SPH 102 - Final Practice Questions
What techniques used in experimental studies are also applicable for cohort studies? > Blinding (Masking) > Placebo > Randomization > Run-in period > Follow-up
> Blinding (Masking) > Follow-up
Selection bias is most likely to occur in what type of study? > Case-Control Study > Retrospective Cohort Study > Prospective Cohort Study > Experimental Study
> Case-Control Study > Retrospective Cohort Study
What is control selection bias? > Bias in cohort studies. > Controls do not represent the population with regards to exposure. > Occurs when criteria for selecting participants related to exposure are different for cases and control.
> Controls do not represent the population with regards to exposure. > Occurs when criteria for selecting participants related to exposure are different for cases and control.
What are two necessary requirements for control selection?
> Controls must come from the same source population as the cases. > Controls must be selected independently of exposure.
What is the main difference between differential and nondifferential misclassification? > Differential always biases towards the null and non-differential always biases away from the null > Differential can bias towards or away from the null and non-differential always biases towards the null
> Differential can bias towards or away from the null and non-differential always biases towards the null
If there is NO ASSOCIATION between the exposure and disease, then the Odds Ratio (OR) is > Greater than 1 > Less than 1 > Equal to 1
> Equal to 1
How is person-time calculated within the context of a cohort study? > Person is asked to join the study > Follow-up period starts > Person develops disease > Person is lost from study > Study ends.
> Follow-up period starts > Person develops disease > Person is lost from study > Study ends.
If the exposure is associated with INCREASED ODDS of disease, then the Odds Ratio (OR) is > Greater than 1 > Less than 1 > Equal to 1
> Greater than 1
Who can be blinded (or masked)? > Participants > Investigators who do the follow-up > Investigators who analyze results
> Investigators who do the follow-up > Investigators who analyze results
If the exposure is associated with DECREASED ODDS of disease, then the Odds Ratio (OR) is > Greater than 1 > Less than 1 > Equal to 1
> Less than 1
Which of the following concepts associated with cohort studies? > Open or Dynamic cohort > Fixed cohort > Retrospective cohort study > Prospective cohort study > Standardized mortality study
> Open or Dynamic cohort > Fixed cohort > Retrospective cohort study > Prospective cohort study > Standardized mortality study
Which of the following scenarios are best suited for cohort studies? > Rare exposure with little known about it > Rare disease with little knowledge > Study population is hard to follow-up > When you want to learn about multiple outcomes of an exposure
> Rare exposure with little known about it > When you want to learn about multiple outcomes of an exposure
Why should we do a case-control study for: Do concussions in youth increase likelihood of Alzheimer's disease? > The time from exposure to disease initiation is long. > Alzheimer's is rare- most people will never develop Alzheimer's Disease. > The cause of Alzheimer's is not well understood.
> The time from exposure to disease initiation is long. > Alzheimer's is rare- most people will never develop Alzheimer's Disease. > The cause of Alzheimer's is not well understood.
Among the following studies, which comparison group comes closest to the counterfactual ideal? > A study that compares male breast cancer rates among U.S. Marines who were stationed at Camp Lejeune, North Carolina, and were exposed to solvents in the local drinking water with breast cancer rates among men from the general U.S. population. > A study that compares male breast cancer rates among U.S. Marines who were stationed at Camp Lejeune, North Carolina, and were exposed to solvents in the local drinking water with U.S. Marines who were stationed at Camp Lejeune, North Carolina, but were not exposed to solvents in the local drinking water. > A study that compares male breast cancer rates among U.S. Marines who were stationed at Camp Lejeune, North Carolina, and were exposed to solvents in the local drinking water with U.S. Marines who were stationed at Camp Pendleton, California, and were not exposed to solvents in the local drinking water.
A study that compares male breast cancer rates among U.S. Marines who were stationed at Camp Lejeune, North Carolina, and were exposed to solvents in the local drinking water with U.S. Marines who were stationed at Camp Lejeune, North Carolina, but were not exposed to solvents in the local drinking water.
After reviewing the questionnaires, you find that out of 100 people who ate at the buffet,20% tested positive for S. typhi. You notice two exposures are most common - donuts and strawberries. 30% of all individuals ate donuts while 15% ate strawberries. Among those who tested positive, 40% reported eating donuts and 20% reported eating strawberries. What about the 30% who ate donuts and 15%who ate strawberries?
