Epidemiology Module 2

Question 1

How to define population?

Answer 1

a group of individuals, population experience (individuals thru time) or population cross-section (individuals at a point in time)

Question 2

What are odds?

Answer 2

Pevent/Pnonevent

Question 3

2 measures of frequency:

Answer 3

prevalence (cases in a pop. at a given time) and incidence (new occurring cases/pop during specified time frame)

Question 4

Prevalence definition

Answer 4

new and existing outcomes of interest. can be measured on different time scales. 2 types: point (# cases/total specified pop. at time of outcome ascertainment) and period (# cases/total specified pop at mid-point of time period)

Question 5

What is prevalence affected by?

Answer 5

number of new cases, duration, rate of recover/death

Question 6

Incidence definition

Answer 6

new events. excludes prevalent cases. 2 measures: CI (incidence proportion) and incidence density (incidence rate)

Question 7

What is cumulative incidence and its assumptions?

Answer 7

fixed population experiencing a new event during a specified time period. Assumes follow up for entire period and static population.

Question 8

Cumulative Incidence (Risk) proportion

Answer 8

occurred in a specified time period/number of total people in population. Can be enumerated like 30-day mortality, monthly risk, or annual incidence (proportion). dimensionless.

Question 9

What are the difficulties with CI?

Answer 9

competing risks, multiple events, dynamic populations, loss to follow-up

Question 10

What is attack rate?

Answer 10

incidence proportion of becoming afflicted during an epidemic (usually infectious)

Question 11

What is case fatality rate?

Answer 11

proportion of people who then proceed to die (often no time period, infectious diseases, i.e. rabies)

Question 12

What is incidence density (IR)?

Answer 12

new events/accumulated person-time of observation. accounts for variable time to event and loss to follow-up, but not potential bias due to loss. i.e. deaths/person-days of patients at risk. count number of injuries instead of number of people. if repeated events, don’t consider independent and subsequent aren’t counted? provides ave. prognosis

Question 13

What are the technical difficulties with ID?

Answer 13

difficult to explain, waiting time. if steady-state conditions, the waiting time is the reciprocal of ID (i.e. 3.57 cases/person-year. 3.57^-1=.28 years, so you would wait about three months to see one case of the outcome

Question 14

CI interpretation vs ID interpretation

Answer 14

time mortality vs mortality; time risk vs outcome per person-time; time incidence (proportion) vs incidence (density). CI goes from 0-1 while ID is 0-infinity.

Question 15

Choosing CI vs ID

Answer 15

cumulative over duration of follow-up vs does not assume a particular time frame; difficulty with differing length of follow-up vs variable length of follow-up ok; difficulty with loss to follow-up vs handles loss to follow-up; difficulty with repeated events vs handles multiple occurrences; assumes fixed cohort vs handles dynamic population

Question 16

Issues with Prevalence and Incidence

Answer 16

Numerator: definition of a case and multiple episodes of a disease? Denominator: who is at risk? competing risks? treat previously occurring disease in the same individual? P = understand burden of disease; I = etiologic associations

Question 17

What is crude rate?

Answer 17

events in a stated time period/population at risk

Question 18

What are problems with crude comparisons?

Answer 18

age associations and issues. other factors may alter death and disease as well

Question 19

What is a specific rate (stratum-specific)?

Answer 19

events among members in a stated time period/total members of category

Question 20

What is an adjusted/standardized rate?

Answer 20

hypothetical event rate that would have occurred if the observed rates happened in a population that is distributed like the standard population. can mask important differences

Question 21

What is age standardization?

Answer 21

removes effect of age. fictional. direct and indirect methods.

Question 22

What is a direct adjustment?

Answer 22

identify study and standard population. e.g. summary age-adjusted rate (disease/death rate in study pop. if study pop. had same age distribution as standard pop.).

Question 23

What is an indirect adjustment?

Answer 23

used b/c limited info on study pop or its small so age-specific rates are unstable. Final measure is a ratio: standardized mortality ratio or standardized morbidity ratio. cannot be compared. only between study and standard.

Question 24

What is SMR?

Answer 24

observed deaths/#expected deaths

Question 25

How do you interpret SMR?

Answer 25

SMR = 1.0 --> no difference; SMR>1.0 --> higher than expected; SMR<1.0 --> lower than expected

Question 26

How to read a prevalence or incidence graph

Answer 26

P: Orange line = health related event; arrows = outcome that was present at beginning or end of the specified time period; circles = incidence of new outcome or resolution; I: circles represent event; arrows are people that have not experienced event at end of follow-up

Question 27

What is a catchment population?

Answer 27

People who use medical facility's services

Question 28

What is a steady state?

Answer 28

I=E in a pop

Question 29

What is a rate?

Answer 29

time is an integral part of denominator

Question 30

CI vs IR vs P

Answer 30

new cases/pop at risk; new cases/person-time at risk; existing cases/total pop

Question 31

What is the formulaic relationship between CI and IR?

Answer 31

CI=sum(IRiXti) --> don't need sum if constant IR and small CI

Question 32

What is the relationship between prevalence and incidence?

Answer 32

P/(1-p)=IRXD (d = duration) for a stable state and incidence and duration do not change. if the frequency of the disease is rare (<10%) the euation becomes P=IRxD