Module 6 for Final Exam (Ch. 6 & 7) Flashcards
What are the types of analytical studies?
Observational Studies and Experimental Studies
Examples of Observational Studies
- Cohort studies
- Case-control studies
- Cross sectional studies
- Ecological studies
Randomized trials are a type of…
experimental study
What is the gold standard for epidemiology research?
Randomized trials
what are reasons that researchers may not conduct a randomized trial?
Ethical issues, difficulties enrolling participants, high costs
What are characteristics of an experimental study (randomized trial)?
- the study researches preventions and treatment for diseases
- the investigator actively manipulates which groups receive the agent (the exposure) in the study
The ________________ is the group that is exposed to the agent in an experimental study.
treatment group
The ____ the group that is not allocated the agent in the experimental study.
comparison group
The comparison group in an experimental study could receive no treatment or:
a) inactive treatment (placebo)
b) another active treatment
c) any of the above
c) any of the above
The hallmark that distinguishes experimental studies from observational studies is the active manipulation of the agent (exposure) by the investigator. True or False?
True. Active manipulation of the agent by the investigator is a key feature of experimental studies.
Vitamins and behavioral modifications are studied to determine if they prevent or delay the onset of disease in healthy (or high risk) people in a ____ trial.
types of experimental studies
a preventative or prophylatic trial
Treatments like surgery or drugs are tested among people who already have a disease to improve quality of life or chances of survival in a ________ trial.
types of experimental studies
a therapeutic trial (or a clinical trial)
what are the basic steps in an experimental study?
- hypothesis is formed
- participants are recruited
- participants are randomly assigned to study groups
- groups are monitored for outcomes
- incidences of outcomes in groups are compared
A ____ is a research question or idea that you want to investigate. You conduct research to support or reject the ______.
A hypothesis is a research question or idea that you want to investigate. You conduct research to support or reject the hypothesis.
What does an institutional review board do in an experimental study?
The institutional review board (or ethics committee) reviews all studies to ensure that the research is ethical and legitimate.
In order for a study to be ethical, a state of ________ must exist. ____________ is characterized by genuine uncertainty about the risks and benefits of the test treatment.
In order for a study to be ethical, a state of equpoise must exist. Equipoise is characterized by genuine uncertainty about the risks and benefits of the test treatment.
There must be sufficient certainty that the test treatment may be beneficial to give to some people, but enough doubt must exist to justify withholding the treatment from other people.
which takes less time a preventative trial or a therapeutical trial?
A therapeutic trial takes a less time than a preventative trial. A preventative trial could be a long term study.
What is an example of an individual trial?
Types of experimental studies
A trial where thousands of individuals with HIV are assigned to different drug treatment groups.
What is an example of a community trial?
Types of experimental studies
An entire community having flouride added to their water and their health results being compared to another community without the flouride.
A parallel trial is when there is one treatment per study group and the groups receive the treatments in the same time period. True or False?
Types of experimental studies
True. In a parallel trial, each study group (the experimental group and the controls) each receive a different treatment in the same time period. The groups will stick with that treatment throughout the study. The experimental group may get a new treatment and the controls may get the standard treatment.
A crossover trial is when each study group goes through the different treatment options in different orders. True or False?
Types of experimental studies
True. The groups switch between multiple treatments in the study (with washout periods between treatments). The groups have the treatments at different times and there are no control groups.
(For example, one group starts with a new treatment, then switches to the old treatment. While another groups starts with the old treatment, then switches to the new treatment.)
What is a simple experimental design?
Types of experimental studies
In a simple design, each group receives one treatment (whether an old or new treatment).
What is a factorial experimental design?
Types of experimental studies
In a factorial design, two or more treatments are given to different groups and multiple hypotheses are tested. For example, there may be one round of randomization to assign controls and an experimental group for testing effects of Vitamin C. But then, each of those groups could be randomized again into additional controls and experimental groups to test effects of Aspirin.
The study groups could potentially be control while testing one hypothesis, but in the experimental group for other hypotheses.
The people who are potential participants in a study are called what?
The study population
The eligibility criteria to participate in a study is based on what types of considerations?
the eligibility criteria is based on scientific, safety, and practical considerations
In a trial testing an HIV drug that may prevent transmission from mother to child, who would you include as eligible?
Women who are HIV positive and are also pregnant
In a trial testing an HIV drug that may prevent transmission from mother to child, who would you exclude from the eligibility criteria?
- Pregnant women whose fetus has an abnormality that could be affected by the drug
- Pregnant women who medically need an antiretroviral drug who could not ethically potentially be assigned to a placebo group
Investigators are limited to enrolling people in studies by the number of eligible people who can actually be recruited to participate. True or False?
