Measures of Association: Descriptive Studies Flashcards
What is the equation we need to know with 3 terms, such that x=y×z? What is x, y, and z
P=IR×D
P=Prevalence
IR=Incidence Rate
D=Duration
Which type of disease frequency (P, CI, IR): Percentage of dental students who get gingivitis over the fall semester.
Cumulative Incidence
Which type of disease frequency (P, CI, IR): Percentage of dental students who have gingivitis on the first day of class.
Prevalence
Which type of Disease frequency (P, CI, IR): Number of new cases of gingivitis in a year per 100,000 person-years.
Incidence Rate
Which type of disease frequency (P, CI, IR): Percent of dental students with gingivitis among 200 dental students.
Prevalence
True or False: Only the population at risk contributes to the denominator of cumulative incidence.
True
True or False: When calculating the incidence rate of disease, it is necessary to follow subjects for the same amount of time.
False
True or False: If the incidence rate of a very serious disease is 75/100,000 person-years and the prevalence in the population is 25/100,000 years, then the average duration of this disease must be 3 years.
False
answer=⅓ year
What is index determinant(group)?
- people w/characteristic
- exposed group
What is reference determinant (Group)?
- People w/o characteristic or contrasting level of characteristic or alternative characteristic
- unexposed group
What is the purpose of measures of association?
- compares disease frequency:
- prevalence
- cumulative incidence
- Incidence rate
What are the 2 types of Measures of Association?
-
Relative Measure of Association:
- Ratio b/w index(exposed) & reference (unexposed)
-
Absolute Measure of Association:
- Difference b/w index & reference
Relative Measures of Association
- oRatio b/w index vs reference groups
- No Units
- General Formula: RR=Re/Ro
-
Prevalence:
-
Prevalence Ratio:
- PR=Pe/Po
-
Prevalence Ratio:
-
Cumulative Incidence:
-
Risk Ratio: (ratio of CI)
- RR=CIe/CIo
-
Risk Ratio: (ratio of CI)
-
Incidence Rate
-
Rate Ratio:
- RR=IRe/IRo
-
Rate Ratio:
If RR=1….
- No Association
- b/w exposure & disease
If RR>1….
- Exposure is associated with INCREASED frequency of disease
If RR<1…
- exposure is associated with DECREASED frequency of disease
Interpret RR=0.75
- Subjects with X had 0.75 times the risk of developing Y compared to those without X……OR
- subjects with X had 25% less risk of developing Y compared to those without X
How do you calculate Excess relative Risk?
Excess Relative Risk=(RR-1)x 100%
What incidence ratio corresponds to a 20% increased incidence rate in the exposed vs unexposed group?
1.2
Excess Relative Risk=(RR-1)x100%
20%=(x-1)x100%
.20=x-1
1.2=x
What incidence ratio corresponds to a 200% increased incidence ratio rate in the exposed vs unexposed group?
3.0
Excess Relative Risk=(RR-1)x100%
+200%=(x-1)x100%
2=x-1
x=3.0
What incidence rate ratio corresponds to a 43% decreased incidence rate in the exposed vs unexposed group?
0.57
Excess Relative Risk=(RR-1)x100%
- 43%=(x-1)x100%
- .43=x-1
x=0.57
What is the excess RR if RR=7.43? and interpret
- Excess RR=(RR-1) x100%
- Excess RR=643%
- 643% increased risk
Absolute Measures of Association
- Difference (not ratio)
- No Units:
- EXCEPT Incidence rate
- General Formula: RD=RI-Rr
-
Prevalence
-
Prevalence Difference
- PD=Pe-Po
-
Prevalence Difference
-
Cumulative Incidence
-
Risk Difference
- RD=CIe-CIo
-
Risk Difference
-
Incidence Rate
-
Rate Difference
- RD=IRe-IRo
- Units=person-time<u>-1</u>
-
Rate Difference
Investigators used a longitudinal cohort study design to assess whether exclusive long-term pipe smoking was associated with the development of head and neck cancers. Among pipe smokers, 171 developed head/neck cancer over 155,086 person-years of follow-up. Among never smokers, 155 developed head and neck cancer over 620,344 person-years of follow-up
Based only on this information, which absolute measure of association can we calculate:
Incidence Rate Difference
=IRe/IRo
IR={#new casespopulation @risk}xtime
How do Relative and Absolute Measures of Association Differ?
- Both provide different information & perspectives
- Absolute measure
- Public Health Impact
- Relative Measure
- strength/magnitude of association
What are the different types of descriptive studies?
- Case Reports
- Case Series
- Ecologic Studies
- Cross-sectional studies
What are the different type of analytical studies?
- Case-control study
- Cohort study
Case Reports
- Descriptive Study of 1 person/case
What are case reports useful for?
- recognize & describe:
- new diseases
- new manifestations of old diseases
- Detect drug side effects
- Provide:
- disease mechanism
- information
- help develop hypotheses
What are the disadvantages of case reports
- No comparison group
- can’t be used for hypothesis testing
Case Series
- many studies on a specific type of case
- group of people w/similar problem
What are some strengths and weaknesses of case series?
- Strengths:
- Provide descriptive data on disease characteritics
- Weakness:
- no control group
- can’t be used for hypothesis testing
- no control group
Ecological Studies
- aka Correlational study
- examines rate of disease in relation to prevalence of exposure
- no individual-level info collected
- population usually identified by:
- place
- (city, state, neighborhood, country)
- time
- or combo
- place
- exposure and outcome measured at population level
Ecological Studies: Strengths & Weaknesses
- Strengths:
- quick & inexpensive
- done on available data
- good for early stage of knowledge
- when you want to study ecologic relationships
- Easy analysis when using correlation coefficient or linear regression
- Weakness:
- Ecological Fallacy
- No control for confounding factors
- Correlation does not imply cause relationship
Cross-Sectional Studies
- study/survey that examines relationship b/w exposure status and disease status at one point in time (SNAP SHOT)
- Exposure & disease are measured in same individual
- Measures exposure prevalence vs disease prevalence
- individual level data (unlike ecologic)