A+B
After reviewing the questionnaires, you find that out of 100 people who ate at the buffet,20% tested positive for S. typhi. You notice two exposures are most common - donuts and strawberries. 30% of all individuals ate donuts while 15% ate strawberries. Among those who tested positive, 40% reported eating donuts and 20% reported eating strawberries. Where would we put "100" for the total number of people in our study?
A+B+C+D
After reviewing the questionnaires, you find that out of 100 people who ate at the buffet,20% tested positive for S. typhi. You notice two exposures are most common - donuts and strawberries. 30% of all individuals ate donuts while 15% ate strawberries. Among those who tested positive, 40% reported eating donuts and 20% reported eating strawberries. Where do the 20 cases that are positive for typhoid fever go?
A+C
State the main advantages and disadvantages of case-control studies.
Advantages include that case-control studies are ideal for studying cases with very long periods of induction and latents, studying cases with rare diseases, using less time and money to conduct, and providing information on many risk factors. Disadvantages include that case-control studies face trouble in establishing the timing between exposure and disease due to its retrospective nature and that bias can be present and prominent.
What type of cohort study is being conducted below? A study of the relationship between pesticide exposure and subsequent risk of non-Hodgkin lymphoma began in 2016. In this study, men who worked as pesticide applicators in the 1970s-1980s were compared to men who worked as fertilizer applicators in the 1970s-1980s. The men were followed until 2015 to determine the incidence rate of non-Hodgkin lymphoma in each group. In 2017 the study investigators will receive additional funds to extend follow-up until 2027 so that a larger number of cases of non-Hodgkin lymphoma could be included in the analysis.
Ambidirectional Cohort Study
Describe one advantage and one disadvantage of using population controls in a case-control study.
An advantage is that the population controls originate from the same population as the cases, which allows for comparability. A disadvantage is that population controls do not have subjects that are as cooperative and compliant as hospital/clinic controls.
Why do investigators use the odds ratio used in a case-control study (instead of the risk or rate ratio) to measure the strength of the association between an exposure and a disease?
Because rate denominators are unknown in case-control studies, the count of exposed and unexposed subjects are needed to calculate the risk.
Consider the table below, what is the cumulative incidences for the exposed and unexposed columns?
CI (Exposed) = Disease/Total = 20/80 CI (Unexposed) = Disease/Total = 30/150
A study was conducted to identifyrisk factors for bovine cysticercosison farms in Switzerland from 2005to 2006. One group consisted offarms with infected cattle identifiedat slaughter in 2005 and 2006. Theother group was comprised ofrandomly selected farms with cattleslaughtered in the same period, butwith no evidence or history ofinfection. What kind of study was performed?
Case-Control Study
Typhoid fever is a severe gastrointestinal disease caused by the bacteria Salmonella typhi (S. typhi). S. typhi can bespread by fecal-oral. Often, typhoid fever is associated with contaminated food or water. You are an epidemiologist working for Epiland County. You notice that several people are testing positive for S. typhi at the local hospital. All state that they ate at a local buffet in the past week. However, you are concerned with what foods, drinks, or other exposures these people might have You are able to retrieve contact information for all patrons in the past week. Every individual fills out a food-borne illness questionnaire detailing their exposures from the restaurant. What type of study design is this?
Case-Control Study
State the main similarity and main difference between cohort and experimental studies.
Cohort and experimental studies compare multiple exposure groups and involve outcome rate metrics. Nonetheless, cohort studies have patients that choose their exposures while investigators choose the exposure in experimental studies.
Compared to unexposed subjects, subjects who were highly exposed to PCE-contaminated drinking water during gestation and childhood experienced increases in the risk of using two or more major illicit drugs later in life. There is a risk ratio of 1.6. What is the interpretation of the risk ratio?
Compared to unexposed subjects, subjects who were highly exposed to PCE-contaminated drinking water during gestation and childhood experienced 1.6-times the risk or a 60% increased risk of using two or more major illicit drugs later in life.
Which type of comparison group is being used? A study of the relationship between pesticide exposure and subsequent risk of chronic kidney disease began in 2015. In this study, Nicaraguan men who worked on a sugar cane plantation in the 1980s were compared to Nicaraguan men who worked in a gold mine in the 1980s. The men were followed until 2014 to determine the incidence rate of chronic kidney disease.
Comparison Cohort
Describe the strengths and weaknesses of the three types of comparison groups used in cohort studies. Which one comes closest to the counterfactual ideal?
