EPIDEM: Chapter 9

Observational Studies

used to identify risk factors, and to estimate the quantitative effects of the various component causes that contribute to the occurrence of disease.

investigations

based on analysis of natural disease occurrence in populations by comparing groups of individuals with respect to disease occurrence and exposure to hypothesized causal factors

Observational Studies

the investigator is not free to allocate factors to individuals

Experimental Studies

the investigator is free to allocate factors to individuals at random.

Cohort study

animals exposed to a hypothesized causal factor (HCF), and a group not exposed to the factor are selected and observed to record development of disease in each group.

Case control study

a group of diseased animals (case) and a group of non-diseased animals (controls) are selected and compared with respect to presence of the hypothesized risk factor.

Cross sectional study

involves the selection of a sample of n individuals from a sample of a larger population, and then the determination, for each individual, of the simultaneous presence or absence of disease and hypothesized causal factor.

Prospective Study, Incidence study, follow-up study, longitudinal study

Other terms for Cohort Study (4)

Cohort Study

Analytic, Observational, Prospective, Longitudinal, Exposure-based

Cohort Study

Exposed grp and Unexposed grp (both free of disease of interest!)

Cohort

grp of people who share the same char.

Closed fixed cohort

fixed grp ffd up from a certain point in time until a defined endpoint w/o losses to follow up

Open or dynamic cohort

persons enter or leave the study anytime; changing exposures;

Prospective cohort

exposed and unexposed groups are followed up over time for the development of disease of interest

Prospective cohort

Study participants are free of dz, and at risk of dev’g d dz

Retrospective cohort

individuals who are free of the dz/outcome of interest at some point in the past are identified and classified according to exposure levels

Retrospective cohort

the specified follow up period has elapsed, both disease and outcome status of subjects have occurred

Selection bias

occurs when the grp actually studied does not reflect d same distribution of factors (age, race, smoking, etc) as occurs in d general population.

Follow-up bias

if the rate of disease among those lost to follow up is different in the exposed as in the unexposed group, then there will be distortion of the estimates

Confounding

variables which are both related to the exposed and the outcome which, when not taken into account, can cause the investigator to conclude that a factor increases risk for the disease when in fact it is not (or may fail to detect an association that exist)

Misclassification bias

results from inaccurate meas’t of factor or disease

Non-differential misclassification

occurs when errors are in similar proportions in the exposed and unexposed group, happens when same data collection method is used for both groups (bias usually towards null)

Differential misclassification

occurs when rates of error differ in the 2 grps (there may be an over- or underestimate of rates)

Confounding

effect of an extraneous variable that distorts d relationship between exposure (E) and disease (D)

Bias

systematic deviation from the truth (systematic error), consequence of study design, occurs at any stage of the research process