Biostatistics Flashcards by Grace Whitham

What is continuous data?

Data with a logical order
Values that continuously increase or decrease – have theoretically infinite options
Examples: age, height, weight, time, BP, temperature

How well did you know this?

Not at all

Perfectly

What are types of discrete (categorical) data?

Nominal and ordinal

How well did you know this?

Not at all

Perfectly

What is nominal data?

Categories – order does not matter

Examples: gender, ethnicity, marital status, mortality

How well did you know this?

Not at all

Perfectly

What is ordinal data?

Categories are ranked in logical order but difference between categories is not equal
Examples: NYHA functional class I-IV; 0-10 pain scale

How well did you know this?

Not at all

Perfectly

What is the spread of data?

How much it varies

How well did you know this?

Not at all

Perfectly

What is the range of data?

Difference between highest and lowest value

How well did you know this?

Not at all

Perfectly

What is the standard deviation (SD) of data?

Indicates how spread out the data is, and to what degree the data is dispersed away from the mean

How well did you know this?

Not at all

Perfectly

What is Gaussian distribution?

Normally shaped “bell curve”

How well did you know this?

Not at all

Perfectly

On a bell curve, ___% of values fall into one standard deviation of the mean, ___% fall into 2 standard deviations of the mean

68%

95%

How well did you know this?

Not at all

Perfectly

Independent vs dependent variables

Independent: Changed by researcher
Dependent: affected by independent variables

How well did you know this?

Not at all

Perfectly

What does the null hypothesis state?

No statistically significant difference between groups

***Researcher tries to reject this

How well did you know this?

Not at all

Perfectly

What does the alternative hypothesis state?

There is a statistically significant difference between groups
***Researcher tries to prove this

How well did you know this?

Not at all

Perfectly

What is the alpha value?

Threshold for rejecting null hypothesis
“error margin”
Usually 0.05 or 5%

How well did you know this?

Not at all

Perfectly

If p-value is ___ alpha, null is rejected and result (is/is not) statistically significant

Less than alpha

How well did you know this?

Not at all

Perfectly

What is the confidence interval? How is it calculated?

Provides same info about significance as p-value, plus precision of the result
1-alpha = CI

How well did you know this?

Not at all

Perfectly

If a confidence interval (CI) includes ____ it is not statistically significant

0: when comparing difference data
1: when comparing ratio data

How well did you know this?

Not at all

Perfectly

Which indicates high precision, narrow or wide CI?

Narrow

How well did you know this?

Not at all

Perfectly

What does a 95% confidence interval indicate?

You are 95% sure that the true value falls between the range given

How well did you know this?

Not at all

Perfectly

Type 1 error = false _____

Type 2 error = false _____

Type 1: false positive

Type 2: false negatives

How well did you know this?

Not at all

Perfectly

What does the study power indicate?

The probability that a test will reject the null hypothesis correctly
Power to avoid type II error

How well did you know this?

Not at all

Perfectly

If beta is 0.2, what is the study power?

80%

How well did you know this?

Not at all

Perfectly

What is relative risk?

Risk in the exposed group (treatment) divided by risk in the control group

Risk formula

Total number of subjects in a group

Relative risk formula

Risk in control group

What does relative risk of 1, >1, or <1 mean?

``` 1 = no difference in risk between group >1 = greater risk in treatment group <1 = less risk in treatment group ```

Relative risk reduction calculation

(% risk control group - % risk treatment group) ------------------- 5 risk control group OR 1-RR

What is relative risk reduction?

Indicates how much risk is reduced in treatment group

Absolute risk reduction calculation

(% risk in control group) – (% risk in treatment group)

What does an absolute risk reduction (ARR) of 12% mean?

12 out of every 100 patients benefit from treatment

Number needed to treat calculation

1/(risk in control group – risk in treatment group) OR 1/ARR

Which way do you round number needed to treat (NNT)? Number needed to harm (NNH)?

NNT – round up | NNH – round down

What is odds ratio (OR)?

Used to assess the odds of an outcome occurring with an exposure compared to odds of the outcome occurring without the exposure

What is hazard ratio?

Rate at which an unfavorable event occurs within a short period of time

Hazard ratio (HR) calculation

Hazard rate in treatment group ----------------------- Hazard rate in control group

What is a composite endpoint?

Combined endpoints into one measurement

What test should be used with numerical/continuous data with one group

One-sample t-test

What test should be used with numerical/continuous data with two groups (e.g treatment and control)

Independence/unpaired student t-test

What test should be used with numerical/continuous data with three or more groups

ANOVA

What test should be used with discrete/categorical data with two groups

Chi-squared test or Fisher’s exact test

What is correlation

Statistically technique that is used to determine if one variable changes or is related to another variable

What is regression used to describe?

The relationship between a dependent variable and one or more independent variables or ho much the value of the dependent variable changes when the independent variables changes

What does sensitivity/specific indicate?

Sensitivity – true positive; positive in patients with the condition Specificity – true negative; negative in patients without the condition

What is intention-to-treat analysis?

Includes data for all patients originally allocated to each treatment group even if they did not complete the trial according to study protocol

What is per-protocol analysis?

Includes data for all patients who completed the study according to protocol

What are equivalence and noninferiority trial designs?

Equivalence – demonstrates new treatment has ~equal effect the old treatment Non-inferiority – new treatment is no worse than current standard based on predefined non-inferiority margin

What do forest plots provide?

Cis for difference data or ratio data

Characteristics of case-control studies

Retrospective | Compare cases to controls

Characteristics of cohort studies | Limitations

Retrospective or prospective Compare patients with an exposure to those without an exposure Limitations: can be influenced by confounders

Characteristics of randomized controlled trials

Prospective | Compare patients who were randomly assigned to groups

Characteristics of meta-analyses

Analyzes the results of multiple studies

Characteristics of systematic review article

Summary of the clinical literature that focusses on a specific topic or question

Characteristics of case report or case series

Describes adverse event or condition in a single patient (report) or a few patients (series)

What does pharmacoeconomics do?

Identifies, measures, and compares the costs and consequences of pharmaceutical products and services

What are some indirect and intangible medical costs?

Indirect: lost work time, low work productivity, morbidity, mortality Intangible: pain, suffering, anxiety, fatigue