Exam 1 Flashcards

Front 1

Cumulative Incidence

Back 1

CI = # new cases of disease during specified time period / # people in at-risk population at beginning of time period

"The 10-day cumulative incidence of disease is ___%."

- Technically proportion (not rate)

- Estimates the probability (risk) that a person will develop disease during a specified time

- Assumes entire pop is followed entire time

Front 2

Attack Rate

Back 2

# new cases of disease / total pop. at risk for a ‘limited’ period of observation

"The 1-day cumulative incidence of disease is ___%."

- Technically proportion (not rate)

- Type of CI measure

- Useful when observing outbreak (often due to very specific exposure)

Front 3

Case-Fatality Rate

Back 3

Number of deaths from disease / # cases of that disease in specified time period

"The 10-day cumulative incidence of disease is ___%."

- Type of CI measure - Reflects severity of disease

Front 4

Incidence Rate

Back 4

IR = # new cases disease during specified time period / total at-risk person-time of observation

"The incidence rate of disease is ___ per 100,000 PYs. (can use 100, 1000, 1,000,000"

- pay attention to what PY you are using

- True rate (not proportion) - a.k.a. "incidence density"

- more accurate b/c doesn’t rely on same assumptions as CI (more accurate rep of real world scenarios

Front 5

Mortality Rate

Back 5

total # deaths / total person-time of follow-up

- Not true rate b/c didn't follow all 231 mil ppl for full year (see ex in slides)

- Can be cause-specific or age-specific, depending on what you're observing

Front 6

Prevalence

Back 6

P = # of existing cases of disease at a specified time / total # of people in population at that time

"The prevalence of disease in the specified population is ____%."

- Estimate of probability that a person in a population has a disease at a specified time

- Proportion (not rate) - can be Point (at 1 specified time) or Period (over certain period of specified time)

Front 7

Birth defect rate

Back 7

# children born with defects / total # of births

- type of point prevalence (probability at specific time of birth - 1 point in time)

Front 8

Cumulative Incidence & Incidence Ratio (their relationship)

Back 8

CI ≈ IR x time period

- Relationship between CI and IR

- Only holds for small CI (<10%).

- Also assumes population is fixed.

- %

Front 9

Point Prevalence & Incidence Ratio (their relationship)

Back 9

P / (1-P) = IR x D

P ≈ IR * D (when prevalence is low, disease is rare, <10%)

- D=avg duration of disease; IR=incidence rate; P= POINT prevalence

- Assumes incidence and duration constant over time (steady state)

- If incidence is low but duration is long, means point prevalence is high relative to incidence

- If incidence is high but duration is short, means point prevalence is low relative to incidence

- Changes in prevalence can be due to changes in duration or incidence or both

- %

Front 10

Age-adjusted rate

Back 10

∑(age-specific rate * adjustment weight)

- forces weighted average of each age group to equal that of the general pop for each observed subgroup

Front 11

Risk Difference

Back 11

"The excess risk during the ___ period of x disease among the exposed group is ___%."

- how much greater is the risk of outcome if you're exposed than if you're not exposed

- (i.e. if RD=0, exposure has no effect on outcome)

- a.k.a. cumulative incidence difference, incidence rate difference, and prevalence difference

Front 12

Excess Number of Cases

Back 12

RD x number exposed

"The excess number of cases that is associated with the exposure is ____ [e.g. 117.78]."

- What numbers of the disease outcome are due to being exposed

Front 13

Population Risk Difference (1)

Back 13

"The excess risk of [disease] for the total population is ___%."

- 1st population risk difference formula

- Rtot = risk of total pop; Runexp = risk among unexposed

- i.e. PRD will be low if exposure is rare

Front 14

Population Risk Difference (2)

Back 14

"The excess risk of [disease] for the total population is ___%."

- 2nd population risk difference formula

- Pexp_tot = proportion of tot pop that's exposed

Front 15

Attributable Proportion of Exposed (1)

Back 15

"___% of [disease] cases that occurred among [exposed group] is associated with them being [exposed]."

- Rexp = risk (IR, CI) among exposed

- Runexp = risk (IR, CI) among unexposed

- a.k.a. etiologic fraction, attributable risk percent, attributable risk percent among exposed

Front 16

Attributable Proportion of Exposed (2)

Back 16

"___% of [disease] cases that occurred among [exposed group] is associated with them being [exposed]."

- Useful when you only have RR (not the factors that make up RR)

Front 17

Attributable Proportion of Total Pop. (1)

Back 17

"___% of cases that occurred in the total population are associated with being [exposed]."

- Rt = risk (IR, CI) among total population

- Ru = risk (IR, CI) among unexposed

Front 18

Attributable Proportion of Total Pop. (2)

Back 18

"___% of cases that occurred in the total population are associated with being [exposed]."

- Pe = proportion of total population that is exposed. Useful when you only have RR, but must have Pe too.

Front 19

Relative Risk

Back 19

"If you are [exposed], you are ___ [i.e. 17..33] times as likely as a [non-exposed] to develop [disease]."

- Tells you how many times higher or lower the disease outcome is among the exposed compared to the unexposed

- Based on ratio of 2 measures of frequency

- Tells us magnitude or strength of association

- dimensionless (no units) and ranges 0 to infinity

- RR=0 => no assoc

- RR>1 => exposed have > risk than unexp

- RR<1 => exposed have < risk than unexp

Front 20

Put RR into a %

Back 20

RR = (RR-1) x 100

"[Exposed] are ___% more likely to develop [disease] as [non-exposed]."

Front 21

Cumulative Incidence Ratio

Back 21

Front 22

Incidence Rate Ratio

Back 22

Front 23

Prevalence Ratio

Back 23

- Not considered measure of relative risk

- Dr. Niccolai said she probably won't test this one