1
Q
what is the purpose of diagnostic testing
A
comparing results of a diagnositc test to the reults of an established gold standard
2
Q
how are results from studies of diagnostic tests are reported in several forms
A
sensitivity and specificity
positive and negative prediction values
positive and negative likelihood ratios
receiver operator characterstic curves
3
Q
sensitivity
A
the number of illnesses or injuries that are correctly diagnosed by the clinical examination procedure being investigated
4
Q
how to calculate sensitivity
A
those diagnosed as having the condition by the clinical examination procedure/ those diagnosed as having a conidtion based on the gold standard
5
Q
specificiity
A
the number of indiviuals correctly classified as not having the condition of concern based on the test being investigated
6
Q
how to calculate specificity
A
those diagnosed as not having a condition by clinical exam procedure/ those diagnosed as not having a condition by the gold standard
7
Q
what do high sensitivities mean
A
- tests are good at ruling out a conditiom
- you will have more confidence that someone doesnt have the condition if a negative result is here
8
Q
what do high specificity mean
A
- tests are good at ruling in a condition
- it will not find the condition when it isnt there so you can be confident in a positive result
9
Q
positive prediction value
A
the probability that the condition is actually present if the test is positive
10
Q
negative predicted value
A
estimates that the target condition is not present when the diagnositc test is negative
11
Q
what does a high PPV indicate
A
that it would be efficient for detecting the condition
12
Q
what does a high NPV indicate
A
it would be efficient in not detecting the condition
13
Q
prevalence
A
proportion of a population who have or had a specific illness/condition in a given time period
14
Q
what happens to PPV and NPV values as prevalance changes
A
when prevalance falls PPV values also fall and NPV values will rise
15
Q
what is more stable estimates of results
A
sensitivity and specificity are more stable compared to PPV and NPV
16
Q
what are liklihood ratios
A
- values derived from estimates of sensitivity and specificity
- influence the level of certaintiy that a condition does or does not exist
17
Q
positive liklihood ratio
A
impact of a positive examination finding on the probability that the target condition exists
18
Q
how to claculate +LR
A
sensitivity / (1-specificity)
19
Q
what does a large +LR indicate
A
helps to rule in a disorder
20
Q
negative liklihood ratio
A
impact of a negative exmaination on the probbaly that the condition in question is present
21
Q
how to calculate -LR
A
(1-sensitivity)/specificity
22
Q
what does a small -LR mean
A
helps to rule out a disorder
23
Q
what does an LR > 1 mean
A
indicates probability that the target/problem/disease is present
24
Q
what does an LR < 1 mean
A
indicates a decreased probability that the target/problem disease is present
25
what does an LR = 1 mean
no change in the probability
26
what does probabilty imply
uncertainty
27
pre test probability
refers to when a patient comes to see there is a level of probabilty that they will have the condition
28
how to calculate pretest odds
pretest probabilty/(1-pretest probability)
29
how to calculate posttest odds
pretest odds*LR
30
what is a receiver operating characteristic curve
ROC
- helpful for determining a cut off score for these types of decision when tyring to find out how many positive special tests are required to determine a diagnosis
31
what does an area under ROC curve of 0.5 mean
useless test
its no better at identifyin true positives than flipping a coin
32
what does an area under ROC curve of 1.0 mean
one that has zero false positives and zero false negatives
33
what is the best choice for a cut off for giving a balance between SN and SP
when the point of the ROC curve turns
34
what is the difference between reliabilty and validity?
- being reliable is when results are pretty consistent but not ideal
- valid but not reliable will have results not clustered but all around the mark
- valid and realible is consistent and desired outcomes
35
internal validity
- validity of a study's experimental design
- if an experiemnt can conclusively demonstrate that the indepent variable has a definite effect on the dependent variable
36
confounding variables
extraneous factors that may result in false relationships
37
what can influence validity
bias and blinding
38
selection bias
characteristics that subject have before they enroll in a study may influence the results of the study
39
deliminations of bias
decisions that investigators make to improve the internal validity of their studies
40
what is a way decrease bias
blinding
41
what are the ways that blinding can occur
subject may be blind
members of the experimental team
clinicans who are treating patients
42
how can a subject be blinded
whether they are receiving an experimental treatment or control treatment
43
how can a member of the experimental team be blinded
blinded to the group assignment of individual subjects and the values of previous measurements for individual subjects
44
how can clinicians be blinded
blinded to the group assignments of individual subjects
45
external validity
relates to the degree to which the results of a study are generalizeable to the real world
46
what affects external validity
tightly controlled subject selection, administration of interventions, and control of confounding factors are less generalizable to general population
47
what are the types of internal validity
face validity
content validity
construct validity
48
face validity
examines if an instrument appears to measure what it is supposed to measure
49
what is the weakest form of measurement validity
face validity
50
content validity
examines if the items within an instrument adequetly comprise the entire content of a given domain reported to be measured by the instrument
51
what are the domains of content validity
fine motor
visual perception
expressive language
receptive language
52
consruct validity
examines if an instrument can measure an abstract concept
53
what are the validity of measures
criterion validsity (concurrent and predictive)
54
criterion validity
a validity of measures
