ABIM 2015 - General IM Flashcards
The degree to which the investigator’s conclusions are supported by a research study is called the study’s?
INTERNAL Validity
The generalizability of an investigator’s conclusions beyond the internal confines of the research study is called the study’s?
EXTERNAL Validity
In therms of what specifically is the error for a given study shown?
Confidence Interval (CI) - usually 95%
What does it mean when there is a “wide” Confidence Interval?
It means that the research study has less certainty and precission
What causes a research study’s Confidence Interval to be less precise or less certain (“wide”)?
Small sample size
The probability of detecting the difference between two groups when a difference exists is affected by a small sample size and is called?
Power of a study
Non-random errors that occur in a research study are called what?
Bias (can occur in selection of patients, measurement and analysis)
What type of bias exists when patients chosen for a study group have characteristics that can affect the results of a study?
Selection Bias
A third factor that influences BOTH exposure (treatment) and outcome (false conclusion: “smokeless tobacco in the form of snuff poses a greater risk for developing CAD than cigarette smoking” could occur because being MALE which is associated with a higher risk of CAD, is more likely among snuff users than cigarette smokers (male and female) thus distorting the cause-effect relationship. This factor is known as?
Confounder (MALE)
How can confounders be minimized in research studies?
By using a randomized study design
What type of study determines patient selection, treatment and analysis from the outset in order to minimize errors and bias by blinding patients, treating physicians and investigators?
EXPERIMENTAL Study
What is the objective of an EXPERIMENTAL Study?
To prove that the “new” therapy is “non-inferior” to existing therapies (but is cheaper, more convenient to take, etc.)
True differences between studied subjects can be masked if the sample size is too small, what is this called?
Low-power
If a study has great precision and internal validity however the treatment protocol is difficult to implement outside of the research setting, the study lacks what?
Generalizability (External Validity)
What is a study design called when randomization of patients is UNETHICAL or UNFEASIBLE however by its design, data can be compared in the same group of patients both BEFORE and AFTER the intervention?
Quasi-Experimental Study
What are studies called when instead of randomizing individual patients, entire GROUPS of patients are randomized?
Cluster-Randomized Study
In what types of studies does the study investigator have no role in assigning individuals to interventions but rather just compares the effects of exposures or treatments among two or more observed groups?
Observational Study
What are the weaknesses of Observational studies over Experimental studies specifically because in Observational studies, the investigator “observes differences among treated or exposed groups and has therefore NO role in randomizing them?
Confounding error and Bias
The ability to include a broader spectrum of disease (rather than very narrow spectrum of disease (well designed Randomized Control Studies), disease exposures that are rare and where treatments are administered in a “REAL WORLD” environment rather than in a lab, is the advantage of what type of study?
Observational Study
What types of studies are COHORT and CASE-CONTROL?
Observational Studies
The study that compares the outcomes of GROUPS forward in time (prospectively) or backward in time (retrospectively) with and without exposure or treatments NOT initiated by the investigator (rates of lung cancer between smokers and non-smokers) is called what?
Cohort Study
Which type of Cohort study, prospective or retrospective, minimizes a certain type of bias?
Prospective Cohort Studies (minimize “recall” bias - inaccurate recall of past events)
This type of study RETROSPECTIVELY compares the EXPERIENCE of patients who HAVE a disease with those who do not have the disease (patients with and without lung cancer can be compared with respect to their exposure to asbestos)?
CASE (disease)-Control Study
What study design is best for studying RARE diseases or those that occur many years after specific exposures?
CASE (disease)-Control Study