1
Q
What is evidence based practice based on?
A
Patient values
Clinical expertise
Best avilable evidence
2
Q
What is the ultimate goal of clinical research?
A
Maximize the effectiveness of clinical practice
3
Q
Qualitative Research
A
measurement is based on subjective info
Obtained using: focus groups, interviews, observations
Purpose: describe the state of conditions, generate hypothesis, explore associations
4
Q
Quantitative Research
A
Measurement outcome utilize numberical data under standardized conditons
info obtained: formal instruments that address physical, behavioral, physiological parameters, putting subjective info into numerical scale
utilizes scientific method
5
Q
Efficacy
A
The maximum ability of a medication to produce a result
6
Q
Effectiveness
A
How well the medication works in real world scenarios
7
Q
Efficiency
A
The ratio of the cost of medication to the health benefits it provides
8
Q
Basic research
A
Bench research
Acquisition of new knowledge
9
Q
Applied research
A
clinical research
Advance development of new dx tests, drugs, therapies, prevention stategies
10
Q
Descriptive research
A
Involves collection of data through interview and observation
11
Q
Exploratory research
A
Observational designs used to examine a phenomenon of interest and examine how it relates to other factors
12
Q
Explanatory research
A
Utilizes various types of experimental design to compare two or more conditions or interventions
13
Q
Synthesis of research
A
Meta analysis
Systematic review
Scoping review
14
Q
Where can we find evidence?
A
Databases
- library, pubmed
15
Q
What is publication bias?
A
Bias towards referencing only a specific journal
Journals are less likely to publish studies with negative or findings of no significant difference
Bias toward only publishing well known researches
16
Q
What does peer reviewed mean?
A
Content was looked at by other experts double checking methodology and results
17
Q
Grey literature
A
Anything not produced by a commercial publisher
ex. gov docs, reports, fact sheets, practice guidelines, conference proceedings, dissertations
Results of inquires that never made it to formal publication
18
Q
PICO
A
Population, problem, person
Intervention
Comparison
Outcome
19
Q
Boolean operators
A
AND, OR, NOT
use in databases to search
20
Q
Levels of evidence
A
1: Systematic reviews
2: RCTs, observational studies with strong designs
3: Study designs with poor control of bias, retrospective cohorts
4: Descriptive studies such as a case series
5: Mechanistic reasoning
21
Q
What is an independent variable?
A
The intervention being tested
22
Q
Methodological research
A
Investigation of reliability and validity
23
Q
True experiment
A
Participants are randomly assigned to at least two comparison groups
Provides the strongest evidence for causal relationships
24
Q
Between subjects design
A
aka completely randomized design
Subjects are assigned to independent groups using a randomization procedure
25
Within-subjects design
aka repeated meaures design
Subejcts act as their own control
26
Quasi-experimental design
Does not include random assignment or control group
27
Single factor design
Has one independent variable with any number of levels
28
Multi facotr design
Contains two or more independent variables
29
Factorial design
Two or more independent variables
Subjects randomly assigned to various combinations of levels of variables
Larger sample sizes are needed
30
Analysis of Factorial designs
Main effects of each independent variable
Interaction effects
Two way or three way analyses used
31
Repeated measures design strength
Differences observed are more likely to reflect tx effects vs variability b/n subjects
32
Repeated measures design potentail threats
Potential for practice effects/learning effects
Carryover effects (fatigue)
Order effects
33
Cross over designs
Participants are randomized to a treatment sequence
Used to control for order effects
Should be used when condition i stable
Considerations for washout period
34
What type of data is the number of home runs in baseball games
Discrete
35
What are the levels of measurement
Nominal, Ordinal, Interval, Ratio
36
Numerals
Used as labels, with no quantitative maeaning
37
Number
Represents a known quantity of a variable
38
Precision
Exactness of measure
39
Rules of measurement
The process of assigning numerals to variables to represent quantities of characteristics according to certain rules
40
Interval scale
Rank order and known and equal interavals between values
NO true zero
Examples; temp, calendar years
41
Ratio scale
Highest level of measurement
Absolute zero point - total absence of what is being measured
Neg values not present
Ex. height, weight, force
42
Parametric tests
Data: interval or ratio
Apply arithmetic manipulations
43
Nonparametric tests
Data: ordinal and nominal
44
Reliability
Are (instrumentation) measurements consistent
45
Validity
Is the measurement appropriate
46
Internal validity
Threats to the study design
ex. time, maturation, attrition, sampling
47
External validity
Threats to the generalization of study conclusions
Ex. does the sample represent population, study influence on test groups (hawthorne effect, rosethal effect)
48
Hawthorne effect
When subjects of an experimental study change or improve their behavior because it is being evaluated or studied
49
What type of measurement is blood type?
