Glossary Terms Flashcards
hypothesis stating that there is no difference between the study groups
null hypothesis
probability that a difference between study groups would have occurred if the null hypothesis was true
P value
range in which we can be approximately 95% certain that the population value lies
95% confidence interval
When the null hypothesis is incorrectly rejected; a difference between groups is statistically significant although a clinically important difference exists
alpha error; false positive
Type I Error
When the null hypothesis is incorrectly accepted; a difference between groups is not statistically significant although a clinically important difference exists
Type II Error
number of new cases of a condition that develop in a population during a defined time period (cases/year)
incidence
total number of people in a population with a condition during a given point in time (cases at birth)
prevalence
statistic used to test whether the rate of an outcome is significant;y different between two or more exposure groups.
This test provides a probability that the outcome and exposure are independent
Chi-Square test
statistic used to test whether there is a linear trend for an outcome to increase or decrease over the range of an ordered categorial exposure variable
Chi-Square test for trend
the probability of an event or outcome occuring - such as infection, death, cure
risk
ratio of the probability of the outcome occurring in the exposed group, divided by the probability of the outcome occurring in the non-exposed group
relative risk
the probability of an event (p) occurring divided by the probability of that event not occurring (1-p)
odds
radio of the odds of the outcome occuring in one group divided by the odds of the outcome occurring in another group
odds ratio
initial trial of a new treatment to assess SAFETY and feasibility in a small group of volunteers who do not have the disease or patients with symptoms
Phase I Trial
clinical trial to measure EFFICACY - the effect of treatment under ideal conditions in patients with the disease
Phase II Trial
large randomized control trial or multi-center study to measure EFFECTIVENESS in the community - the effect of treatment in GENERAL CLINICAL PRACTICE
Phase III Trial
POST MARKETING survery to measure rare adverse events
Phase IV Surveillance
All participants are analyzed in the group to which they were allocated, regardless or subsequent events such as non-complicance or withdrawal from the study. This provides a conservative estimate of treatment effect that is not influenced by cofounders
Intention-to-Treat analysis
only participants with final study outcomes are included in the data analysis but participants are maintained in the group to which they were allocated - results may be influenced by bias and cofounders
Available-case analysis
participants are re-grouped according to the treatment they actually received, irrespective to te treatment to which they were allocated - by using this method there is no control of cofounders
treatment-received-analysis
number of people who need to receive a new treatment to prevent one adverse event from occurring
number-needed-to-treat (NNT)
frequency of the outcome in the control (current best practice treatment or placebo) group
control event rate (CER)
frequency of the outcome in the experimental (new treatment) group
experimental event rate (EER)
the reduction in risk (probability of the outcome) that is conferred by the new treatment
absolute risk reduction (ARR)
the outcome measurement in a study - the variable of interest such as the primary illness or disease status indicator
outcome variable
characteristic that is hypothesized to influence the outcome variable - in clinical studies this is often the group to which patients have been randomized. In cross sectional and cohort studies these are often exposure variables
explanatory variables
stats used when the outcome measurement has a distribution that is approximately normal
parametric statistics
squared term that describes the total variation in the sample
variance
meausre of variability that describes how far the data spreads on either side of the central mean value. It is the square root of the variance and therefore is in the same units as the data values
standard deviation (SD)
data points at the extremities of the range or separated from the normal range of data values. Data points are MORE THAN 3 SD from the mean are usually considered this
outlier
calculated by dividing a mean value by its SE - gives a number from which the probability of the event occurring is estimated from a t-distribution
T-value
this is closely related to a normal distribution but depends on the number of cases in the sample
T-distribution
test to measure whether a continuous outcome variable with a normal distribution is significantly different between two groups - example between male or female or between an intervention and a control group
independent samples T-test
test used to compare the mean values of two independent samples using a normal distribution - only used when the sample size is very large or the mean and SD of the population are known
unpaired Z-test
distance between two mean values, described in units of their SD that describes the relative magniftute of the difference between two groups
Effect size
Pearson’s correlation coefficient that measures the strength of a linear relationship between two continuous normally distributed variables
R value
coefficient of determination is equal to the squared correlation coefficient and provides an estimate of the percent of variation in one variable that is explained by the other variable
R ^2
sample taken from a population in which all people have an equal chance of being selected
random selection
range of values in which the majority of people in a population expected to lie
normal values
distance between an observed value and its predicted value - in this case the value predicted by the regression line
residuals
regression line through a set of data points calculated to minimie the sums of squared residuals
line of best fit
study which is conducted to measure rates of disease in a population or to meausre associations between exposures (risk factors) and disease
observational study
study conducted to test the effect of a treatment or intervention
experimental study
study which is conducted to measure whether a new treatment is superior or equivalent to no treatment or an existing treatment - participants are randomly allocated to the study groups
randomized controlled trial
study in which participants receive two or more treatments given consecutively usually in random order. The response to the first treatment can be contrasted with the response to the second treatment in the same participants
cross-over trial
test to measure whether the means of two related continuous meausres are different from one another, typically measurements taken from the same participants on 2 occasions
Paired T-test
outcome of interest which is typically death but can be non-fatal or favorable outcome such as discharge from hospital
event
used to describe participants who withdraw from the study or who do not experience the outcome of interest
censored observations
statistic used to compare the event rate over time between two or more study groups - also called long-rank test
Kaplan-Meier statistic
risk of the event in a study group divided by the risk of the event in a reference group
Hazard ratio
test used to confirm a disease in people who present with signs or symptoms
diagnostic test
test regarded as the most accurate method available for classifying people as disease positive or negative
gold standard
probability of a positive test in a person with the disease compared to the probaility of a positive test in a person without disease
likelihood ratio
proportion of test-negative people who do not have the disease
negative predictive value
proportion of test-positive people who have the disease
positive predictive value
test used for early identification of disease in a population without symptoms
screening test
proportion of disease positive people who are test-positive
sensitivity
proportion of disease-negative people who are test-negative
specificity
meausre of precision with which the mena value has been measured
Standard Error (SE)