Comparison cohorts can be comparative to the exposed group but sufferers from confounding exposures in the results, making it not really preferable. The general population is easy to obtain data from, but it is difficult to gather information and limit confounding factors. Therefore, the internal comparison group is the best option and the one most similar to the counterfactual ideal because the group originates from the same source population as the exposed group, maximizing comparability. The only issue is that the internal comparison group is difficult to identify and gather subjects.
Calculate the odds ratios for donuts and strawberries.
Donuts = (8 x 58) / (12 x 22) = 1.75 Individuals who ate donuts had 1.75 times the odds of testing positive for typhoid fever compared to individuals who didnʼt eat donuts during the last week. Strawberries = (4 x 69) / (16 x 11) = 1.56 Individuals who ate strawberries had 1.56 times the odds of testing positive for typhoid fever compared to individuals who didnʼt eat strawberries during the last week.
After reviewing the questionnaires, you find that out of 100 people who ate at the buffet,20% tested positive for S. typhi. You notice two exposures are most common - donuts and strawberries. 30% of all individuals ate donuts while 15% ate strawberries. Among those who tested positive, 40% reported eating donuts and 20% reported eating strawberries. How many sick individuals ate donuts? strawberries? Where should the numbers go?
Donuts: 20 x 40% = 8 Strawberries: 20 x 20% = 4 They should go in A in their respective tables.
What is the exposure and what is the outcome in the example below: Do concussions in youth increase likelihood of Alzheimer's disease?
Exposure: Concussions Outcome: Alzheimer's Disease
A case-control study is the most efficient design for studying the health effects of rare exposures, whereas a cohort study is the most efficient design for studying the risk factors for rare diseases. > True > False
FALSE
Bias is introduced primarily during analysis stage. > True > False
FALSE
Bias is introduced primarily during the analysis stage of a study. > True > False
FALSE
Case control studies must always have the same ratio of cases to controls (1:1). > True > False
FALSE
Case identification is generally more difficult than control identification in case-control studies. > True > False
FALSE
Including a large sample size reduces self-selection bias. > True > False
FALSE
It is possible to fix bias during the analysis phase of a study. > True > False
FALSE
It is possible to obtain a valid estimate of disease prevalence from a typical case-control study. > True > False
FALSE
It is possible to remove bias during the analysis phase of a study. > True > False
FALSE
Losses to follow-up can be a problem in a cohort study but not in an experimental study. > True > False
FALSE
Open-ended interview questions are the best way to ascertain exposure information. > True > False
FALSE
Poor recall and recall bias are synonymous terms for the same concept. > True > False
FALSE
Poor recall and recall bias are synonymous terms for the same concept. > True > False
FALSE
Special cohort studies are the most sensible design for examining many exposures in relation to a single disease. > True > False
FALSE
The control group in a case-control study should never include individuals who have the case's disease. > True > False
FALSE
The odds of illness are mathematically equivalent to the risk of illness. > True > False
FALSE
The prevalence rate of a disease can be calculated using a case-control study. > True > False
FALSE
The purpose of a control group in a case-control study is to provide information on the disease distribution in the source population that produced the cases. > True > False
FALSE
Using an inaccurate case definition increases the likelihood of nondifferential misclassification of the disease. > True > False
FALSE
This type of controls are selected because they are likely to share some of the cases' characteristics or habits but we need to be careful because this can lead to biased results if the study hypothesis involves these exposures.
Family/Friends Controls
Describe the three strategies—survivor sampling, case-base sampling, and risk set sampling—that investigators use to select controls.
For survivor sampling, researchers choose controls from survivors who did not transform into cases by case definition during observation. For base sampling, researchers choose controls from the at-risk population before observation begins. For risk-set sampling, researchers choose controls from the at-risk population while diagnosis occurs (cases are established).
Which type of comparison group is being used? A study of the relationship between pesticide exposure and subsequent risk of chronic kidney disease began in 2015. In this study, mortality rates from chronic kidney disease among Nicaraguan men who worked on a sugar cane plantation in the 1980s were compared to mortality rates from chronic kidney disease in the 1980s in the entire male population of Nicaragua.
General Population
Which type of comparison group is being used? A study of the relationship between pesticide exposure and subsequent risk of chronic kidney disease began in 2015. In this study, Nicaraguan men who worked as pesticide applicators on a sugar cane plantation in the 1980s were compared to Nicaraguan men who worked as irrigators on the same sugar plantation in the 1980s. The men were followed until 2014 to determine the incidence rate of chronic kidney disease.
Internal Comparison
Describe the situations in which it is desirable to conduct a case-control study.
