Clin Epi and Stats Flashcards
Define construct validity
The ability of a test to measure what it is designed to measure (and not something else)
Ex. test for depression should measure that and not anxiety
Define criterion-related validity
The ability of a test to correlate to the gold standard, how accurately it measures what it was designed to
What percent did older adults 65+ increase from 2016 to 2021?
19.3% increase
What age group was the fastest growing from 2016 to 2021?
Millenials
85+ is one of fastest growing
100+ grew by 16%
What is the life expectancy for a man/woman born in 2021?
Man 80 years
Woman 84 years
How many people are 85+ yo?
861, 000 (doubled since 2001)
What percent of people 85+ live in collective dwelling?
> 25%
What is the definition of life expectancy?
The average number of years that members of a population are expected to live
What is the definition of life span?
The maximum duration or greatest age reached by individual in a population or species (122 years for humans)
How do you calculate relative risk?
RR = % disease exposed / % disease not exposed
= (a/a+b) / (c/c+d)
What is the definition of relative risk?
The RATIO of the probability of an event occurring in an exposed group to non-exposed group
RR>1 means exposure increases risk of disease
RR<1 means exposure decreases risk of disease
How do you calculate relative risk reduction?
CER-EER / CER
Control event rate
Exposed event rate
How do you calculate absolute risk reduction?
CER - EER
How do you calculate NNT?
1/ARR
ARR = CER-EER
What is the definition of attributable risk?
The portion of disease rate attributable to a certain exposure
For example AR of smoking and lung cancer is 75% which means that 75% of lung cancers are caused by cigarette smoking
What are likelihood ratios?
Used to interpret diagnostic tests
The higher the +LR the more likely a patient has disease or condition
LR+ = Sn / 1-Sp
LR- = 1-Sn/Sp
What is definition and formula for Sn and Sp?
Sensitivity = SnOUT = High Sensitivity means if it is negative you can rule out
Probability positive test in patient with disease
Sn = TP / TP+FN
Specificity = SpIN = High Specificity means if positive you can rule in
Probability of negative test in person without disease
Sp= TN / TN+FP
What is the PPV?
Positive predictive value
TP / TP+FP
Probability that subjectives with positive test actually have the disease
NPV = TN/TN+FN
What are some types of bias?
- Selection bias
- Recall bias
- Detection bias
What is a cohort study?
Look at those with and without an exposure
Follow them to see if a disease develops
What is a case control study?
Look at those who have a disease vs. those without and compare exposures
Define a confounder
A variable that is associated or has a relationship with both the exposure and the outcome of interest
What is the difference between a cross sectional study and a longitudinal study?
Cross sectional - collect data from a specific point in time
Longitudinal - collect data repeatedly over an extended period of time
Define intra-rater reliability vs. inter-rater reliability?
Intra-rater reliability - degree of agreement of same examiner doing test repeatedly
Inter-rater reliability - degree of agreement of different examiners doing the same test
Define power
The ability of a test to detect an effect, if the effect actually exists
Detects the probability of rejecting a false null hypothesis
What is quality assurance in protocol design?
Program to minimize the possibility of systematic discrepancies in treatment management amongst participating institutions
Define generalizability
The degree to which the results of a study are generalizable to the population more broadly (Beyond those included in study)
Define validity
The degree to which the results of study are likely to be true, believable and free of bias
Define reliability
The reproducibility of a measurement procedure
What is the ceiling effect?
Test items aren’t challenging enough for a group
Test score won’t improve even though patient has clinically improved because already reached highest score
MMSE has strong ceiling effect, MOCA not as much
What is the floor effect?
When data can’t take on lower value than particular number
What is a type I vs. type II error?
Null hypothesis = no difference between treatment and control
Type I error = rejecting null hypothesis when it is true (false positive)
Type II error = failing to reject null hypothesis when it is false (false negative)
According to the Canadian Dementia Strategy executive summary what percent of Canadians 65+ were living with diagnosed dementia?
7%
According to the Canadian Dementia Strategy executive summary what is the average number of hours spent by family/caregivers to support a person with dementia?
26 hours
According to the Canadian Dementia Strategy executive summary what is the projected total health care costs and out of pocket caregiver costs of dementia in 2031?
$8.3 billion total cost
$16.6 billion total health care cost and out of pocket caregiver costs
