of 5.0:

Cases were 400% more likely to have been exposed than controls - 5.0 - 1 = 4.0 X 100 = 400%

Descriptive studies, which describe health-related states or events by person, place, and time allows us to do all of the following except:

Confirm causal associations.

For this question, plot the epicurve for the following data, then select the type of curve it best represents. Day 1 = 0 casesDay 2 = 4 casesDay 3 = 6 casesDay 4 = 2 casesDay 5 = 7 casesDay 6 = 5 casesDay 7 = 11 casesDay 8 = 8 casesDay 9 = 10 casesDay 10 = 5 cases

Continuing source transmission

WHICH STUDY DESIGN AM I? In order to determine the prevalence of leishmaniasis infection (a disease spread by the bite of the female sandfly) among children and factors associated with its transmission in a high incidence area for the disease in Ethiopia, a sample of 600 children aged 4-15 years were surveyed using a serological test to determine their Leishmania status and a questionnaire covering socio-demographic, health and dietary factors. Associations between Leishmania status were examined for a number of factors including if the child slept outdoors, if the child lived in a straw roofed house, and if the family owned sheep.

Cross-sectional study

When you are finished, pick the item with the highest relative risk. That item is most likely to have caused the outbreak. Did you wade through the flood water? Exposed (45, 110), Non Exposed (11, 84) Did you swallow any flood water? Exposed (11, 68), Non Exposed (45, 126) Did you drink the shelter milk? Exposed (35, 82), Non Exposed (20, 113) Did you eat the shelter salad? Exposed (27, 78), Non Exposed (30, 115) Did you eat the shelter hamburger? Exposed (54, 159), Non Exposed (3, 34)

Did you drink the shelter milk? Wade through flood water: (45/155=29.0%)/(11/95=11.6%) =2.5 Swallow flood water:(11/79 =13.9%)/(45/171=26.3%)= 0.53 Drink shelter milk: (35/117=29.9%)/(20/133=15.0%)= 1.99 Eat shelter salad: (27/105=25.7%)/(30/145=20.7%) = 1.24 Eat shelter hamburger:(54/213=25.4%)/(3/37)=8.1%= 3.14

In case control studies, matching is used to address

Differences in known confounders between cases and controls

With respect to broad categories of epi study types, which kind of a design is an RCT? (Note: you may need to refer to previous lecture notes to determine this.)

Experimental and analytic

An SMR of less than 100% indicates that the study population has a higher rate of death/illness than the comparable standard population.

False

In mechanical vector-borne transmission, it is common for the agent to multiply or undergo some physiological change within the vector.

False

It's rare that we see age related differences or variation in specific disease rates. In other words, people are just as likely to get a disease when they are old as when they are young.

False

Prophylactic trials are designed to treat (i.e., reduce) or cure disease.

False

healthy persons are more likely to gain employment or remain employed than unhealthy persons.

False

In experimental studies using factorial designs, all patients receive treatment and also serve as their own controls.

False - this is cross over design

Calculate the overall attack rate for the outbreak. Assume all shelter residents were susceptible to the infection.

Overall attack rate = # clinical cases / # susceptibles * 100 = 58/250 * 100 =23.2 %

Which of the following is an example of Selection bias?

Participation bias Exclusion bias

What is the pathogenicity of E. coli in this outbreak?

Pathogenicity = # with clinical disease / # infected * 100 = 58/130 * 100 = 44.6%

Which offers the best interpretation of the risk ratio calculated in the previous question?

People who smoke are xx% more likley to develop CHD compared to non-smokers

Randomized controlled trials (also known as Randomized Clinical Trials or RCTs) occur in phases. Which phase tests a treatment (typically, a new treatment) against other therapies or a placebo?

Phase III

Changes in the frequency and pattern of disease over long periods of time (10+ years).

Secular trends

The second choice offers an interpretation of a risk difference, not an etiologic fraction.

Subjects with smoking exposure have an incidence of CHD that is xx% higher than subjects without smoking exposure.

An odds ratio that is LESS than 1.0 (i.e., 0.35, 0.45, 0.72, etc. ) indicates

The exposure is associated with a reduction in risk for the health outcome

An risk ratio (RR) that is equal to 1.0 indicates:

The exposure is not associated with any risk or benefit for the disease under study

Which of the following may be true about this finding: OR=4.75 (95% CI: 0.14,18.76)

The study may have had an inadequate sample size (not enough people in the study)

Mercury in fish

neurological effects