174 Ch2

Prevalent Cases

All cases, individuals with outcome of interest regardless when diagnosed

Incident Cases

New cases, individuals who change in status over a specified period of time

When to use prevalence and when counnted

Counted in prevalence until they recover or die. Appropriate for stable chronic conditions

Incidence Risk

• Appropriate when population at risk is fairly fixed

• Assumes that all people contributed the same amount of time

• Denotes the probability of developing the disease

  • Interpretation: The risk of [developing disease] is:

Incidence Rate

• Appropriate when the population at risk is dynamic

• Denominator is the total person time

• Interpretation: IR of the exposed developed the disease each year OR there were IR cases per 1000 persons

Person Time

length of time that an individual stays at risk, used when population is dynamic.

Stops when subject: Dies, leaves study, develops disease, LTFU

Counting Person Time

Case 1: Known Obs Period

• LTFU or Developed = 0.5

• At risk = 1

Case 2: Unkown Obs Period

[(n_start + n_end)/ 2] * T

Odds

• ratio of the probability that an event will occur to the probability that the event will not occur

• easier to calculate, has clear meaning to clinicians and lay-people and can be used when risk or rate data are not available

• Interpretation: The odds of disease is [odds] or [odds]^-1

Ratio Effect Measures

relative strength of the association between the exposure and a disease or health outcome compared with the absence of exposure or less exposure

Risk, Rate, Odds, Prevalence Ratio

Difference Effect Measures

measures answers the question “How much could have been prevented if we eliminated the exposure?”

indicate whether the exposure (risk factor) is associated with a large or small number of cases (O+)

Interpretation for RR, RD, and Prev%

• RR: Those who are exposed were approximately RR times more likely to have the disease than those who are unexposed. The risk for the outcome was RR-1% more likely among those who are exposed compared to unexposed.

• RD: Among those who are exposed, the risk of disease is RD higher than those who are unexposed.

• Prev%: We could have prevented Prev% of the E+O+ if they were unexposed

Interpretation for IRR and IRD

• IRR: Those who are exposed were approximately IRR times more likely to have disease compared to unexposed. The risk for diseases was IRR-1% more likely among exposed than unexposed

• IRD: Among those who are exposed, the rate of disease was IRD cases higher per 1000pm

Further Interpretation of IRR and RR

RR = 1 ; No association, risk in exposed is equal to unexposed

RR > 1 ; Exposure is a risk factor, exposed higher

RR < 1 ; Exposure is a protective factor, exposed is lower

Odds Ratio

ratio between the odds of the health outcome in the exposed to the odds in the unexposed

Interpretation of Odds Ratio

• OR = 1 ; No association, exposure unrelated

• OR > 1 ; exposure is risk factor, positively related to disease

• OR < 1 ; exposure is protective factor, negatively related

Exposure increases the estimated risk of disease by OR

When is odds ratio good as an approximation of RR/IRR

1. When the cases studied are representative of all people with disease in the population

2. When the controls are representative of all people without disease from population

3. When the disease being studied is rare