14. Biostats Flashcards
2 types of continuous data
interval data and ratio data
Interval data definition and example
No meaningful zero (zero does not equal none)
Example: temperature scale (0 celcius does not mean no temp)
Ratio data definition and example
Meaningful zero (zero equals none)
Example: HR (0 BPM is cardiac arrest)
2 types of discrete data
nomial and ordinal
Nomial data definition and example
Subjects are sorted into arbitrary categories (names) , order of categories does not matter
Example: gender, ethnicity, martial status, mortality
Ordinal data defintion and example
Ranked and has a logical order
Example: pain scale
Note: ordinal scale categories do not increase by the same amount (pain scale 4 is not necessarily 2x more than pain scale 2)
In normal distributions, ___% of values fall within 1 standard deviation of the mean
68%
In normal distributions, ___% of the values fall within 2 standard deviations of the mean
95%
In normal distributions, 99.7% of the values fall within ___ standard deviations of the mean
3 SDs
More lower values in data set and outliers are high values, data is skewed to (left/right), aka (negative/positive) skew
Right, positive
More higher values in data set and outliers are low values, data is skewed to (left/right), aka (negative/positive) skew
Left, negative
T/F: the null hypothesis and alternative hypothesis are always complementary; when one is accepted, the other is rejected
True
Define null hypothesis
There is no statistically significant difference between groups
Alpha (error margin) is the threshold for rejecting the null hypothesis. In medical research, alpha is commonly set at ____
5% or 0.05
T/F: If the p-value is less than alpha, the null hypothesis is rejected
True
T/F: If the p-value is greater than alpha, the null hypothesis is rejected
False - if p-value is greater than alpha, the study failed to reject the null hypothesis and results are not statistically significant
If alpha is 0.05, the study reports ___% confidence intervals
95%
If p-value is <0.01, it means __% probability (confidence) that the conclusion is correct; less than __% change it’s not
99% confidence, 1% chance it’s not
When comparing difference data (means, based on subtraction), the result is statistically significant if the CI range does not include ___
zero
When comparing ratio data (relative risk, odds ratio, hazard ratio, based on division), the result is statistically significant if the CI range does not include ___
one
A (narrow/wide) CI range implies high precision and a (narrow/wide) CI range implies poor precision
narrow, wide
False positive is considered to be a type __ error (null hypothesis was rejected in error)
Type 1 error
False negative is considered to be type __ error (null hypothesis is accepted when it should have been rejected)
Type 2 error
Study __ is the probability that a test will reject the null hypothesis corrected (i.e. the power to avoid a type 2 error)
Study power
Type 2 error is denoted as __
beta (ß)
Power equation
Power = 1 - ß
If beta is set at 0.2, the study has __% power which means there is a __% chance of missing a true difference and making a type ___ error
80%, 20%, type 2 error
____ is the ratio of risk in the exposed group (treatment) divided by risk in the control group
Relative risk
Relative risk formula
RR = risk in treatment group / risk in control group
RR=1 (100%) implies ____ risk of outcomes in the treatment group
no difference (compared to control)
RR > 1 (100%) implies ___ risk of outcomes in the treatment group
greater
RR < 1 (100%) implies ___ risk of outcomes in the treatment group
lower
Relative risk reduction (RRR) formula
RRR = (%risk in control - % risk in treatment) / % risk in control group
RRR = 1 - RR (decimal form)
Absolute risk reduction (ARR) formula
ARR = (% risk in control) - (% risk in treatment)
ARR of 12% interpretation
12 out of every 100 patients benefit from the treatment
For every 100 patients treated, 12 fewer patients will have the outcome
____ is the number of patients who need to treat for a certain period of time in order for 1 patient to benefit
Number needed to treat (NNT)
Number needed to treat (NNT) and Number needed to harm (NNH) formula
NNT (or NNH) = 1 / (risk in control - risk in treatment)
NNT (or NNH) = 1 / ARR
*risk and ARR expressed as decimals
If ARR is 12%, what is the NNT?
