1
Q
research designs
A
- experimental
- quasi-experimental
- nonexperimental
- case report/anecdote
2
Q
experimental designs used to
A
examine cause + effect relationships b/w independent variable and an outcome (dependent variable)
3
Q
independent variable
A
predictor variable
could be:
- condition
- intervention
- characteristic
4
Q
dependent variable
A
outcome variable
*response or effect that is presumed to vary depending on the independent variable
5
Q
randomized controlled trials
A
- clinical in nature
- attempt to determine if an intervention or preventative measure improves significantly compared to a control group
6
Q
quasi-experimental designs
A
- only one subject group or in which randomization to more than one subject group is lacking
- controlled manipulation of subjects is preserved
7
Q
nonexperimental (observational) designs
A
- controlled manipulation of subjects is lacking
- if groups are present, assignment is predetermined based on naturally occurring subject characteristics or activites
8
Q
case report/anecdotes
A
- detailed description of the management of a patient that may serve as a basis for future research
- describes overall management of an unusual case or condition that is infrequently encountered in practice or poorly described in literature
9
Q
what is the gold standard for intervention studies
A
randomized controlled (clinical) trial
*reduces bias by ensuring similarity b/w groups
10
Q
randomized controlled (clinical) trials are designed to evaluate
A
efficacy : effect of treatment under highly controlled conditions
11
Q
recruiting a study sample
*consecutive
A
all patients referred to a given location with this problem
12
Q
recruiting a study sample
*selective
A
based on referrals; less ideal but common and practical
13
Q
recruiting a study sample
*prospective
A
recruited before intervention
*easier to control through observation, measurements, interviews
14
Q
recruiting a study sample
*retrospective
A
recruited after intervention
*usually via medical records; harder to fit inclusion/exclusion criteria
15
Q
attrition
A
participants leave a study
16
Q
intention-to-treat analysis
A
- statistical analysis that considers attrition
- subjects analyzed in groups to which they were initially assigned even if they did not receive their assigned treatment
17
Q
sensitivity (Sn)
A
SnNout =
- high sensitivity
- negative test result
- rule out disorder
18
Q
sensitivity (Sn) example
A
lachmans test - 84.6% Sn
19
Q
specificity (Sp)
A
SpPin =
- high specificity
- positive test result
- rule in disorder
20
Q
specificity (Sp) example
A
lachmans test - 95% Sp
21
Q
positive likelihood ratio (LR+)
A
likelihood that a positive test result was observed in a person with the disorder
22
Q
negative likelihood ratio (LR-)
A
likelihood that a negative test result was observed in a person with the disorder versus a person without the disorder of interest
23
Q
LR+ > 10 or LR- < .10
A
large + conclusive change
24
Q
LR+ = 5-10 or LR- = 0.10-0.20
A
moderate change
25
LR+ = 2.5 or LR- = 0.20-0.50
small but sometimes important change
26
LR+ = 1-2 or LR- = 0.50-1.0
negligible change in pre-test probability
27
if LR = 1.0
represents 50:50 chance of increasing or decreasing the probability of a diagnosis
28
prognosis
process of predicting the future about a patients condition
*identifies risk of developing a future problem
*predict outcomes of impairments or functional problems
*predicts results of PT interventions
29
prognosis - 3 elements
- outcomes that are possible
- likelihood that outcomes will occur
- time frame for outcome achievement
30
prognosis - personal factors
sex, age, previous injury, current symptoms, etc
31
prognostic factors
influence the outcome of a condition
32
prognostic factors can only be determined by
assessing individuals who already have the condition
33
risk factors
associated with CAUSING a condition
34
case-control study
*retrospective comparison of 2 groups
- 1 group w/ disorder , 1 group w/o
- look at proportion of each group who had the risk factor or prognostic indicator of interest
35
cohort study
*prospective comparison of 2 or more groups before they have the disorder
- monitor groups to see who develops the disorder + identify what characteristics they have
36
predictive model
*retrospective or prospective data
- regression model determines which relevant factors predict outcome of interest
37
odds ratio
odds that an individual with the prognostic or risk factor will develop the outcome of interest
38
odds ratio - if outcome is negative
- OR > 1 : odds in favor of adverse outcome (problem will develop)
- OR < 1 : odds against adverse outcome (prognostic factor is protective against problem)
39
odds ratio - if outcome is positive
