Biostats Flashcards
Continuous increases by the same amount and can be characterized as either:
Ratio data (true, meaningful 0) ex. age, height, B Interval data (no meaningful 0) ex. temperature scales
Discrete data is also known as categorical data and can be characterized as either…
Nominal data (no set order of categories) ex. gender Ordinal data (ordered, ranked categories) ex. 0-10 pain scale
mean is the average value and is preferred for…
continuous data that is normally distributed
median is the value in the middle and is preferred for…
ordinal data or continuous data that is skewed
mode is the value that occurs most frequently and is preferred for…
nominal data
The spread (variability) of data can be described as either…
1) the range (difference between the lowest and highest value)
2) standard deviation (how spread out the data is from the mean)
Large sample sets of continuous data tend to form a Gaussian or…
“normal” bell-shaped distribution
When the distribution of data is normal, the curve is symmetrical. At the center point, the…
mean, median, and mode are the same value. 68% of the values fall within 1 SD of the mean and 95% of the values fall within 2 SDs of the mean.
Data with skewed distributions is not…
symmetrical (usually occurs when the sample size is small and there are outliers in the data)
With skewed distributions, the mean is pulled away from the center and the best measure of central tendency is the…
median
The skew can be either…
positive (right) or negative (left)
In a study, the independent variable is part of the study design and is…
changed and selected by the investigator to determine the effect on the dependent variable (outcome)
Examples of independent variables are the…
inclusion criteria (age, gender, etc), intervention/study drug
In a study, the dependent variables are the outcomes being measured and may include…
HbA1c, BP, cholesterol, etc.
A null hypothesis states that there is no…
statistically significant difference between groups (ex. drug efficacy= placebo efficacy)
The alternative hypothesis states that there is a…
statistically significant difference between groups. (if so, the researcher rejects the null hypothesis, or fails to accept it)
In a study, alpha refers to…
the maximum permissible error margin. It is the threshold for rejecting the null hypothesis (commonly set at 5%, or 0.05)
The confidence interval (CI) shows the precision of a result and is = to:
1- alpha
If alpha=0.05, this means that there is 95% probability or…
confidence that the conclusion is correct
When comparing difference data, the result is statistically significant if the CI…
does not include 0
When comparing ratio data (RR, OR, HR), the result is statistically significant if the CI…
does not include 1
Type I errors are related to alpha and occur…
when the null hypothesis is rejected in error
Type II errors are denoted as beta and occur…
when the null hypothesis is accepted when it should have been rejected
Beta is typically equal to…
0.1 or 0.2
Power is the probability of avoiding a type II error and is equal to…
1- beta (typically 0.9 or 0.8)
Risk =
(# of subjects in a group with an unfavorable event) /(total # of subjects in a group)
Relative Risk (RR)=
risk in treatment group/risk in the control group
If RR =1, then there is…
no difference in risk
If RR>1, then there is a…
higher risk in the tx group
If RR<1, then there is a…
a lower risk in the tx group
In a metoprolol study, the RR of HF progression was 57%. This means that…
patients treated with metoprolol were 57% as likely to have progression of disease as placebo-treated patients
The relative risk reduction (RRR) indicates…
how much the risk is reduced in the treatment group, compared to the control group
RRR=
1-RR
RRR=
(% risk in control group- % risk in the treatment group)/
(% risk in the control group)
If RRR=43%, this means that metoprolol-treated patients were…
43% less likely to have HF progression than placebo-treated patients
The absolute risk reduction (ARR) is more useful in practice and is the…
absolute difference in outcome rates between 2 groups
ARR=
(% risk in the control group) - (% risk in treatment group)
If ARR=12%, this means that for every 100 patients treated with metoprolol…
12 fewer patients will have HF progression
The number needed to treat (NNT) is the # of patients who need to be treated for a certain period of time in order for…
1 patient to benefit
NNT=
1/ [(risk in control group) - (risk in treatment group)]
NNT=
1/ARR (as a decimal)
If the NNT=8.3, then the NNT is…
9 (round up)
The number needed to harm (NNH) is the # of patients who need to be treated for a certain period of time in order for…
1 patient to experience harm (adverse event)
NNH=
1/ARR (as a decimal; use absolute value if negative)
If the NNH=90.9, then the NNH is…
90 (round down)
Case-control studies are not suitable for RR calculations. In this case, the…
odds of unfavorable events, or odds ratio (OR), is calculated instead. May also be used in cohort or cross-sectional studies.
OR=
(AD)/(BC)
In a case-control study of serotonergic ADs, an OR=23% means..
that serotonergic ADs are associated with a 23% increased risk of falls with fracture
In a case-control study of serotonergic ADs, an OR=1.23 means that serotonergic ADs…
are associated with a 23% increased risk of falls with fracture (in this case, OR>1)
In a survival analysis, a hazard rate (HR) is used instead of risk. A HR is the rate at which…
an unfavorable event (usually mortality) occurs within a short period of time
HR=
(hazard rate in the tx group)/(hazard rate in the control group)
If OR/HR=1, the event rate is…
the same in both groups
If OR/HR>1, the event rate is…
higher in the tx group
If OR/HR<1, the event rate is…
lower in the tx group
The primary endpoint is the main result that is measured to see if the tx had a significant benefit.
Ex. HF progression
The composite endpoint combines multiple individual endpoints into one measurement.
Often used to increase the likelihood of reaching a statistically significant benefit with a smaller, less costly trial. However, all endpoints must be similar in magnitude and have similar, meaningful importance to the patient.