how well the test performs and whether it is useful when judged against a standard
55
concurrent validity of measure
refers to how well one measure is correltated with an existing gold standard measure
56
reliability
the consistency of a specific measurement
57
intratester reliability
the ability of the same tester to produce consistent repeated measures of a test
58
what is intratester realiabilty also known as
intrarater reliabilty
test-retest reliability
59
intertester reliability
abilty of different tester to produce consistent repeated measures of a test
60
what is intertester reliability also called
interrater reliability
61
whta is the common measure for reliabilty
ICC
62
pearsons r
assesses the association betweent wo continious measures across a sample of subjects
63
what is r value used to measure
the stregnth of the association between pairs of data that are interval or ratio
64
how do you estimate the association between multiple variable
multiple regression (R^2 value)
65
what is an R^2 value
variance in the criterion variable explained by a predictor variable
66
what is the range of values for an r value
-1 - 1
67
what is the range of values from an R^2 value
0-1
68
what indicates strong correlation
values closer to 1
with pearson r values, it can have strong negative correlation or strong positive correlation
69
what are measures of association used for
- to help build useful predictive models
- they cant necessarily imply cause and effect
70
precision of measurement
how confident one is in the reporducibility of a measure
71
how is precision reported as
standard error of mean (SEM)
72
what is used most to analyze ordinal data
spearman rank order correlation
73
what is used to analyze nominal data
Chi square calculation
74
what are nonparametric statistics
statistics that do not assume a distrubtion
spearman r and cramers V
75
what is STARD
standards for the reporting of diagnostic accuracy studies
76
what is STARD
its a checklist for manuscripts to help ensure that other people can reproduce the study
77
what is the framework of risk identification and injury prevention
- establishing the extent of the injury prolem
- establish etiology and mechanisms of sports injuries
- introducing a preventative measure
- assessing its effectiveness by repeating step 1
78
in the framewokr of risk identification what is the goal of step 1
- estbalishing the extent of the injury via collecting data on incidence and severity
- done through prospective injury surveilance
79
in the framework of risk identification how does step 2 work
- identify the mechanism of injury
- identify risk factors that may be involved in the etiology of a specific injury
80
in the framework of risk identification how does step 3 work
intervention should be based on information gained in the first two steps of the paradigm
81
in the framework of risk identification how does step 4 work
if severity or incidence of injuries have been substantially reduced implementation of intervention is likely warranted
82
what is the gold standard for studies of injury risk factors
prospective cohort
83
what studies are retrospective in nature
case control studies
84
prevalence of an injury or illness
proportion of a sample that has a given injury or illness at a single time point
85
how to calculate prevalance
(injured subjects/total subjects) * 100
86
incidence
the number of new cases of the pathology in a given period of time
87
incidence rate
the number of new cases that occur per unit of person time at risk
88
incidence proportion
the number of newly injured individuals in a defined population over a given period of time
89
injury risk
refers specifically to the probability of new injury per individual
90
injury rate
specfically refers to the number of new injuries per unit of exposure time
91
prevalence ratio
the ratio of the injury prevalence estimates between two groups is taken
92
risk ratio
the ratio of injury risk estimates between two groups is calculated
93
rate ratio
the ratio of injury rate estimates betwen two groups is determined
94
relative risk
provides a proportion of injury incidence between two groups and is identical to the calculation of risk ratio
95
what is used to assess injury risk between two groups
relative risk
96
how is relative risk reduction calculated
1 minus the relative risk and multiplying by 100
97
relative risk increase
experimental condition is found to lead to heightened risk of injury
97
relative risk reduction (RRR)
represents the percentage that the experimental condition reduces injury risk compared to the control condition
98
absolute risk reduction
difference betwen the injury rate/risk in the intervention group and the injury rate/risk in the control group
99
how is the absolute risk reduction interpreted
as a percentage it is the reduction in number of injuries per 100 people who received the intervention
100
number needed to treat
represents the number of patients that need to be treated with the experimental treatment to prevent one injury compared to receiving the control condition
101
how is the number needed to treat calculated
taking the inverse of the absolute risk reduction comparing two groups
102
what is the ideal number to treat
1
103
what is an odds ratio used for
compares the odds of a condition occuring in one group compared to another
104
how are data categorized
nominal
ordinal
interval
ratio
105
nominal data
assignment of numeric values for analysis of nominal data is arbitrary
106
ordinal data
orderd in a particular and meaningful manner
107
interval data
data that is measured along a scale with no absolute zero
108
ratio data
data that is measured numerical that has an absolute zero
109
nonparametric data
methods of comparison to analyze nominal and ordinal data
will not be associated with a distribution
110
parametric data
statistics that are used for analyzing interval and ratio data
111
what are measures of dispersion for interval and ratio data
variance and standard deviation
112
what are measures used for reporting ordinal data
median and range
113
what is the analysis of variance
compares means between 2 or more groups when the outcome is a continious variable
114
what does it mean when its a one way or two way ANOVA
refers to the number of IVs or factors studied for the impact on the DV