- OR > 1 : odds are in favor of positive outcome
- OR < 1 : odds are against positive outcome (prognostic factor is harmful)
40
risk ratio (RR)
ratio of risk for developing the adverse outcomes in patients with risk factor vs patients without risk factor
*comparable to odds ratio but depends on incidence of outcome in both groups
41
risk ratio (RR) - most commonly used for adverse outcomes :
- RR > 1 : increased risk of adverse outcome
- RR < 1 : decreased risk of adverse outcome
42
p-value
probability that the result occurred due to chance
43
95% confidence inverval
range of values within which the true value is estimated to lie within a 95% probability
44
if OR = 1.0
represents 50:50 chance of increasing or decreasing the odds that the outcome will occur
45
if RR = 1.0
represents 50:50 chance increasing or decreasing the risk that the outcome will occur
46
a 95% CI for odds ratio or relative risk that include 1.0 in its range means :
one possibility for the true value of the OR is a 50:50 chance
47
most to least filtered info
*filtered info
- systematic reviews
- critically-appraised topics
- critically-appraised individual articles
*unfiltered info
- randomized controlled trials
- cohort studies
- case-controlled studies
- case reports
48
narrative review
- general or focused topic
- summary of findings from individual studies
- may or may not have established search + selection methods
- limited to no critical analysis of individual studies included in review
49
systematic review
- focused question
- summary of findings from selected individual studies
- rigorous search + selection methods established ahead of time
- extensive critical analysis of individual studies included in review
50
systematic review purpose
to draw a conclusion from the cumulative weight of the evidence
51
research design
- subjects = individual studies
- inclusion and exclusion criteria
- included studies undergo standardized eval process
52
meta-analyses
- statistical analyses of data pooled form individual studies included in review
- interventions and data compatibility from individual studies are necessary
53
benefits of meta-analyses
- increase statistical power
- improve estimates of effect size
- reduce uncertainty
- improve external validity
54
rater reliability
- interrater
- intrarater
55
types of measurement validity
- face validity
- content validity
- construct validity
- criterion validity
56
face validity
- does measurement appear to assess what is intended
- addressed from the standpoint of the tester and/or the standpoint of the patient or family member
57
content validity
- extent to which measurement is judged to reflect the meaningful elements of a variable
- judged by content experts or people with experience with the variable
58
criterion validity
- extent to which one measure is systematically related to other measures or outcomes
- requires direct comparison of index measure with a standard measure or with known outcome
59
construct validity
- validity of abstract concepts that underlie the measure
- achieved via operational definitions, logical arguments, theoretical arguments, research evidence
60
mean
- sum of observations divided by total number of data points
- uses info from all data elements
- extreme values may skew
61
median
- middle score
- score above which half of distribution lies
- calculated diff. for odd vs even # of data points
- cannot be skewed by extremes
62
mode
- score that occurs most frequently in distribution
- often used for nominal data
63
range
- difference b/w highest and lowest values in distribution
- can be represented as a single score or in range format
64
standard deviation
- how much, on average, scores deviate from the mean of the data set
- uses same units of measurement as original score or data
65
coefficient of variation
- allows comparison of the variability in different measures of the same phenomenon
- standard deviation/mean expressed as a %
- shows extent of variability in relation to the mean of the population
*lower coefficient of variation is better and demonstrates LESS variability amongst the measures
66
power
likelihood that one will detect a difference (relationship) when one exists
67
methods for increasing power
- control extraneous variables + apply methods consistently
- use homogeneous groups
- increase sample size
68
sample size determinants
- desired level of power
- alpha level to be used in research
- estimate of effect size
(clinically meaningful relationship or difference between groups)
69
interrater vs intrarater
interrater : how consistent are different raters with each other?
intrarater: how consistent is one rater with themselves?
UTHSC EBP
flashcards
Decks in class (3)
# Cards