Nominal
50
Box and whisker plot
Visually demonstrates the spread of scores in a distribution
Range: whiskers
Median: horizontal line, 50th percentile
Interquartile range: the box
51
Standard deviation
Square root of the variance
Variance is the sum of squared differences of each score and the mean/number of scores minus 1
52
Inferential statistics
Allows researches to estimate unknown population characteristics from sample data
Probability, sampling error
53
Sampling error
Estimation of population characteristics from a sample data is based on the assumption that samples are random and valid represnetations of the data
54
Standard error of the mean
An estimate of the population standard deviation
55
Null hypothesis
Group means are not different
56
Alternative/Research hypothesis
Observed differences between two population means is not due to chance, a difference exists
57
Level of significance
Set by the researcher
Acceptable risk of making a type I error
Alpha of 0.05 means there is a 5% chance you are wrong, but are accepting the risk
58
p value
The result of your statistical test is a p value
Probability that the finding is occuring by chance
59
Statistical power
Power is the probability of attaining statistical significance
60
Determinants of statistical power
sample size, alpha level of significance, variance in the data, effect size
SAVE
61
Effect size
The is the magnitude of the observed difference
62
Correlation
Pairs of observations are examined if they go together
between -1 and 1, closer to 1 is stronger. 0.00 = no correlation
63
Regression
predict the relationship between variables and can use it to identify which variables can predict the outcome variable
64
Parametric correlation test
Pearson r
65
Parametic regression test
Simple
Multiple
66
Nonparametric test
Spearman r
67
t test
statistical ratio used to compare two group means
68
Independent or unpaired t test
used to compare means from two independent groups
Assumption of homogeneity of variance
69
Paired t test
Used to compare means from the same group after an intervention
70
ANOVA
Used to compare 3 or more goups
Independent groups or repeated measures
Uses F statistic
Assumptions: interval/ratio data, normal distribution, homogeneity of variance
71
One way ANOVA
Used to compare 3 or more independent group means
The design involves one independent variable with 3 or more levels
72
F statistic
When the treatment effect is significant, the between groups variance is lage, yielding F ratio greater than 1.
73
Two way analysis of variance
Indicates two independent variables
74
Repeated measures ANOVA
All subjects are tested under k treatment conditions
Ex. one independent variable, where all levels of treatment are administed to all subjects
75
Mixed design
involves one independent factor and one repeated facor
76
Effect size: Cohen's d
This is the magnitude of the observed difference
.2=small
.6=med
.9=large
77
Calculating a reasonable sample size
Pilot study results
Similar work published by others
Minimum differene that would be considered important by educators/experts
78
Minimal clinically important difference
Is the smallest difference that clinicians and patients would perceive as meaningful
79
Minimal detectable change
Is the smallest change in a score that is due to real change, rather than measurement error
80
Reliability is the extent to which a measurement is consistent and free from erro
True
81
Relative reliability
Refects true variance as a proportion of total variance in a set of scores
Measured as a unitless coefficient
Intraclass correlation coeffients and kappa coefficients are commonly used
82
Absolute reliability
Indicates how much a measured value, expressed in the originial units, is likely due to error
Standard error of measurement is commonly used
83
Intrarater reliability
Stability of data recorded by one individual across two or more trials
84
Interrater reliability
Variation between two or more raters who measure the same group
85
Minimal decteable change
The amount of change ina variable that must be achieved to reflect a true difference
The smallest amount of change an instrument can accurately measure
86
Measurement validity
The extent to whhich an instrument measures what it is intended to measure
87
Types of evidence to support validity
Content validity
Criterion related validity
Construct validity
88
Face validity
Appears to test what is is supposed to
89
Content validity
Adequately sample the universe of content that defines the construct being measured
Most useful with questionnaires, scales
90
Criterion-related validity
Based on the ability of the test to align with results obtained on an external criterion
The target test is compared to reference or gold standard measure
91
Criterion-related validity
Concurrent
When target and criterion test scores are obtained at about the same time
Useful when new or untested tool is potentially more efficient,easier, or safer
92
Criterion-related validity
Predictive
Determine whether a measure will be a valid predictor of some future criterion score or behavior
useful for validating screening procedures used to identify risk factors
93
Construct validity
Refects the ability of an instrument to measure the theoretical dimention of a construct
Known group method: when a test can discriminate b/n individuals
Convergence: extent to which a test correlates with other tests of closely related constructs
Discrimination; Extent to which a test has different results or low correlation of contrasing constructs
94