True. The number of people eligible for te study will be greater than the number of people enrolled.
Participants who are ill and participate in therapeutic trials are usually ______________.
Participants who are ill and participate in therapeutic trials are usually patients of a clinic, hospital, or physician that’s linked to the study.
Trials that enroll healthy participants recruit through:
a) hospitals and clinics
b) tv
c) newspapers
d) online
e) all of the above
Trials that enroll healthy participants recruit through
a) hospitals and clinics
b) tv
c) newspapers
d) online
e) all of the above
They may also recruit using billboards and radio
What is a single site trial?
A trial that happens a single clinical site
what is a multicenter trial?
A trial that involves 2 or more clinical sites. It is collaborative and involves more participants. Each site carries out the same experiment at their location and data is analyzed together.
For many therapeutic and clinical trials, investigators have to enroll a large number of people to detect whether a treatment has a meaningful effect on the outcomes of the treatment compared to the comparison groups. True or False?
True
what does statistical power mean?
Statistical power is the ability of a study to demonstrate an association if one exists. Studies need a high enough number of participants to have statistical power.
Do investigators expect a big or small improvement when comparing a new drug to an old one?
Investigators expect a small improvement
If investigators expect a small difference between the treatment group and the comparison group, do they need a large or small sample size?
They need a large sample size
What is informed consent?
Participants agreeing to be part of a study/trial.
Which of the following are the investigator’s responsibility in gaining informed consent?
* Use simple language
* Offer the potential participant snacks
* Disclose potential risks and benefits
* Discuss the study and study procedures
* Promise the potential study participant a discount on medications
* Describe how privacy will be maintained
* Inform the participant that they may withdraw from the study at any time
* Get verbal and written consent
- Use simple language
- Disclose potential risks and benefits
- Discuss the study and study procedures
- Describe how privacy will be maintained
- Inform the participant that they may withdraw from the study at any time
- Get verbal and written consent
________ and ____ assignments are the two general methods for assigning treatments to people.
nonrandom and random assignments
Nonrandom assignment methods can lead to ________ assignment. The most common nonrandom methods used is ________, where investigators alternate between assigning people to the treatment or comparison group based on order of enrollment. (1st person in treatment, 2nd person in comparison)
Nonrandom assignment methods can lead to biased assignment. The most common nonrandom methods used is alternation assignment, where investigators alternate between assigning people to the treatment or comparison group based on order of enrollment. (1st person in treatment, 2nd person in comparison)
Another type of nonrandom is assigning people based on the days of the week they are enrolled
Random assignment (or randomization) is better than nonrandom methods because _____________
it is less prone to bias
Assigning people to treatment or comparison group by coin toss is an example of?
Randomization
what are the goals of randomization?
1) to ensure that treatment assignment is unbiased
2) To achieve baseline comparability between compared groups on factors related to outcome (compared groups should have the same distribution of characteristics)
Randomization is best achieved using large groups. True or False?
True
Describe the process of using blocking to randomly assign people to groups
With blocking, study participants are enrolled slowly in small numbers and then evenly, randomly split into treatment and comparison groups. Example: Taking 6 months to enroll 100 people, enrolling 10 people at a time and then splitting them.
What are the advantages of blocking?
It ensures time of enrollment doesn’t influence whether someone is put in a certain group and controls for shifts in the nature of people enrolled over the course of the trial
Describe the process of using stratification to randomly assign people to groups
With stratification, study participants are separated by specific confounding characteristics. Then, within those groups, they are assigned to be part of the treatment or comparison group.
Examples: Stratifying by gender or age groups
Why is stratification helpful?
It ensures that confounding characterstics are equally divided among treatment and comparison groups.
A pharmacologically inactive substance given as a substitute for an active substance is called a what?
A placebo
a placebo pill is usually sugar coated and inert substance
what is a sham procedure?
It is a bogus procedure meant to mimic the legitimate one, so study participants cannot tell whether they are in the treatment group (who is receiving the experimental procedure) or comparison group.
A placebo is a good method of measuring…
treatment outcome and side efffects
Study participants, investigators, and healthcare providers are ____ if they don’t know whether the study participant is in the treatment group or not.
Study participants, investigators, and healthcare providers are masked if they don’t know whether the study participant is in the treatment group or not.
- In a single masked study, only the____ are masked.
- In a double masked study, the ____ and investigators who ____ are masked.
- In a triple masked study, the ____ , investigators who ________ are masked, and investigators who ________are masked.
- In a single masked study, only the participants are masked.
- In a double masked study, the participants and investigators who administer treatment are masked.
- In a triple masked study, the participants, investigators who administer treatment are masked, and investigators who monitor the effects of treatment are masked.
