Assessing Quality of Medical Literature Flashcards
describe the tool BICEP
-used for assessing the sources of error in a research study
B=bias I=internal validity C=Confounding variables E=external validity P=power
what is bias and what are the different types
factors arising from the design and conduct of a particular study that skews the data in a particular direction
selection bias, observer bias, participant bias, withdrawal bias, recall bias, measurement bias, publication bias, lead time bias
what is internal validity
- are assessment tools truly measuring the things they are intending to measure?
measuring blood pressure only in a doctor’s office may not be a good indication of overall blood pressure control
- this statement is a question of:
internal validity
what are confounding variables
are the variables which the research failed to measure impacting outcome of the study?
An unmeasured variable may show spurious associations, eg a racial group may appear to be genetically predisposed or protected from disease, when dietary differences should be responsible
-this statement shows the problem of:
confounding variables
what is external validity
-are the study outcomes relevant/valid in the “real world”
a study finds a very small but statistically significant decrease in blood pressure after meditation
-this statement is a question of:
external validity
what is power?
is the study able to detect an association of effect if one is present?
- size of study is usually the most important determinant of power
- inadequate power can result in type II errors, false negatives
a study evaluating an intervention to prevent a rare illness includes too few subjects (or is too short) to detect a difference between groups
-this statement is a question of:
power - the study is not powerful enough and could lead to false negatives
what is a surrogate marker?
any laboratory or physical results that is thought to predict a clinical outcome
ex: cholesterol levels, ECG findings, plaque in arteries
what is a clinical outcome?
sometimes called patient centered outcome
- direct measures of how a patient feels, functions, or survives
ex: stroke, myocardial infarction, death
how are surrogate markers and clinical outcomes used to assess study outcomes
- surrogate markers are easier to measure (studies are smaller, shorter, or less expensive)
- clinical outcomes are generally harder to measure- studies often need to be larger and quite long (esp of incidence of outcome is low)
which is considered superior? studies assessing surrogate markers or clinical outcomes?
clinical outcomes
-surrogate markers do not always predict clinical outcomes
discuss primary, secondary, and composite outcomes
primary= those identified as most important
secondary=of lesser importance
composite=several separate outcomes combined into a single outcome variable