When a composite endpoint is used, the FDA requires each individual endpoint…
to be measured and reported when the composite endpoint is used
If continuous data is normally distributed, which parametric statistical tests can be used?
- if 1 group (one-sample t-test or paired t-test if 1 group has before and after measures)
- if 2 groups (unpaired student t-test)
- if 3 groups: ANOVA (F-test)
For categorical (nominal or ordinal) data, which statistical tests should be used?
- if 1 group: Chi-square test
- if 1 group has before and after measures: Wilcoxon Signed-Rank test
- if 2 groups: Chi-square or Fisher’s exact test. Mann-Whitney (Wilcoxon Rank-Sum) test may be preferred for ordinal data.
- if 3 or more groups: Kruskal-Wallis test
If continuous data is not normally distributed, which non-parametric statistical tests can be used?
- If 1 group: sign test or Wilcoxon Signed-Rank test (if one group has before and after measures)
- If 2 groups, Mann-Whitney (Wilcoxon Rank-Sum) test
- If 3 or more groups: Kruskal-Wallis test
Fisher’s exact test is basically the same as Chi-Square, except that…
it is used for smaller sample sizes
Correlation is used to determine if…
one variable changes, or is related to, another variable
A correlation does not prove a…
causal relationship between variables
Spearman’s rank-order correlation, or Rho, is used to test correlation with…
ordinal, ranked data
Pearson’s correlation coefficient, or r, is used for…
continuous data (r ranges from -1 to 1) 1= strong positive correlation; -1= strong negative correlation
Regression is used to describe the relationship between a dependent variable and…
one or more independent variables
There are 3 types of regression…
1) linear, for continuous data
2) logistic, for categorical data
3) Cox regression, for categorical data in a survival analysis
Senstivity refers a true positive and 100% means that…
a test will be positive in all patients with the condition
Specificity refers to a true negative and 100% means that…
a test will be negative in all patients without the condition
Intention to treat analysis includes all…
data for all patients originally allocated to each tx group, even if the patient did not complete the trial according to the study protocol
A per-protocol analysis uses…
data only from the subjects who completed the study according to the protocol (less applicable to real-world)
Equivalence trials are designed to show…
that a new treatment has roughly the same effect as a reference treatment
Non-inferiority trials are designed to show…
that a new treatment is no worse than the current standard based on the predefined non-inferiority (delta) margin
The delta margin is…
the minimal difference in effect between the 2 groups that is considered clinically acceptable base on previous research
Forest plots are used either for a single study or more commonly, to show results from a meta-analysis. They provide CIs for…
different data or ratio data. The vertical line is the line of no effect (0 for difference data and 1 for ratio data).
In a forest plot, a significant benefit has been reached when…
the data falls to the left of the line (data to the right indicates significant harm)
Economic outcomes include…
direct, indirect, and intangible costs
Clinical outcomes include…
medical events that occur as a result of the treatment or intervention
Humanistic outcomes include…
consequences of the disease or treatment as reported by the patient or caregiver (e.g. patient satisfaction, quality of life)
The average cost-effectiveness ratio is the cost per outcome of one treatment independent of other alternatives.
Ex. if a tx costs $50 to generate successful outcomes in two patients, the average cost-effectiveness ratio is $25/treatment success ($50/2 successfully treated patients)
Incremental Cost-Effectiveness Ratio (ICER)…
represents the change in costs and outcomes when two tx alternatives are compared.
ICER=
(C2-C1)/(E2-E1)
Cost-minimization Analysis (CMA) compares..
costs of interventions with therapeutic equivalence (easiest to calculate, but limited)
Cost-benefit analysis (CBA)…
compares benefits and costs in monetary units (can be hard to assign a $ amount to a benefit)
Cost-effectiveness analysis (CEA) is the most common type and is used to compare…
the clinical effects of 2 or more interventions to the respective costs. Outcomes are similar to clinical trials (e.g length of stay, cure rate, change in BP). Inputs are measured in dollars (e.g. pharmacist time, drug costs)
Cost-utility analysis is a specialized form of CEA that includes 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)
Order the medical studies from most-reliable to least-reliable (with 1 being the most reliable.
1) Meta-analysis 2 RCTs 3) Cohort studies 4) Case-control series 5) Case series
A clinical trial is conducted on a new drug, LoSod. LoSod is found to reduce mean serum sodium to 131 mEq/L (95% confidence interval: 128.7-133.3 mEq/L). What is the correct interpretation of this study result?
There is a 95% chance that the CI range contains the true population value for reduction in serum Na with LoSod.
cohort study
start with the exposure, looking for disease
case-control study
start with the disease, looking for exposure
When data is normally distributed the…
mean, median, and mode are all the same value
A small, randomized, double-blind trial was conducted to evaluate a new tx in patients with acute MI. The primary endpoint, reduction in all-cause mortality, was not statistically significant. Which of the following is an accurate conclusion?
the trial may not have adequate statistical power (performing a larger trial will increase power, which increases the probability that a type II error will be avoided.
A case-control study evaluated the incidence of pulmonary fibrosis due to amiodarone tx. It was determined that of 250 patients who were exposed to amiodarone, 22 developed pulmonary fibrosis. This was compared to a control group of 250 patients who did not have exposure to amiodarone and in which only 2 patients developed pulmonary fibrosis. The OR for developing pulmonary fibrosis with amio was calculated to be 12. Which of the following is correct?
Patients exposed to amiodarone were 12 times more likely to develop pulmonary fibrosis than those not exposed.
Pharmacoeconomic analyses…
identify and measure the costs and outcomes to ensure healthcare resources are used wisely
A high statistical power increases the confidence that…
the null hypothesis was rejected correctly