Basic Concepts
Sample vs Population
Variables
- Random
- Independent
- Dependent
Analysis
- Univariable
- Bivariable
- Multivariable
Types of Data
- Non-parametric
- Parametric
Distributions
- Normal
- Bimodal
- Rectangular
- Skewed
- Positive
Skew
- Negative
Skew
- Kurtosis
Descriptive Statistics
Measures of Central Tendency
- Mean
- Median
- Mode
- Gaussian
or Normally distributed data graph
- Positively
skewed data graph
- Negatively
skewed data graph
Measures of Variability
- Range
- Interquartile
Range
- Variance
- Standard
Deviation
- Coefficient
of Variation
- Standard
Error of the Mean (aka standard error)
- Confidence
Intervals
- Interpretation
of Superiority Traits
- Interpretation
for Differences
- Interpretation
for Ratios
- Interpretation
of Equivalence Trials
- Interpretation
of Non-Inferiority Trial
Sensitivity, Specificity, Predictive Value, and Hypothesis Testing
Sensitivity
Specificity
Positive Predictive Value
Negative Predictive Value
Prevalence
Incidence
Hypothesis Testing
- Null
Hypothesis
- Alternative
Hypothesis
- Type
1 or Alpha Error
- Type
2 or Beta Error
Sample size calculations
Correcting for estimated Drop-out Rate
Delta
One-tailed vs Two-tailed tests
Effect of change in Alpha
Effect of change in Beta
Effect of change in Delta
Effect of One-tailed vs Two-tailed design
Effect of change in expected incidence
Statistical Inference Techniques for Hypothesis Testing w/ Parametric Data
T-test
- Non-Paired
- Paired
- One-Tailed
- Two-Tailed
- Bonferroni
- Dunn-Sidak
ANOVA
- ANOVAs
for Non-Paired samples
- 1-way
ANOVA
- Multifactorial
ANOVAs
- ANOVAs
for Paired samples
- Repeated
Measures ANOVA
- 2-way
Repeated Measures ANOVA
- Repeated
Measures Regression
- Multiple
Comparison Methods (Post Hoc Analysis)
- Least
Significant Difference
- Layer
(aka stepwise) Methods
- Experiment-wise
Methods
- Dunnett
- Dunn
- Tukey
- Scheffe
Statistical Inference Techniques for Hypothesis Testing w/ Non-Parametric Data
Non-Parametric Tests for Nominal Data
- Chi-Square
Tests
- Test
of Independence
- Goodness
of Fit
- McNemar
Chi-Square
- Mantel-Haenszel
Chi-Square
- Log-Linear
Analysis
- Post
Hoc Analysis
- Fisher's
Exact
- Rows
by Columns Test
- Table
Noting Proper Test to Use
Non-Parametric Tests for Ordinal Data
- Mann
Whitney U and Wilcoxon Rank Sum
- Kolmogorov-Smirnov
- Wilcoxon
Signed Rank
- Kruskal-Wallis
- Friedman
- Table
Noting Proper Test to Use
Correlation and Regression
Correlation
- r-value
- population
correlation coefficient
- coefficient
of determination (r2)
- sample
correlation coefficient
Regression
- Multiple
regression
- Linear
regression
- Logistic
regression
- Cox
Proportion Hazards
Error vs Bias
Bias in the Publication Process
- Funding
Sources
- Publication
- Investigator
- Reviewer
- Small
study
Compliance
Prevalence
Selection Bias
- Inclusion
in study
- Participation
in study
- Admission
rate
- Nonresponse
- Membership
- Diagnostic
or Detection
- Insensitive
Measure
- Procedure
Bias that falsely identify or inflate risk factors
- Recall
- Observer
- Interviewer
- Misclassification
- Random
Misclassification
- Channeling
Controlling for Bias
- Design
of Trial
- Randomization
- Selection
Criteria
- Blinding
- Matching
- Using
Controls
- Analysis
Controlled vs Non-Controlled
Various Types of Controls
- Placebo
- Historical
- Crossover
- Standard
or Active Treatment
- Within
Patient Comparison
Blinding in Trial Design
Non-blinded
Single-blinded
Double-blinded
Double-dummy
Triple-blind
Total-blind
Note concerning trials that claim blinding, but may not be blinded.
Run-In-Phase
Confounding
Arrow Chart Examples
Co-linearity
Controlling for Confounding
- Design
of Trial
- Randomization
- Selection
Criteria
- Blinding
- Matching
- Analysis
- Stratification
- Multivariate
Regression Analysis
- Note
concerning various types of ANOVA used
Validity
Internal
External
Prospective vs Retrospective
General Considerations for Assessing Causality
- Relative
strength of causal relationships
Retrospective Trials
- Case-Control
- Cohorts
- Retrospective
Cohort
- Prospective
Cohort (retrospective)
- True
Prospective Cohort
- Benefits
- Problems
Odds Ratios, Relative Risks (aka Risk Ratios, Rate Ratio, or Incidence Rate Ratio)
Odds Ratio (see below for further explanation of examples)
Relative Risk (see below for further explanation of examples)
Hazard Ratio
Interpretation
Absolute Risk
Absolute Risk Reduction or Increase
Numbers Needed to Treat or Harm
Relative Risk Examples for RCTs and Cohort Trials
- Example
1: Relative Risk Reduction example w/ Calculations and Interpretations
- Absolute
Risk
- Absolute
Risk Reduction
- Numbers
Needed to Treat
- Relative
Risk
- Relative
Risk Reduction
- Difference
vs Cohort Trial
- Example
2: Relative Risk Increase example with Calculations and Interpretations
- Absolute
Risk
- Absolute
Risk Increase
- Numbers
Needed to Harm
- Relative
Risk
- Relative
Risk Increase
- Difference
vs Cohort Trial
Odds Ratio Examples for Case Control Trials
-
Example
1 w/ Calculations and Interpretation
-
Example
2 w/ Calculations and Interpretation
-
Note
concerning change in Incidence's effect on estimation of
Relative Risk
Randomized Controlled Trials
Parallel Design
Cross-over Design
Considerations
Disadvantages
Various Types of Analyses
Intention of Treat (ITT)
As Treated
Per Protocol
Post Hoc
Note concerning trials that claim to be ITT, but may not truly be ITT.
Run-In-Phase
Interim Analysis
Subgroup Analysis
Survival Analysis
-
Hazard
-
Life
table
-
Actuarial
Method
-
Kaplan-Meier
-
Statistical
tests
-
Log-Rank
-
Wilcoxon-type
-
Cox
Proportion Hazards
Meta-Analysis
Combined Endpoints
Miscellaneous
Breslow Day test
Rule of Four
Abbreviations
References
Index
|
This book is a primer for basic medical literature evaluation. It is not intended to be used as a complete critical literature evaluation text. It is simply a “survival guide” for students and clinicians trying to learn the basics of literature evaluation.