Epidemiology Week 2: 2. Measures of Disease Frequency

what are the measures of disease frequency?

• risk

• incidence rate

• prevalence

definition: measure of number of new cases of disease/health outcome of interest that develops in population at risk during a specified time period

incidence

definition: Measure of the number of new cases of a disease/health outcome of interest that develops in a population at risk during a specified time.

risk

what is this?

• Measured in a closed population

• The proportion of the population (BLANK) for the disease who becomes diseased within a given period of time

• Also known as cumulative incident (BLANK) or incidents proportion

• Must interpret only with specified time period

risk

which term?

• Probability of those who will develop the disease

• Does not include those who already have the disease or cannot get the disease

• Example: someone with hysterectomy vs endometrial cancer

risk

which term?

• numerator: # of new cases during time period of interest

risk

which term?

• denominator: includes only people at risk

risk

solve for risk and cumulative incidence: 5000 residents w/o diabetes and at end of 5-yrs 100 new cases

risk: 100/5000 =0.02

cumulative incidence: 2/100 over five years (0.2×100=2)

how to calculate person-incidence rate and why would you need to?

• number of new cases/person-time

• cohort w unequal follow-ups, people enter and drop at various time points

what does person-time account for?

• time at risk

• persons at risk

what is person-time?

a measure that represents the total time that participants are at risk of developing a health outcome during a study. It is calculated by summing the time each individual contributed to the study while they were still eligible to experience the outcome

calcualte person-time

4+7+2+5+4+3+1= 26

2 cases

2/26=0.077 or 7.7 cases per 100 person-years

what are the three time units incidence rate can be interpreted by?

• person-days

• person-months

• person-years

definition: proportion of a defined population that has disease at a specific point in time or during a specified time period

prevalence

which term

• disease status rather than occurrence

• reflects incidence and survival of the disease

prevalence

which term?

prevalence

calculate prevalence: 44,000 pts over five yra. 1100 had clicking or crepitus. what is the prevalence rate of clicking/crepitus in this group?

(1100/44000) = 0.025 ×100 = 2.5 cases per 100

what are the two options for resolution?

recovery or death

what is the relationship btwn prevalence and incidence of chronic diseases?

• low incidence and long duration

• high prevalence

• ex: diabetes, HIV, Crohn’s

what is the relationship btwn prevalence and incidence of acute diseases?

• high incidence, short duration

• low prevalence

• ex: common cold

why would incidence stay low but prevalence continues to rise for a chronic disease like HIV?

• new drugs (1996)

  • better prognosis

  • life expectancy increased significantly

general things to report w frequency measure:

• who

• what

• where

• when

• who: describe your population

• what: condition of interest

• where (location)

• when (time frame)

definition: an exposure that is statistically related in someway to an outcome

risk factor

Definition: an environmental, behavioral, or biological factor confirmed by temporal sequence usually in a longitudinal studies which if present directly increases the probability of a disease occurring and if absent or remove reduces the probability

risk factor

T or F: risk factors are part of the causal chain or expose host to causal chain

true

T or F: once disease occurs removal of a risk factor results in a cure

false

which type of studies are necessary to demonstarte risk factors?

prospective

definition: an exposure which is associated with an outcome only in cross-sectional data

risk indicator

can a risk indicator be a probable risk factor?

yes but caution is needed bc cross-sectional relationships can be deceptive

risk factor implies…

causality, so apply only when time sequence s established by prospective studies

when to use risk indicator?

wen you need to impute risk from cross-sectional data