Masking of subjects and investigators helps prevent ____________, particularly endpoints that involve __________.
Masking of subjects and investigators helps prevent biased ascertainment of the outcome, particularly endpoints that involve subjective assessment (measures of pain, happiness).
What is the Hawthorne effect/halo effect?
When study participants’ condition improve just because of the attention received from being part of a study.
A study participant’s failure to adhere to the study protocol is called:
noncompliance
what are reasons for noncompliance?
- bad reaction to treatment
- lack of interest
- inconvenience
- want different treatment
- participant becomes clinically worse
Noncompliance can only occur in the treatment group. True or False?
False. Noncompliance can happen in the treatment group, comparison group, or both.
In the treatment group what does noncomplaince look like?
In the treatment group, noncompliance would be not following the protocol for treatment (not taking as directed) or not participating in the intervention.
In the comparison group what does noncomplaince look like?
Noncompliance in the comparison group would be those participants taking the treatment, doing the intervention, or changing their behavior.
Noncompliance makes the treatment and comparison groups more similar. True or False?
True
Ways to increase compliance
Do most studies have high levels of compliance?
No, most studies do not have a high level of compliance.
What does a run-in or lead-in period in a study accomplish?
A run-in or lead-in period is a test run of the study with potential study participants to observe their complaince with the regimen. It helps investigators eliminate study participants who tend be noncompliant.
The ouctomes of interest in a preventative trial are __________
The ouctomes of interest in a preventative trial are precursors of disease or incidence of disease
The ouctomes of interest in a therapeutic trial are __________
The ouctomes of interest in a therapeutic trial are symptom improve, length of survival, or disease reoccurence
The main condition that the trial has been designed to evaluate is know as the …
primary outcome
this is used for samle size calculation
The other outcomes of interest that are considered less important the main condition is known as …
secondary outcomes
Dropouts in a studay are…
people who withdraw from a study
Losses to follow up in a study are…
people who cannot be located or contacted by the investigator
Dropouts and losses to follow up negatively affect follow up. True or False?
True
What are the goals when ascertaining outcomes?
Having high follow up rates and uniform follow-up for all compared groups (the penalty of non-uniform follow up is bias)
What are some reasons for dropouts and losses to follow up?
- relocation
- waning interest
- adverse reactions to treatment
____ helps ensure accurate outcome data. ________ means investigators verify or corroborate the occurrence of the outcome using several sources.
Validation helps ensure accurate outcome data. Validation means investigators verify or corroborate the occurrence of the outcome using several sources.
this can be done by checking medical records, having a masked investigator verify outcomes
In intent-to-treat analysis, investigators analyze ____________
In intent-to-treat analysis, investigators analyze the data of all people in the treatment and comparison groups, regardless of noncompliance.
This is the classic data analysis approach
What are the benefits of intent-to-treat anaylsis?
- it preserves randomization
- it tells you about the effectiveness of the treatment
- it promotes unbiased results
Answers the question “how wel does treatment work among people who are offered it?”
In efficacy analysis, investigators analyze…
how well the treatment works among people who complete the treatment as directed.
This analysis ignores people who were noncompliant
In subgroup analysis, investigators analyze outcomes for different groups of interest. What is a potential problem with this type of analysis?
Problems with statistical power because the subgroups have less people in them than the general study population
Generalizability is …
the ability of a study’s conclusions to be apply to general medical care and a more diverse population than the study participants.
A study is done with a study population, but you want to apply the findings to ______________
the reference population
basically members of the general public who the results can be applied to
Is an observational study or experimental study best, if you want to gain knowledge about preventing or treating disease?
experimental study
Has high validity due to randomization
Is an observational study or experimental study best, if you you want to study all exposures, treatments, and preventions?
Observational study
less validity because the investigator has less control over extraneous factors
Most epidemiologic research is conducted using an ____________ study which takes advtange of the fact that people are regularly exposed to noxious and healthy substances.
Most epidemiologic research is conducted using an observational study which takes advtange of the fact that people are regularly exposed to noxious and healthy substances.
The two principle types of observation studies are?
case-control studies and cohort studies
A ________ is a type of observational study that looks at the risk and outcome relationship at a point in time in a defined population.
A cross-sectional study is a type of observational study that looks at the risk and outcome relationship at a point in time in a defined population.
does not give informationon past exposires and uses disease prevalence
A ________ is a type of observational study that looks at population level risk factors and health outcomes to determine exposure/outcome associations.
An ecological study is a type of observational study that looks at population level risk factors and health outcomes to determine exposure/outcome associations.
Potential problem is “ecological fallacy” which is when investigators make conclusions about individuals, but base them on population data
