SISPE Study Design & Variable Types

Cohort Studies

100

Three types of cohort studies

Retrospective, prospective, ambispective 

200

Measures of association that can be estimated

Absolute and relative measures of risk and rates

300

The statistical model that should be used to estimate the association between an exposure and binary outcome in a cohort study

Dist=binomial, link=log (relative) or identity (absolute)

Problems with convergence: 

Dist=Poisson, link=log,  repeated subject=id/type=unstr

Also the logistic regression IS sometimes OK to use in a cohort study with a binary outcome (rare outcomes) as a reasonable approximation of the RR- just not always. This is worth discussing in response to your prompt (Why shouldn’t a logistic regression be used?)

400

Sources of bias in cohort studies

Selection bias (differential loss to follow-up, collider stratification) 

Information bias (non-differential classification, differential misclassification, holding covariates fixed over time) 

Confounding (time-varying confounding)

500

True/false -- non-differential misclassification of the exposure always biases the effect estimate towards the null.

False

Case-Control Studies

100

Measures of association that can be estimated from a case-control study. 

Relative measures of effect

200

Controls in a case-control study should be sampled independently of 

Exposure

300

True/false -- case-control studies are just cohort studies done in reverse

False -- case-control studies are more efficient forms of cohort studies

400

Sampling hospital controls can introduce this

Selection bias - namely, Berkson's bias. 

500

Types of control sampling in nested case-control studies.

Density sampling (IRR -- no rare disease assumption needed), cumulative sampling (OR -- need rare disease assumption if interpreting as RR, otherwise "odds" interpretation remains), case-cohort sampling (RR --no rare disease assumption needed)

Cross-Sectional Studies

100

True/false -- Participants are sampled with respect to their exposure status.

False -- everyone is sampled at a point or period of time. 

200

Cross-sectional studies have trouble determining

Temporality

Exception -- if questions are posed to define temporality

300

Two types of cross-sectional studies

Descriptive and analytical

400

Outcome=binary

Model=logistic regression

Estimate=?

Prevalence odds ratio

500

1 strength and 1 limitation of cross-sectional studies

Strengths: typically inexpensive, population-based needs assessment, straightforward to analyze

Limitations: lead-time bias, limited temporality, other traditional biases (selection, information, confounding)

Variables

100

Definition of a confounder

1) Risk factor for the outcome among the unexposed

2) Associated with exposure in the source population

3) Does not mediate the exposure-outcome relationship

200

Definition of a mediator

An intermediate variable along the causal path between the exposure and outcome relationship

300

Heterogeneity of effect within strata of this type of variable

Effect measure modifier

400

Methodological and analytical strategies to control for confounding

Randomization (with sufficient sample size), stratification, regression adjustment, matching, standardization

500

Confounding or effect measure modification?

Crude OR = 2.0

Strata 1 OR= 1.2

Strata 2 OR = 2.8

Effect measure modification