Chapter 6: Study Designs – Flashcards
Unlock all answers in this set
Unlock answersquestion
This type of study seeks to identify possible causes of disease by finding out how the two groups differ with respect to an exposure
answer
Case-control Study
question
The unit of observation and the unit of analysis for case-control and cross-sectional studies
answer
Individual
question
The unit of observation and the unit of analysis for ecologic stuies
answer
Group
question
Sources of Controls (3)
answer
Population-based control, Patients from the same hospital as cases, Relative or associated of cases
question
A type of measure of association used in a case-control study. Signifies the "ratio of the probability of occurrence of an event to that of nonoccurrence."
answer
Odds
question
What does an OR of 1.0 mean?
answer
Implies that the odds of exposure are equal among the cases and control. Also suggests that a particular exposure is not a risk factor for the disease in the study.
question
What does an OR of 2.0 mean?
answer
Indicates that the cases were twice as likely as the controls to be exposed. Also may imply (if properly considered for causality) that this particular exposure is a risk factor for the disease.
question
If an OR is less than 1, what may be implied?
answer
May indicate that exposure may be a protective factor of the disease. However, it cannot be inferred if the confidence interval contains 1.0.
question
Also termed "Prevalence study". Starts by selecting a sample of subjects and then determining the distribution of exposure and disease.
answer
Cross-Sectional Studies
question
Type of experimental design studies where there is a manipulation of the study factor, but subjects were not randomized. An example: Community Trials
answer
Quasi-Experimental Studies (Given example)
question
A common type of experimental design where the exposure of interest is controlled, and the subjects are randomly assigned to the conditions of the study.
answer
Clinical Trials
question
Examines a group as a unit of analysis. The rate of exposure and rate of disease are known. The number of non-exposed persons and noncases may be inferred.
answer
Ecological Study
question
Involves an assessment of the correlation between exposure rates and disease rate among different group/populations over the same time period. Also called cross-sectional ecological studies.
answer
Ecological Comparison Studies
question
Involves the correlation of changes in exposure with changes in disease over time within the same group/aggregate unit to ascertain trends.
answer
Ecological Trend Studies
question
The bias that may occur because an association observed between variables on an aggregate level does not necessarily represent the association that exists at an individual level.
answer
Ecologic Fallacy
question
In case control studies, what are the benefits of studying incident (as apposed to prevalent) cases?
answer
1- subject's better recall of past exposure 2-reduced likelihood that exposure has changed as a consequence of the disease ( prevalent cases create challenges in separating causal exposures from consequential exposures)
question
In case control studies, what's the main caveat in the selection of cases derived from hospitals and tertiary facilities?
answer
Nonrepresentativeness of the cases (these institutions may receive only the most severe cases)
question
What are the two methods of matching the cases and controls?
answer
1-matched pairs ( individual matching) 2- frequency matching ( group matching)
question
True or False: 1-case control studies are useful for investigation of rare disease 2- case control studies may be insufficient , if the exposure is rare in the population
answer
True
question
The textbook provides an example of a situation where the residents of Palm Springs California (a desert resort) have higher mortality rates from emphysema than residents of LA (a highly industrialized area). If we concluded based on this information that areas with lower pollution levels have higher emphysema mortality rates what would this be an example of?
answer
Ecologic fallacy
question
The important characteristic of ecologic studies is that the level of exposure for each individual in the unit being studied is __________
answer
unknown
question
If you choose to sample every person whose name starts with J from a master list, what kind of sampling is this?
answer
Systematic sampling
question
If you want to derive estimates of the magnitude of a health problem in a relatively small subset of the population, you should use this type of sampling which requires that the population be divided into mutually exclusive and exhaustive strata
answer
Stratified sampling
question
A _______ sample has the characteristic that every element in the population has a nonzero probability of being included in the sample
answer
Probability
question
Name 3 examples of exposure data, as listed in the lecture notes/video
answer
1 aggregated measures (ex literacy rates) 2 environmental measures (ex levels of mercury) 3 global/group measures (ex birth rates)
question
True or False? When a new health concern emerges and relatively little is known about it, it is best practice to conduct costly, resource intensive studies to investigate before the concern worsens.
answer
False (as knowledge increases and the complexity of research questions increase, more rigorous study designs may be merited)
question
A community intervention that is oriented towards smoking cessation is an example of which type of study?
answer
Quasi-experimental
question
What are the two main subtypes of observational studies?
answer
1-descriptive studies 2-analytical studies
question
If in a 2 by 2 table you start by selecting a sample number and then determine each subject's exposure and disease status, which approach are you using?
answer
Cross-sectional study
question
What are the 2 major types of ecologic studies?
answer
1 - ecologic comparison; 2 -ecologic trend
question
True or false? The principle weakness of cross-sectional designs arise from their limited usefulness for inferring disease etiology
answer
True
question
True or false? A control is always free of disease
answer
False
question
Suitable controls should have the potential to become ________
answer
Cases
question
True or false? Cross-sectional studies are well suited to study diseases of low frequency
answer
False (a major limitation is that they are not)
question
One of the major limitations in cross-sectional studies is that disease and exposure histories are taken at the same time. Why is this potentially problematic?
answer
Temporality issue- it is difficult to determine whether the disease or exposure came first
question
How would you calculate the proportion of cases not exposed in this table?
answer
75/(75+25)
question
How would you calculate the number of exposed cases in this table?
answer
25 (just look at cell a)
question
How would you calculate the odds of exposure among the case group in this table?
answer
25/75
question
How would you calculate an odds ratio to evaluate whether the odds of exposure among the case group are different from the odds of exposure among the control group in this table?
answer
(25 x 95)/ (5 x 75)
question
Based on the odds ratio that you conducted from this table, what would you conclude about smoking as a risk factor for lung cancer?
answer
The lung cancer cases were about 6.33 times more likely to have smoked than the controls, which implies a strong association between smoking and lung cancer
question
True or false? Case-control studies are often the method of choice in infectious disease research
answer
True
question
According to the lecture slides what are three major limitations of case-control studies?
answer
1 selection bias 2 use of indirect estimates of risk 3 can only measure one outcome
question
Under what conditions is an odds ratio a good approximation of the risk associated with a given exposure?
answer
1 the controls are representative of the target population (in the frequency of exposure of interest) 2 the cases are representative of all cases (regarding severity and diagnostic criteria) 3 the frequency of the disease in the population is small
question
According to Dr. Moraros' video capture lecture on cross-sectional studies, what is the #1 advantage of this type of study?
answer
They are quick and cheap to conduct
question
What is the ideal ratio of cases to controls if you want to maximize statistical power?
answer
1:4 (cases to controls)
question
How would you classify an association with an odds ratio between 1-1.2?
answer
No association
question
How would you classify an association with an odds ratio between 1.2-1.5?
answer
Weak association
question
How would you classify an association with an odds ratio between 1.5-3?
answer
Moderate association
question
How would you classify an association with an odds ratio between 3-10?
answer
Strong association
question
How would you classify an association with an odds ratio greater than 10?
answer
Extremely strong association
question
A ____________ is associated with the exposure and with the outcome, but is not present in the causal pathway between the exposure and the outcome
answer
confounder
question
What are 4 uses of cross-sectional studies? (As listed in the lecture notes)
answer
1 hypothesis generation 2 intervention planning 3 estimate the magnitude and distribution of a health problem 4 infer causation
question
According to the chapter 6 video capture lectures, most of what we know about the population is determined with which kind of study?
answer
Cross-sectional