NNT = 1/0.12 = 8.3 –> rounded to 9
____ is the number of patients need to treat for a certain period of time in order for 1 patient to experience harm
Number needed to harm (NNH)
NNT is rounded (up/down) and NNH is rounded (up/down)
Up, down
NNH 90 interpretation
One additional case of outcome expected to occur for every 90 patients taking treatment vs placebo
___ is used to estimate the risk of unfavorable events a/w treatment or intervention
Odds ratio
Odds ratio (OR) formula
OR = AD/BC
A-Treatment outcome present
B-Treatment outcome absent
C-Control outcome present
D-Control outcome abset
In a survival analysis, ___ is the rate at which an unfavorable event occurs within a short period of time
Hazard ratio
Hazard ratio (HR) formula
HR = Hazard rate in treatment / hazard rate in control
OR or HR = 1 interpretation
event rate is same in both groups
OR or HR > 1 interpretation
Event rate is higher in treatment group
OR or HR < 1 interpretation
Event rate is lower in treatment group
___ t-test is used when data from a single sample group is compared with known data from the general population
One-sample t-test
___ t-test is used when a single sample group is used for a pre-/post-measurement
Paired t-test
___ t-test is used when the study has 2 independent samples: treatment and control groups
Student t-test
___ is used to test for statistical significance when using continuous data with 3 or more samples or groups
Analysis of variance (ANOVA) or F-test
For nominal or ordinal data, a ___ test is used to determine statistical significance between treatment groups
Chi-square
(sensitivity/specificity) describes how effectively a test identifies patients WITH the condition
Sensitivty
(sensitivity/specificity) describes how effectively a test identifies patients WITHOUT the condition
Specificity
Sensitivity formula
A-Positive test, have condition
B-Positive test, does not have condition
C-Negative test, have condition
D-Negative test, does not have condition
Sensitivity = A / (A+C) x 100
Specificity formula
A-Positive test, have condition
B-Positive test, does not have condition
C-Negative test, have condition
D-Negative test, does not have condition
Specificity = D / (B+D) x 100
Sensitivity of 28% interpretation
Only 28% of patients with the condition will have a positive RF result, test will be negative in 72% of patients with disease (missed dx)
Specificity of 87% interpretation
Test will be negative in 87% of patients without the disease but 13% of patients without the disease may test positive (false dx)
___ analysis includes data for all patients originally allocated to each treatment group even if they did not complete the trial
Intention to Treat
___ analysis only include data for patients who completed the trial
Per protocol
Name type of study: retrospective comparisons of cases (patients with a disease) and controls (patients without a disease)
Case-control study
Name type of study: retrospective or prospective comparisons of patients with an exposure to those without an exposure
Cohort studies
Name type of study: prospective comparison of patients who were randomly assigned to groups
Randomized controlled trials
Name type of study: analyzes the results of multiple studies
Meta-analysis
What type of study is most reliable?
Systematic reviews and meta-analysis
List type of studies in order of most reliable to least reliable:
Expert opinion, cohort studies, case series and reports, case-controlled studies, randomized controlled trials, systematic reviews and meta-analysis
Systematic reviews and meta-analysis
Randomized controlled trials
Cohort studies
Case-controlled studies
Case series and case reports
Expert opinion
Examples of direct medical costs
Drug prep and administration (including home infusion supplies)
Inpatient direct costs: hospital bed, administration, medical staff, surgeries, procedures, labs
Outpatient direct costs: office and clinical visits
Examples of direct non-medical costs
Travel and lodging costs for patients and caregivers traveling to hospital or clinic
Household costs such as childcare or elder care
Home health aides, to help with dressing/bathing and other ADLs
Examples of indirect costs
Lost work time
Low work productivity
Morbidity: costs from having the disease and related disability
Mortality (death)
Examples of intangible costs
Pain, suffering, anxiety, fatigue
What type of cost?: A caregiver was hospitalized and needs to pay for eldercare
Non-medical direct cost
What type of cost?: Traveling and lodging costs to get to hospital/clinic
Non-medical direct cost
What type of cost?: pain/suffering
Intangible
What type of cost?: Lost work time
Indirect
What type of cost?: Morbidity (costs from having disease and related disability)
Indirect
____ represents the change in costs and outcomes when 2 treatment alternatives are compared
Incremental cost-effectiveness ratios
Incremental cost ratio formula
Incremental cost ratio = (C2 - C1) / (E2 - E1)
C-costs
E-effects
____ is used when 2 or more interventions have demonstrated equivalence in outcomes and the costs of each intervention are being compared
Cost-minimization analysis (CMA)
____ is a systematic process for calculating and comparing benefits and costs of an intervention in terms of monetary units (dollars)
Cost-benefit analysis (CBA)
____ is used to compare the clinical effects of 2 or more interventions to the respective costs
Cost-effective analysis (CEA)
Cost minimization analysis limitation
Can only compare alternatives with demonstrated equivalent outcomes
(example: captopril vs lisinopril, both considered therapeutically equivalent in literature)
Cost effectiveness analysis limitation
Cannot compare different types of outcomes
(example: asthma program vs diabetes program)
____ is a specialized form of CEA that includes QOL component of morbidity assessments (QALY or DALYs)
Cost-utility analysis
