RxPrep: Biostatistics

What are the two types of continuous data?

interval data

ratio data

What is the definition of continuous data?

logical order with values that continuously increase or decrease by the same amount

What type of continuous data has no meaningful zero?

interval data

What are some examples of interval data?

temperature

What type of continuous data has a meaningful zero?

ratio data

What are some examples of ratio data?

heart rate

blood pressure

What are the two types of discrete data?

Nominal

Ordinal

What is another name for discrete data?

Categorical data

What is the definition of nominal data?

subjects are sorted into arbitrary categories

What is an example of nominal data?

male or female

What is the definition of ordinal data?

ranked with a logical order, categories do not increase by the same amount

What is an example of ordinal data?

pain scale; pain does not double for 2 to 4

What are measures of central tendency?

typical descriptive values including the mean, median, and mode

What is the mean?

An average, all of the values are added up and divided by the number of vavlues.

What type of data is the mean preferred for?

continuous data that is normally distributed

What is the median?

value in the middle when values are arranged lowest to highest

What type of data is the median preferred for?

ordinal data

skewed continuous data

What is the mode?

value that occurs most frequently

What type of data is the mode preferred for?

nominal data

What methods are used to describe the variability or spread of data?

Range

Standard deviation (SD)

What is the range?

difference between highest and lowest values

What does the standard deviation indicate?

how spread out the data is and to what degree it is away from the mean

What percentage of values fall within 1 standard deviation of the mean in normally distributed data?

68%

What percentage of values fall within 2 standard deviations of the mean in normally distributed data?

95%

What factors can cause skewed data?

small sample size

outliers in data

High values cause what kind of skew?

right

positive skew

Low values cause what kind of skew

left

negative skew

What type of variable is changed by the researcher?

independent variable

What does a trial need to demonstrate to show significance?

That the null hypothesis is not true and should be rejected

What is a null hypothesis?

No statistically significant difference between grouops

What is an alternative hypothesis?

there is a statistically significant difference between groups

What is the alpha level?

maximum permissible error margin.

Threshold for rejecting null hypothesis

To reject a null hypothesis what must be the p-value be in regard to the alpha?

p-value less than alpha

How is the confidence interval calculated?

CI = 1 - alpha

What is a type 1 error?

false positive. When the null hypothesis was rejected in error.

What is a type 2 error?

when a null hypothesis is accepted when it should have been rejected.

What is study power?

probability that a test will reject the null hypothesis. Used to avoid type 2 error.

What is power determined by?

- outcome values

- difference in outcome rates between groups

- alpha level

What is the formula for relative risk?

(# subjects in group with unfavorable event) / (total # of subjects in group)

What is the risk ratio formula?

(risk in treatment group) / (risk in control group)

What does an RR of 1 indicate?

no difference in risk of outcome between groups

What does an RR over 1 indicate?

greater risk of outcome in treatment group

What does an RR under 1 indicate?

lower risk of outcome in treatment group

What is the definition of relative risk reduction?

indicates how much the risk is reduced in treatment compared to placebo

What is the formula of relative risk reduction?

(% risk control - % risk treatment)

/

(% risk control OR

1 minus RR (decimal))

What is the formula for absolute risk reduction?

(% risk control)

-

(% risk treatment)

Why is absolute risk reduction useful?

it includes reduction in risk and the incidence rate of the outcome

What is the formula for number needed to treat?

NNT

1 / (risk in control - risk in treatment)

1 / ARR

**Risk and ARR expressed as decimals

How does rounding work for the number needed to treat?

round UP to the nearest whole number

What is the formula for number needed to harm?

NNH

1/(risk in control - risk in treatment)

ARR

**Risk and ARR expressed as decimals

How does rounding work for the number needed to harm?

round DOWN to the nearest whole number

When is odds ratio calculated?

to estimate the risks associated with a treatment and calculate the odds of an outcome occurring with an exposure compared to no exposure

What is the formula for odds ratio?

OR = AD/BC

A = # outcome w/ exposure

B = # w/o outcome w/ exposure

C = # w/ outcome w/o exposure

D = # w/o outcome w/o exposure

What is the hazard ratio?

the rate an unfavorable event occurs within a short period of time

What is the formula for hazard ratio?

HR = HR in treatment / HR in control

What does it mean if OR or HR is 1?

event rate is the same in treatment and contrl

What does it mean if OR or HR is above 1?

event rate in treatment group is higher than control

What does it mean if OR or HR is less than 1?

event rate in treatment group is lower than control

What is a primary endpoint?

main result that is measured to see if treatment has a significant benefit

What is a composite endpoint?

Combine multiple individual endpoints into one measurement

When should parametric methods be used?

For normally distributed, continuous data

When should nonparametric methods be used?

For data that is not normally distributed and continuous

What is a T-test?

parametric method used when endpoint has continuous data

When is a one-sample T-test performed?

For a single sample group that is compared with known data from the general population

When is a paired T-test used?

for a single sample group for a pre/post measurement

When is a student T-test used?

when the study has two independent samples. The treatment and control group.

When is ANOVA used?

With continuous data with 3 or more samples or groups

What type of data is a Chi-square test used for?

nominal data

ordinal data

What is correlation?

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

What is regression?

describes the relationship between a dependent variable and one or more independent variables

When is regression common?

When multiple independent variables are assessed or when there is a need to control many confounding factors

What is sensitivity?

how effectively a test identifies patients with a condition

What is the formula for sensitivity?

A / (A + C)

A = # with condition and positive test

C = # with condition and negative test

What is specificity?

how effectively a test identifies a patient without the condition

What is the formula for specificity?

D / (B+D)

D = # without condition, negative test

B = $ without condition, positive test

What data is included in intention to treat?

Data for all patients allocated to each treatment group regardless if they completed the trial according to protocol

What data is included in per protocol analysis?

subset of the trial population who completed the study according to protocol

Which type of analysis gives a real world estimate of treatment effect?

intention to treat

Which type of analysis gives an optimistic estimate of the treatment effect?

per protocol

What is the purpose of an equivalence trial?

To show that the new treatment roughly has the same effect as the old treatment

What is the purpose of non-inferiority trials?

to show that the new treatment is no worse than the current standard

What do the boxes, diamonds, and horizontal/vertical lines show in a forest plot?

Box = show the effect estimate

Diamond = pooled results from multiple studies

Horizontal = length of the CI for the endpoint

Vertical = line of no effect

What is a case-control study?

retrospective. Compare cases to controls

What is a cohort study?

retrospective or prospective. Compare patients with exposurer to those without.

What is a limitation of cohort studies?

influence of confounders affecting the outcome

What is a randomized controlled trial?

prospective. Compare patients who were randomly assigned to groups

What is a meta-analyses?

analyze results of multiple studies

What is an advantage of meta-analyses?

allows for greater power

What is the ECHO model?

economic

clinical

humanistic

outcomes

Assesses outcomes associated with diseases and treatments

What is the definition of a cost-minimization analysis?

used when two or more interventions have demonstrated equally in outcomes and costs of each intervention are being compared.

What is the definition of cost-benefit analysis?

systematic process for calculating and comparing benefits and costs of an intervention in terms of monetary units

What is the definition of cost-effectiveness analysis?

used to compare the clinical effects of two or more interventions to the respective costs.

What is the definition of cost-utility analysis?

includes quality of life component of morbidity assessments using common health indices such as quality adjusted life years and disability adjusted life years.