Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
80 Cards in this Set
- Front
- Back
study of factors affecting the occurrence and distribution of diseases in human pop.
|
medical epidemiology
|
|
summarizes and evaluates data obtained from epidemiologic and other studies
|
biostatistics
|
|
ratio of # new cases of disorder/event to the # potential cases
|
incidence rate
|
|
ratio of the # of present cases of disorder/event to # of potential cases
|
prevalence rate
|
|
2 types of prevalence rate
|
point and period
|
|
# of indivs who have an illness at a specific point in time divided by # who could potentially have it on that date
|
point prevalence
|
|
# of indivs who have an illness during specific time period divided by # who could have it during that period
|
period prevalence
|
|
prevalence (P) =
|
incidence rate (IR) x average duration of disease process (t)
P = IR x t |
|
2 ways a person can leave the population of prevalent cases
|
die or recover
|
|
compares incidence rate of disorder among individuals exposed to risk factor with incidence rate of disorder in nonexposed indivs
|
relative risk
|
|
relative risk is used to analyze what type of study?
|
cohort study
|
|
used to determine what would happen in study pop if risk factor were removed
|
attributable risk
|
|
AR =
|
IR (exposed) - IR (nonexposed)
|
|
attributable risk is used to analyze what type of study?
|
cohort study
|
|
odds ratio is used to analyze what type of study?
|
case-control study
|
|
odds ratio =
|
AxD/BxC
A=exposed and ill B=nonexposed and ill C=exposed and not ill D=nonexposed and not ill |
|
types of bias
|
selection, recall, sampling
|
|
type of bias that occurs if subjects are permitted to choose what group to be in rather than be assigned
|
selection
|
|
type of bias that occurs if the investigator purposely chooses which pts go into the groups
|
selection
|
|
type of bias that occurs if subjects who volunteer to be in study don't represent the population being studied
|
sampling
|
|
type of bias that occurs if knowledge of presence of disorder alters the way the subject remembers hx
|
recall
|
|
strategies to reduce bias
|
blind studies, crossover studies, randomized studies
|
|
type of study where pt doesn't know what tx he is getting
|
single blind
|
|
type of study where pt and evaluator don't know what tx he is getting
|
double blind
|
|
type of study where pts get both the placebo and real drug for certain amounts of time
|
crossover study
|
|
type of study where pts are randomly assigned to a group
|
randomized study
|
|
measures whether the results of test are similar when test is admin by a different rater or examiner
|
interrater reliability
|
|
measures whether the results of the test are similar when the person is tested repeatedly
|
test-retest reliability
|
|
measures whether the test assesses what it was designed to assess
|
validity
|
|
measure of the ability of test to identify people who have a disorder
|
sensitivity
|
|
measure of the ability of test to identify people who don't have a disorder
|
specificity
|
|
determines the likelihood that a person with + test has the disorder
|
positive predictive value
|
|
determines the likelihood that a person with - test doesn't have the disorder
|
negative predictive value
|
|
sensitivity =
|
TP / TP + FN
|
|
tests with high sensitivity can
|
rule out disease
|
|
specificity =
|
TN / TN + FP
|
|
tests with high specificity can
|
rule in disease
|
|
measure of % of test results that match the actual dx
|
predictive value
|
|
PPV =
|
TP / TP + FP
|
|
NPV =
|
TN / TN + FN
|
|
# times an event actually occurs divided by # times event can occur
|
clinical probability
|
|
type of incidence rate used to describe disease outbreaks
|
attack rate
|
|
predictive factor that affects an outcome
|
indep variable
|
|
outcome that reflects the effects of changing the indep variable
|
dependent variable
|
|
begins with subjects free of illness; determine incidence rates of illness between exposed and non-exposed
|
cohort study
|
|
2 types of cohort studies
|
prospective and historical
|
|
begins with subjects who have disorder and those who don't
|
case-control study
|
|
begins with info collected from group that provides snapshot in time of disease activity
|
cross-sectional study
|
|
summarizes the data obtained from research studies
|
descriptive statistics
|
|
provides a way to generalize results to entire pop by observing sample of that pop
|
inferential statistics
|
|
average distance of observations from their mean
|
standard deviation
|
|
standard deviation =
|
(variations squared + deviations squared) / (scores - 1), then take the square root
|
|
z score
|
standard normal score
|
|
difference between an indiv score and pop mean in units of standard dev
|
standard normal score
|
|
standard deviation divided by square root of # of scores in a sample
|
standard error
|
|
specifies the limits between which a given % of pop would be expected to fall
|
confidence interval
|
|
bell shaped curve is also called:
|
Gaussian distribution
|
|
what is special about a bell shaped curve?
|
the mean, median and mode are equal
|
|
what % of of pop score falls within 1 standard deviation of the mean
|
68%
|
|
degree to which the mean is resistant to random variation
|
precision
|
|
refers to likelihood of bias
|
accuracy
|
|
if the distribution is positively skewed (to the right) the tail is toward the ___ and the modal peak is toward the ___
|
right, left
|
|
statement based on inference, existing literature or preliminary studies that suggests a difference between 2 groups
|
hypothesis
|
|
hypothesis that postulates that no difference exists between 2 groups or more groups
|
null hypothesis
|
|
error that occurs when the null hypothesis is rejected even though it is true
|
type I (a) error
|
|
error that occurs when the null hypothesis is not rejected although it is false but there may not have been enough power to detect the difference
|
type II (b) error
|
|
chance of a type I error occurring
|
P (probability) value
|
|
ability to detect a difference between groups if it is truly there
|
power
|
|
if a P value is equal to or less than ___, it is unlikely that a type I error has been made
|
0.05
|
|
a P value of 0.05 or less is considered to be
|
statistically significant
|
|
what do the results of statistical tests indicate?
|
whether to reject or not reject the null hypothesis
|
|
statistical test using population parameters and usually used to id the presence of statistically significant diff b/t groups when pop is normal and sample size is lg
|
parametric tests
|
|
commonly used parameter tests
|
t-tests, analysis of variance (ANOVA) and linear correlation
|
|
refers to the degree of relationship between 2 continuous variables and can be assessed using linear correlation coeff that range b/t plus 1 and minus 1
|
linear correlation
|
|
commonly used nonparametric statistical tests
|
Wilcoxon's, Mann-Whitney, Kruschal-Wallis
|
|
statistical test used if distribution isn't normal or if sample size is small
|
nonparametric test
|
|
difference between the means of 2 samples
|
t test
|
|
differences b/t the means of more than 2 samples
|
analysis of variance
|
|
mutual relation b/t 2 continuous variables
|
correlation
|
|
diff b/t freqs in a sample and Fisher's exact probability
|
Chi-sq test
|