Criterion-referenced test
Interpreted according to a fixed standard that represents an acceptable level of performance
95
Norm referencing
Standarized assessment designed to compare and rank individuals within a defind population
96
Floor effect
Not being able to show small differences when patients are near the bottom of the scale
97
Ceiling effect
A patient who is functioning at a high level may show no improvement if a scale is not able to distinguish small changes at the top
98
Naturalistic inquiry
Require observation and interaction with research participants in their natural environments
99
Ethnography
Fieldwork where investigators immerse themselves in settings
Study participants are called informants
Strong ties of trust b/n researcher and particpant
100
Grounded theory
Formalized process for simultaneous data collection and theory development
101
Phenomenology
Built on the premise that lived expereinces and meanings are fully knowable only to those who experience them
102
Case studies
Investigates and describes phenomenon in its real life context
103
A longitudinal design can be retroscpective or prospective
True
104
Factors for causality
Temporal sequence
Strength of association
Biological credibility
Consistency
Dose response
105
Causation: Necessary
Exposure must be in place for outcome to occur
106
Causation: Sufficient
Exposure can cause an outcome, but other outcomes are involved
107
Cross Sectional studies
Possibility of reverse causation, where designated outcome may actually cause the exposure
108
A single subject design is the same a case report
False
109
Single subject design
Serial observatoin of individuals before, during, and after intervention under controlled conditions
110
N of 1 trials
Crossover with random assigned order
Blinding if possible
Answer for the specific patient
111
Split Middle Line
Celebration line:
- Drawing a straight line for rate of change
- describes trends as either accelerating or decelerating
Line divides the data within a phase equally above and below the line
112
Binomial Test
Data points counted above and below the split middle line
-dont count if on the line
113
If a test with high sensitivity is negative, we can rule out a diagnosis
True
114
Sensitivity
Test's ability to obtain a positive test when the target condition is really present
True positive rate
115
Specificity
Test's ability to obtain a negative test when the condition is really absent
True negative rate
116
False negative rate
Probability of obtaining an incorrect negative test in patients who have the target disorder
117
False positive rate
False alarm rate
Probability of an incorrect positive test in those who do not have the target condition
118
Positive predictive value
Estimates the likelihood that a person who tests positive actually has the disease
119
Negative predictive value
Probability that a person who tests negative is actually disease free
120
Receiver operating characteristic curves
Graphic representation of the balance b/n specificity and sensitivity
Can discriminate b/n presence or absence of disease
- area under the curve
-cutoff score
121
What is the major difference between a cohort study and a case control study?
A case control study is conducted after outcome of interest has occurred
122
Active variable
Manipulated by researcher and assigned
123
Attribute variable
Individula characteristic (age/gender)
Cannot be manipulated
124
Allocation concealment
Ensuring that group assignment is doen without the knowledge of those involved in the experimental process
A breakdown in concealment can create a substantial bias in estimates of treatment effects
125
The RCT ideal
Broad specification of exclusion criteria
Specification of standardized protocol
126
Pragmatic clinical trials
Diverse population with minimal exclusion criteria
Participants recruited directly from practice setting
Controls active
Treatment protocol as it would be in clinical setting
Data collection focuses on important clinical outcome
127
Confidence intervals
Helps us understand how a value from a study sample relates to the broader population
A range that represents where the true value lies for a population
128
Likert scale
Ordinal scale
Individuals choose among answer options that range from one extreme to the other to refelct attitudes, beliefs, perceptions, or values
129
What are 3 requisite design characteristics of a true experiment
Control group
Random assignment
Testing independent variable
130
Describe a systematic review
Comparing multiple articles on a specific topic and synthesize a conclusion
131
Examples of evidence syntheses
Systematic review
Meta analysis
Scoping review
Clinical practice guidelines
132
Clinical practice guidelines
Statements that include recommendations to optimize patient care, informed by a systematic review of the evidence and an assortment of the benefits and harms of alternative care options.
133
Validity stakeholder involvement
Included individulas from all relevant professional groups
Sought patients views and preferences
Target users of ther guideline are clearly defined
134
Validity rigor of development
Systematic methods were used to search
Criteria for selecting evidence clearly described
Strengths and limitations of evidence are clearly described
Methods for formulating recommendations clearly described
Health benefits, side effects, risks considered
Explicit link b/n rec and supporting evidence
Extrernally reviewed by experts
Update procedure is provided
Critical Inquiry
flashcards
Decks in class (2)
# Cards
