Chapter 14: Biostatistics Flashcards
Continuous Data
Can be ratio data or interval data
Continuous data has a logical order with values that continuously increase or decrease by the same amount
Ratio Data
Equal distance between values with a true, meaningful 0
(0=none)
E.G. age, height, weight, time, blood pressure
Interval Data
Equal difference between values but without a meaningful 0
(0=/ none)
E.G. Celsius and Fahrenheit temperature scales
Discrete or Categorical Data
Can be nominal data or ordinal data
Discrete data fits into a limited number of categories
Nominal Data
Categories are in an arbitrary order. Order of categories does not matter
E.G. gender, ethnicity, marital status
Ordinal Data
Categories are ranked in a logical order, but the difference between categories is not equal. Order of the categories matters.
E.G. NYHA functional class I-IV, 0-10 pain scale
Independent Variables
Are variables changed by the researcher
Include: drugs, drug doses, placebos, patients included (gender, age, comorbid conditions)
Dependent Variables
Can be affected by the independent variables
Include: HF progression, A1C, blood pressure, cholesterol values, mortality
Null Hypothesis
States that there is no statistical difference between groups
Alternative Hypothesis
States that there is a statistical difference between the groups. This is what the researcher is trying prove or accept.
Alpha
Is the maximum permissible error margin.
Alpha is the threshold for rejecting the null hypothesis and is commonly set at 5% or 0.05
P-values
P-value is compared to alpha. If alpha is set to 0.05 and the p-value is less than 0.05 then the null hypothesis is rejected and the result is termed statistically significant
Confidence Interval (CI)
CI provides the same information about significance as the p-value, plus the precision of the result. Alpha and CI will correlate with each other.
For example if alpha is 0.05 the study reports a 95% CI; and alpha of 0.01 corresponds to a CI of 99%
Confidence Interval (CI) Significance
The result is statistically significant if the CI range does not include 0 when comparing difference data (means)
The result is statistically significant if the CI range does not include 1 when comparing ratio data (relative risk, odds ratio, hazard ratio)
Type I Errors
Null hypothesis is rejected in error
“False Positive”
Type II Errors
Null hypothesis is accepted when it should have been rejected.
“False Negative”
Study Power
Power is the probability that a test will reject the null hypothesis correctly (i.e. the power to avoid a type II error)
Risk Formula
Risk refers to the probability of an event, when an intervention, such as a drug, is given
Risk= # of subjects is a group with an unfavorable event/ total # of subjects in group
Relative Risk
Is the ratio of risk in the exposed group (tx group) divided by risk in the control group
RR= risk in tx group/risk in control group
RR Interpretation
RR = 1 implies no difference in risk of the outcome between groups
RR > 1 implies greater risk of the outcome in tx group
RR < 1 implies lower risk (reduced risk) of the outcome in tx group
Relative Risk Reduction
Indicates how much risk is reduced in the tx group compared to the control group
RRR = 1 - RR
or
RRR = % risk in control - % risk in tx / % risk on control
RR vs RRR
RR = AS likely (vs. the control) RRR = Less likely (vs. control)
Absolute Risk Reduction
Is more useful because it includes the reduction in risk and the incidence rate of the outcome
ARR = % risk on control - % risk in tx
ARR interpretation
For example is ARR is 12%; 12 out of every 100 patients will benefit from tx. Said another way, for every 100 patients treated, 12 patient will benefit.
Number Needed to Treat (NNT)
NNT is the number of patients who need to be treated for a certain period of time in order for one patient to benefit.
NNT = 1/ARR
NNT Interpretation
If NNT is 9 then for every 9 patients treated for 1 year, 1 patient will benefit
Number Needed to Harm (NNH)
NNH is the number of patients who need to be treated for a certain period of time in order for one patient to experience harm.
NNH = 1/ARR
NNH Interpretation
If NNH is 90 then one additional patient will be harmed for every 90 patients taking tx
Hazard Ratio (HR)
A hazard rate is the rate at which an unfavorable even occurs within a short period of time
HR = hazard rate in tx/hazard rate in control
Odds Ratio (OR)
Odds ratio is used to calculate the odds of an outcome occurring with an exposure compared to the odds of the outcome occurring without exposure
OR and HR interpretation
OR or HR = 1 the event rate is the same
OR or HR > 1 the event rate in the tx group is higher than the even rate in the control group
OR or HR < 1 the event rate in the tx group is lower than the event rate in the control group
Type of Statistical Tests for Continuous Data: T-Test
T-Test is used when the endpoint has continuous data and the data is normally distributed.
Type of Statistical Tests for Continuous Data: One-sample t-test
-When data from a single sample group is compared to data from general population a one-sample t-test is used
Type of Statistical Tests for Continuous Data: Paired T-test
When a single sample group is used for pre/post measurements a paired t-test is used
Type of Statistical Tests for Continuous Data: Analysis of Variance (ANOVA)
ANOVA is used to test for statistical significance for continuous data with 3 or more sample groups
Type of Statistical Tests for Categorical or Discrete Data: Chi-Square test
Used to determine statistical significance between treatment groups
One group uses chi-square
Two groups use chi-square or fisher’s exact
Sensitivity
The true positive
Specificity
The true negative
Intention to Treat Protocol
Analysis includes data for all patients originally allocated to each treatment group even if the patient did not complete the trial according to trial protocol
Types of Studies from Most Reliable to Least Reliable
- Systematic Review and Meta-Analysis
- Randomized Controlled Trials
- Cohort Studies
- Case-controlled Studies
- Case Series and Case Reports
- Expert Opinion
Case-Control Study
Compares patients with a disease (cases) to those without the disease (control).
Outcomes are already known and researcher looks back retrospectively
Cohort Study
Compares outcomes of a group of patients exposed and not exposed to a treatment.
The researcher follows both groups prospectively
Case-Report and Case-Series
Describes an adverse reaction or a unique condition that appears in a single patient (case report) or a few patients (case series)
Randomized Controlled Trial
Patients are randomized (have an equal chance) of being assigned to treatment or control group.
Studies can be blinded or double-blinded
Meta-Analysis
Combines results from multiple studies in order to develop a conclusion that has greater statistical power than is possible from the individual studies
Systematic Review
Summary of the clinical literature that focuses on a specific topic or question
Direct Medical Costs
Drug preparation and administration, office visits, hospital bed etc
Direct Non-Medical Costs
Travel and lodging, elder or childcare costs, home health aides
Indirect Costs
Lost work time, low work productivity, morbidity and mortality
Intangible Costs
Pain, suffering, anxiety and fatigue
Incremental Cost-Effectiveness Ratios
= (C2-C1)/(E2-E1)
Cost Minimization Analysis
Is used when 2 or more interventions have demonstrated equivalence in outcomes and the costs of each intervention are being compared.
Limited to compare only alternative with demonstrated equivalent outcomes
Cost Benefit Analysis
Is a systematic process for calculating and comparing benefits and costs of an intervention in terms on monetary units ($)
Cost Effectiveness Analysis
Is used to compare the clinical effects of two or more interventions to the respective costs.
Main advantage is that the outcomes are easier to quantify .
Disadvantage is the inability to directly compare different types of outcomes
Cost Utility Analysis
includes a quality of life component. Uses quality-adjusted life years (QALYs) and disability-adjusted life years (DALYs)
