EPI 201 Summary.docx - EPI 201 Summary[Causal Effect We need to know the risk of the outcome in the presence of exposure in the absence of

EPI 201 Summary.docx - EPI 201 Summary[Causal Effect We...

This preview shows page 1 - 5 out of 13 pages.

EPI 201 Summary [Causal Effect] We need to know the risk of the outcome in the presence of exposure in the absence of exposure [Exchangeability] Exposed individual have a same risk of the outcome compared with unexposed individuals. After randomization & a large number of subjects [Randomization] No randomization Factors that make the two groups non-comparable A possibility that one group have different risk of the outcome c/t other Randomization Exchangeability & No confounding Strictly speaking, selection bias can occur (loss to follow-up etc) The distribution of the characteristics (that should be same at the time of randomization) Age, Sex, Gender etc After randomization, these characteristics do not change The distribution of the characteristics (that should not be the same) BMI, education Could be affected by the exposure and changeable [Counterfactual] Counterfactual outcome: What the outcome of the exposed group would have been in the study if they never had the exposure. Counterfactual reference (contextual specific, related to research question) The comparison of interest that we are making in a given study question. (If we observe the cognitive function at age 12 of a group of children exposed to high levels of lead in their homes from ages 0-10, the relevant counterfactual is the cognitive function at age of 12 of those same children had they not been exposed to high levels of leads in their homes from ages 0-10.) [Competing Risks] Event that removed a subject from being at risk for the outcome of interest (typically death)
Image of page 1
[Loss to follow-up] Voluntary or involuntary dropping out of the study
Image of page 2
[Measure of Occurrence] Prevalence: ¿ existing cases ¿ individuals the study population at a point in time (at age 50, in 1995, at study entry) Range: 0 to 1 Dimensionless ↑occurrence ↑duration ○Cross sectional ×Temporality ×Causal inference Incidence Rate (IR) ¿ incidencecases duringt 0 person timeat risk duringt 0 ¿ t 1 ¿ ¿ t 1 ¿ *incidence density *hazard rate instantaneous probability Range: 0 to ∞ cases/person- year closed cohort open cohort Cumulative Incidence (CI): ¿ incident casesduring t 0 ¿ t 1 ¿ ¿ individualsat risk at t 0 CI ≈ IR×Δ t ( ¿ 0.1 ) CI = 1 e IR × Δt Must state time period (Incidence proportion, risk) *Survival Proportion = 1−CI Range: 0 to 1 Dimensionless ○ short period f/u closed cohort × open cohort × loss to f/u × competing risk Odds: occurrence non occurrence Range: 0 to ∞ Dimensionless Case-control Cohort Q. The reason why prevalence of a disease may decrease while IR constant: The prevalence is a function of both incidence rate and the average duration for which individuals have the disease. The prevalence of the disease could be decreasing despite a constant incident rate because of a decrease in average disease duration (reduced duration: cured or died shortly)
Image of page 3
[Kaplan-Meier Formula (Product-Limit)] Inde x Tim e # Event s # Censore d # At Risk Conditional Cumulative
Image of page 4
Image of page 5

You've reached the end of your free preview.

Want to read all 13 pages?

  • Fall '18

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

Stuck? We have tutors online 24/7 who can help you get unstuck.
A+ icon
Ask Expert Tutors You can ask You can ask You can ask (will expire )
Answers in as fast as 15 minutes