115
what are T-tests
special cases of ANOVA in which there are only two sets of data in the comparison
116
what are the types of t-tests
dependent t-test and independent t-test
117
what does a one way ANOVA test for
whether the varibales being compared are equal or at least one is different
118
what are two way ANOVAs
2 independent variables
119
how to interpret an interactions graph
if two lines cross it probably means there is some type of interaction (non-parallel lines)
120
what is F
- a point on a distribution
- the ratio of explained/unexplained variance)
- larger F = les likely that differences observed where chance occurances
121
what do you do with the hypothesis if the F value is larger
reject the null hypothesis
differences observed were due to the effects of our intervention
122
what is a post Hoc analysis
- method to find group differences
- only done when the ANOVA F-test is significant
123
what does a significant F test mean
evidence that not all group measn are euqal but it doesnt identify where the differences exist
124
what are types of Post Hoc tests
Tukey, Scheffe, Bonferroni procedures
125
MANOVA
multivariant analysis of varinace or cases where more than one dependent measure is analyzed simultaneously
126
Type I error
null is rejected when in fact differences do not exist
127
Type II error
null hypothesis is not rejected but the study of the population finds diffrences between groups
128
what is the probability that a type I error will occur
based on the alpha level/level of significance
129
what is the alpha level normally set at
0.05, meaning that this i sthe acceptable level of risk for making a type I error
- the chances are less than 5% that you have an oddball sample or false positive results
130
what is type I error also
the level of significance
131
what is the probablity a type II error will occur
beta, which is often set at 20% chance of finding a false negative
132
what is statistical power
1-beta, a measure of how powerful an experiment is
133
what does it mean if a type II error is made
power is not adequete
134
what is power
helps to determine the number of subjects needed in a study to detect a statiscially significant difference
135
what does a power of 80% means
that 80% of the time, the researcher will avoid a type II error
136
what is required to decrease the risk of a type II error
statistical pwoer
137
what is power influenced by
the mean difference between groups
the variance within groups
sample size (the only thing that investigators can control
138
does statistical significance indicate clinical significance
no
139
what does statistical significance depend on
the magnitude of the treatment effect, sample, size, reliability of treatment effect, and reliability of the measurements
140
what does a confidence interval indicate
the true effect about 95% of the time
- e.g. there is a 95% chance that the sample mean is between 55.4 and 68.6 if the 95% CI is reported as 68.6+/- 6.6
141
what does a confidence interval off 99% indicate
include the true effect about 99% of the time
142
what is the formular for 95% CI
sample statistic +/- (2 * standard error)
143
what does a CI give
plausible range of values for a population
precision of an estimate
statistical significane
144
minimally clinically important difference
- MCID
- value established taht identigies the change in a helath status measure associated with improvements meaningful to patients
145
what are ways to estimate MCID
- ROC analysis
- standard error measurement values
- effect size estimates
146
effect size
calculation that shows the strength of the intervention
147
what is the most common way to calculate effect size
cohens d
148
what does a large effect size indicate
larger difference in results of intervention to control group
149
what does 0.2-0.49 effect size of Cohen 's D mean
intervention has a small effect
150
what does 0.5-0.79 effect size of Cohen's D mean
moderate effect
151
what does >= 8 effect size of Cohen's D mean
large effect
152
what is the binomial effect size display
- calculation for relative success rate
- helps to further interpret clinical significance when using effect size r
153
how to calculate success rate of intervention group
SR = 50% + (r/2)
154
how to calculate success rate of control group
SR = 50% - (r/2)
155
how to calculate relative success rate
SR of intervention/SR of control
156
what are some procedures for nonparametric data
Mann-Whitney U
Kruskal-Wallis one way analysis
Friedman Two way analysis of variance
157
what is the Mann-Whitney U test
- analogous to paried t-test
- reported as a T value
158
what is a Friednman Two-Way analysis of variance by ranks
analysis of repeated measures within one group over time
159
at is the Kruskal-Wallis One-way analaysis of variance by ranks
test when there are more than two groups
160
the effect size is an important tool for making sense of research results because
it tells us the magnitude of the statistical effect
161
what is the gold standard for research design
large scale, multisite, Randomized control trials
162
what may be a more feasible reserach design
smaller scale RCTS because they were less expensive
163
what is a clinical trial
the research performed on patients to assess an intervention
164
what is a prior power analysis
the number of subjects involved understanding the oucome data and statistics
165
what does the a priori power analysis look at
the level of significance (alpha value)
the statistical power (beta)
the effect size of the experiment
the expected variability in the data
166
what are effect sizes and variability based on
the pilot data and reported findings from existing literature
167
CONSORT
consolidated standards of reporting trials
168
what is CONSORT
- evidence based, minimum set of recommendations for reporting randomized trials
- checklist that guides authors what they should be looking for in strenght of a paper
169
PEDro Sclae
physiotherapy evidence database
170
what is the PEDro scale for
help to identify whihc of the known or suspected RCTs are likely to be internally valid and has sufficient statistical information to make results interpretable
171
what is a good value for PEDro scale
>= 6
PTD 504 EBP
flashcards
Decks in class (7)
# Cards