It is desirable when the disease is rare, the disease have long periods of induction and latency, when the disease is not well-known, the population in question changes, and the exposure data is hard or pricey to procure.
What is the ideal but unattainable comparison group in a cohort study?
It would be the counterfactual ideal where the same individuals in the exposed group would not have the exposure.
Why is it important to minimize losses to follow-up?
It would otherwise diminish the statistical power of the study because there are less subjects in the results analysis. This can also lead to a bias analysis if those lost to follow-up have different results from those who remained for follow-up.
What is the main similarity and the main difference between experiments and cohort studies?
Main similarity: compare exposure groups and see incidence of outcomes. Main difference: ability or inability to assign treatment
Which of the following techniques that are commonly used in experimental studies can also be applied in cohort studies? > Masking > Placebo > Randomization > Run-in period > Intent-to-treat analysis
Masking
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the relative risk (RR) and provide an interpretation of the value.
N/A; this is a case-control study. Relative risk cannot be calculated.
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the risk of the exposed group (Re) and provide an interpretation of the value.
N/A; this is a case-control study. Relative risk cannot be calculated.
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the risk of the unexposed group (Ru ) and provide an interpretation of the value.
N/A; this is a case-control study. Relative risk cannot be calculated.
State whether or not a cohort study is best suited for each of the following scenarios: When little is known about a rare disease
No
State whether or not a cohort study is best suited for each of the following scenarios: When the study population will be difficult to follow
No
Odds Ratio
OD = AD/BC How many times higher the odds of getting disease in exposed individuals compared with the odds in unexposed individuals.
Consider the table below, what are the odds of illness among the exposed and unexposed groups?
Odds (Exposed) = Disease/No Disease = 20/60 = 0.33 Odds (Unexposed) = Disease/No Disease = 30/120 = 0.25
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the odds of cases in the exposed group.
Odds = a/b = 1.5 The odds of being a case were 1.5 times that of being a control in the exposed group.
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the odds of cases in the unexposed group.
Odds = c/d = 0.25 The odds of being a case were 0.25 times that of being a control in the unexposed group.
A pathogen has been reported in a species that has no history of previously being an affected species nor is it commonly considered in a clinical disease in this species in Northern California. Previous seroprevalence studies have reported differing background rates of this infection globally. The objective of this study was to evaluate possible associations between pathogen seroprevalence and clinical signs of disease within this species. To investigate this further, 720 individual samples within the species of interest with signs of disease were compared to healthy, non-symptomatic samples of the same source population and assessed for the presence of pathogen-specific antibodies. Seroprevalence among cases and controls was determined and represented in the 2x2 table below. Calculate the odds ratio (Odds of exposure in cases compared to controls) and provide an interpretation.
Odds Ratio = (a/c) / (b/d) = ad / bc = 5.94 The cases had 5.94 times the odds of being exposed than controls.
Consider the table below, what is the odds ratio?
Odds Ratio = ad/bc = (20 x 60) / (30 x 30) = 1.33 Compared to unexposed subjects, exposed subjects have 1.33 times the odds of getting the disease or a 33% increased odds of getting the disease.
Suppose that a case-control study was conducted in the United States to find out whether a Black woman's exposure to racism during pregnancy influenced her risk of giving birth prematurely. Investigators selected 500 cases who were hospitalized for premature delivery and 1,000 controls. The study found that 90 case mothers and 50 control mothers reported overt incidents of racism during their pregnancy. Calculate the odds ratio. Interpret the outcome.
Odds Ratio = ad/bc = (90 x 950) / (50 x 410) = 4.2 African American women exposed to racism during pregnancy had 4.2 times the odds of premature delivery versus African American women who were not exposed to racism during pregnancy.
How is person-time calculated within the context of a cohort study?
Person-time starts when follow-up begins and ends when either the subject dies, is lost to follow-up, or is positive for the outcome.
This type of controls is time consuming and expensive to identify and it may be difficult to obtain their cooperation because they usually do not have the same level of interest in participating as do cases and controls identified from other sources.
Population Control
Is retrospective cohort or prospective cohort is better in the following situation? You are concerned about the effect of bias.
Prospective Cohort Study
Is retrospective cohort or prospective cohort is better in the following situation? You want to know as much information about baseline characteristics of the exposed and unexposed groups.
