Biostats

ratio data

equal difference between values with a true meaningful difference e.g. 0=none (age, height, time, blood pressure)

interval data

equal difference between values, but without a meaningful zero eg. 0=/=none (Celsius, Fahrenheit)

nominal data

categories are in an arbitrary order (gender, ethnicity, marital status, mortality)

ordinal data

categories are ranked in a logical order, but the difference between categories is not equal (0-10 pain scale, NYHA functional class I-IV)

mean

average

median

middle value

mode

most frequent

Gaussian distribution can be described as _____ shaped

bell shaped (normal distribution)

In normal distribution ___% if the values fall within 1 SD of the mean and ___% fall within 2 SD

68%

95%

If data is right (positive) skewed, there are more ___ values

low

If data is left (negative) skewed, there are more ___ values

high

if p-value is less than alpha…

reject the null hypothesis (e.g. p-value<0.05)

Confidence interval =

1 - alpha

A result is statistically significant if the CI range does not include ____

0

ratio data is statistically significant if the result does not include ___

1

Type 1 error also known as false ____ occurs when…

positive

null hypothesis is rejected despite being true (e.g. there is no real difference between the groups)

Type 2 error also known as false ____ occurs when…

negative

null hypothesis is accepted despite being false (e.g. there is a difference between groups)

power of a study =

1 - beta

a narrow CI implies ____ while a wide CI implies ____

high precision

low precision

power is defined as

the probability that a test will reject the null hypothesis correctly (i.e. avoid type II error)

risk=

number of subjects with an unfavorable event / total number of subjects in the group

RR=

risk in treatment / risk in control

RR of 1 (100%) implies…

no difference in risk between the groups

RR > 1 (100%) implies…

greater risk in treatment group

RR < 1 (100%) implies…

lower risk in treatment group

RRR= ____ OR ____

(% risk in control - % risk in treatment) / % risk in control

OR

1 - RR

RRR indicates

how much risk is reduced in the treatment group compared to the control group

why is absolute risk reduction used?

It is used because it includes the reduction of risk and the incidence of risk in the outcome. For example, if there is a 50% decrease in risk of nausea between two drugs, but the risk of nausea was very low to begin with, the large risk reduction has little practical benefit.

ARR =

% risk in control group - % risk in treatment group

NNT=

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

OR

1/ARR

for NNT round ___ and for NNH round ___

NNT: round up

NNH: round down

Odds ratio is used to…

estimate the risk of unfavorable events associated with a treatment

odds ratio =

AD/BC

A: has treatment; outcome present

B: has treatment; outcome absent

C: no treatment; outcome present

D: no treatment; outcome absent

hazard ratio is the rate at which…

unfavorable event occurs within a short period of time

HR=

hazard rate in treatment / hazard rate in control

OR or HR = 1

event rate is the same

OR or HR > 1

event rate in treatment group is higher than in control group

OR or HR < 1

event rate in treatment group is less than the event rate in control group

a composite endpoint…

combines multiple individual endpoints into one measurement (e.g. death from MI, stroke, and cardiovascular disease)

continuous data uses these statistical tests

T-test and ANOVA

t test is used when

a single sample group is used for both pre and post measurement (e.g. the subject serves as their own control)

ANOVA is used when

continuous data with 3 or more samples

chi-squared test is used for…

nominal or ordinal data to determine statistical significance between two groups

regression is used to describe…

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

regression is used when researchers need to assess…

multiple independent variables or need to control for many confounding factors

sensitivity desscribes

how effectively a test identifies patients with a condition

specificity describes

how effectively a test identifies a patient without a condition

sensitivity formula

[A / (A+C)] * 100%

A: positive test and have the condition

C: positive test and don’t have the condition

specificity formula

[D / (D+B)] * 100%

B: positive test and don’t have the condition

D: negative test and don’t have the condition

intention to treat

analysis includes data for all patients originally allocated to each treatment group, even if the patient did not complete the trial according to study protocol

per protocol analysis

analysis includes only those who completed the study according to protocol

In a forest plot, a result is not statistically significant if…

it crosses the center line

Study types from most to least reliable

systemic review / meta analysis

randomized control trials

cohort studies

case-control studies

case series and case reports

expert opinion

case-control study

retrospectively compares patients with a disease to those without a disease

cohort study

prospectively compares outcomes of a group of patients exposed and not exposed to a treatment

pharmacoeconomics definition

identifies, measures, and compares the cost (direct, indirect, and intangible) and the consequences (clinical, economic, and humanistic) of pharmaceutical products and services

direct medical costs

drug preparations and administration, including home infusion supplies

inpatient direct costs: bed, administration, staff, surgeries, procedures, labs

outpatient direct costs: office and clinic visits

direct non-medical costs

travel and lodging costs

household costs like childcare or eldercare

home health aides

indirect costs

lost work time

low work productivity

morbidity: costs from having disease and related disabilities

mortality

intangible costs

pain, suffering, anxiety, fatigue

incremental cost ratio =

(C2 - C1) / (E2 -E1)

cost-minimization analysis is used when…

and its outcome units are…

Two or more interventions have demonstrated equivalence in outcomes and the cost of each intervention are being compared.

Outcome units are assumed to be equivalent

cost-benefit analysis is used when…

and its outcome units are…

Calculating and comparing benefits and costs of an intervention in terms of monetary units.

Outcome units are dollars

cost-effectiveness analysis is used when…

and its outcome units are…

Comparing the clinical effect of two or more intervention to the respective costs.

Outcome units are natural units like life-years gained, BP, % at treatment goal, etc.

cost-utility analysis is used when…

and its outcome units are…

Analyzing a quality-of-life component of morbidity assessments using common health indices such as quality-adjusted life years (QALYs) and disability-adjusted life years (DALYs)

Outcome units are QALY or other utilities