Ch. 12 Biostats & Eco Flashcards
ECHO stands for
Economic outcomes
Clinical outcomes
Humanistic Outcomes
Cost Minimization Analysis (CMA)
Used when two or more interventions have already demonstrated equivalency in outcomes and the costs of each intervention are being compared
Cost Benefit Analysis
Systematic process for calculating and comparing benefits and costs of an intervention in terms of monetary units
Cost Effectiveness Analysis (CEA)
Compares clinical effects to net costs. Can only compare one type of outcomes. Uses natural units for the outcome (BP, life-years gained)
Cost Utility Analysis (CUA)
Includes quality of life in terms of years gained in quality-adjusted life years (QALYs). Does not measure the “quality” or “utility” of those years
Standard deviation
Indicates how spread out the data is, and to what degree the data is dispersed away from the MEAN
Negative skew
Has more high values
Positive skew
Has more low values
In skewed data, it is best to use this central tendency value
The median
Type I Error
False positive (most common error), CI= 1 - alpha Rejects the null hypothesis in error
Type II Error
False negative (beta) Accepts the null hypothesis in error
Study Power
Probability that a test will reject the null hypothesis correctly (the power to avoid Type II error correctly)
Power= 1 - beta
Relative Risk (RR)
Risk in tx group/Risk in control group
Interprets how likely the tx group risk will occur compared to the control group
Relative Risk Reduction (RRR)
1-RR
Interprets how much less likely the tx group risk will occur compared to the control group
Absolute Risk Reduction (ARR)
Risk in control group - Risk in tx group
Number Needed to Treat (NNT)
1/(Risk in control group-Risk in tx group) OR 1/ARR
Round UP to avoid overestimating
Number Needed to Harm (NNH)
1/ARR
Round DOWN to avoid underestimating
Odds Ratio
The probability that an event will occur, versus the probability that it will not occur (used in case-control studies)
OR= AD / BC
Hazard Ratio
Survival analysis
HR= Hazard rate in the tx group / Hazard rate in the control group
Types of discrete data
Nominal (name) i.e. male, female
Ordinal (order) i.e. pain scale, NYHA class
Types of continuous data
Ratio, has a meaningful 0 value (i.e. blood pressure)
Interval, does not have a meaningful 0 (i.e. degrees)
Use of Chi Square Test
In discrete data; tests for difference between/among tx groups
Use of ANOVA/F-test
In continuous data with 3 or more sample groups
Use of Student T-test
In continuous data with one sample group
Correlation
Determines if one variable is related to another (can be positive with increase to right, or negative with decrease to the right)
Types of correlation
Spearman’s Rank Order: (Rho) tests correlation in ordinal, ranked data
Pearson’s Correlation Coefficient: (r) used in continuous data (indicates the strength and direction of the relation between variable)
Linear Regression use
Continuous data
Logistic Regression use
Categorical data
Cox Regression
Categorical data in survival analysis
Sensitivity
True Positive
A / (A+C) x 100
Specificity
True Negative
D / (B+D) x 100
Medical studies from most to least reliable
Systematic reviews and Meta-analysis > RCTS > Cohort > Case-control > Case Series and Reports > Expert opinion