Prospective Cohort Study
What type of cohort study is being conducted below? A study of the relationship between alcohol intake and subsequent risk of non-Hodgkin lymphoma will begin in 2017. In this study, men who drink alcoholic beverages will be compared to men who do not drink alcoholic beverages. The men, who are identified through drivers' license records, will be followed until 2027 to update information on alcoholic beverage consumption and determine the incidence rate of non-Hodgkin lymphoma in each group.
Prospective Cohort Study
Consider the table below, what is the risk difference of exposed to unexposed?
RR = (20/80) - (30/150) = 0.05 There is an excess risk of 5% or 5 more cases per 100 exposed persons.
Consider the table below, what is the relative risk of exposed to unexposed?
RR = (20/80) / (30/150) = 1.25 There is 1.25 times the risk or a 25% increased risk.
Relative Risk
RR = [a/(a + b)] / [c/(c + d)] How many times higher is the incidence of disease in exposed individuals as compared to unexposed individuals in that time frame.
Is retrospective cohort or prospective cohort is better in the following situation? You have a limited budget to study the effects of your exposure.
Retrospective Cohort Study
Is retrospective cohort or prospective cohort is better in the following situation? Your disease of interest has a long latent period.
Retrospective Cohort Study
What type of cohort study is being conducted below? A study of the relationship between pesticide exposure and subsequent risk of non-Hodgkin lymphoma began in 2016. In this study, men who worked as pesticide applicators in the 1970s-1980s were compared to men who worked as fertilizer applicators in the 1970s-1980s. The men were followed until 2015 to determine the incidence rate of non-Hodgkin lymphoma in each group.
Retrospective Cohort Study
Suppose that a case-control study was conducted in the United States to find out whether a Black woman's exposure to racism during pregnancy influenced her risk of giving birth prematurely. Investigators selected 500 cases who were hospitalized for premature delivery and 1,000 controls. The study found that 90 case mothers and 50 control mothers reported overt incidents of racism during their pregnancy. Set up the two-by-two table for these data.
See the image.
A retrospective cohort study is more efficient than a prospective cohort study for studying diseases with long latent and induction periods. > True > False
TRUE
A study must be valid before its results can be generalized. > True > False
TRUE
Bias can pull an estimate of association either toward or away from the null. > True > False
TRUE
Cohort studies are the observational equivalent of experimental studies, but the researcher does not allocate exposure but rather locates a "natural" experiment to observe relationship between exposure and disease > True > False
TRUE
General cohort studies are good for studying common exposures, whereas special cohorts are good for studying rare ones. > True > False
TRUE
If diseased and non-diseased subjects are not accurately classified as cases or controls, the results of the study may be biased. > True > False
TRUE
In a case-control study, we cannot calculate the cumulative incidence. > True > False
TRUE
In general cohort, researcher can study effects of multiple exposures on one or more outcomes. > True > False
TRUE
In special cohort, researcher can study effects of single exposure on one or more outcomes. > True > False
TRUE
It is preferable to use incident (that is, newly diagnosed) cases in a case-control study. > True > False
TRUE
The risk of illness does not necessarily equal the odds of illness. > True > False
TRUE
Using incentives to ensure high rates of participation in a study will decrease selection bias. > True > False
TRUE
When a positive association is biased toward the null, the true association is underestimated. > True > False
TRUE
Suppose that a case-control study was conducted in the United States to find out whether a Black woman's exposure to racism during pregnancy influenced her risk of giving birth prematurely. Investigators selected 500 cases who were hospitalized for premature delivery and 1,000 controls. The study found that 90 case mothers and 50 control mothers reported overt incidents of racism during their pregnancy. The investigators also ask you to describe the purpose of the control group in a case-control study. What would you tell them?
The purpose is to describe the exposure distribution in the population with the cases.
Suppose that a case-control study was conducted in the United States to find out whether a Black woman's exposure to racism during pregnancy influenced her risk of giving birth prematurely. Investigators selected 500 cases who were hospitalized for premature delivery and 1,000 controls. The study found that 90 case mothers and 50 control mothers reported overt incidents of racism during their pregnancy. Suppose that the investigators hire you as an epidemiological consultant to help them design this study. They ask you what type of control group is most appropriate for the study. Briefly describe the control group that you would advise them to select, and justify your choice.
We would advise women who experienced full-term deliveries in the same medical location as women with exposure. This control group would work because the controls and the cases are comparable with little bias, and we can establish that a premature delivery can be counted as a case.
State whether or not a cohort study is best suited for each of the following scenarios: When little is known about a rare exposure
Yes
State whether or not a cohort study is best suited for each of the following scenarios: When you want to learn about multiple effects of an exposure
Yes
