Class Notes (836,128)
United States (324,350)
BIOS 1500 (11)
Lecture 9

BIOS 1500 Lecture 9: Lecture_9.2_Prospective_Cohort_Design_2017

6 Pages
11 Views
Unlock Document

Department
Biostatistics
Course
BIOS 1500
Professor
Kevin O'brien
Semester
Spring

Description
Prospective Cohort Study BIOS 1500 Spring 2017 Prospective Cohort Study The problem is assessing whether an exposure increases the incidence of a disease. The measurement design is longitudinal. An assessment of disease is made of all subjects at the start, so that all are known to be disease free at the beginning of the study. Further assessments of disease are made as the study progresses and incident occurrences of disease are recorded. Prospective Cohort Study The design fixes as known: the total sample size, and the marginal distribution of the independent (exposure status): a+b and c+d. At the start we know how many are exposed and how many are not exposed and the total sample size. We also establish that all participants are evaluated and known to be disease free at the start. Prospective Cohort Study This design has forward direction due to the samples being based on the independent variable (exposure). The study then proceeds to establish the outcome (disease) in each sample. hence the directional argument is Forward: from knowing exposure to disease. Timing The timing is prospective due to the observation of new or incident cases of disease. The design is observational as no investigator intervention or randomization that is employed. We only observe what disease occurs in established or existing groups over time. Prospective Cohort Design The quantities known at the start are: A+B, c+d and the total sample size n. Cohort Study Estimation From such a design, you can estimate the Marginal distribution of disease The two Conditional distributions of disease: That given exposure and that for unexposed. It makes no sense to talk about the proportions exposed or not exposed as they were established by the research design. We selected a stated size sample of those exposed, and those not exposed. Cohort Studies An important aspect of this design is the ability to estimate risk or the incidence of disease. The hope is that the two cohorts are as similar as possible in all other ways except exposure status. If that assumption is true, then differences in disease incidence between the groups can be attributed to the exposure. Prospective Cohort Study All subjects are assessed to be disease free at the start. Each Subject in each exposure group is followed over some set time interval. Disease status is assessed at the final time point or as they occur. (new cases of disease =>Incidence). Longitudinal component. Cohort Studies Note that events in this type of study are incident events—new cases of disease. The longitudinal aspect (disease assessment at the start and at the end) allows the study to measure absolute risk (Incidence) of disease in each of the exposure cohorts. It also allows estimation of two measures of association: the risk difference, and the relative risk. Relationship to Experiments The Prospective cohort design is the observational design closest to an experiment. It has forward directionality and has prospective timing. BUT and experiment has two additional aspects: a planned intervention (exposure say) and randomization of participants to the levels of that intervention. Cohort Studies Risk is defined as the probability of getting the disease of interest during some specified time interval. This assessment is made under the knowledge that the subject did not previously have the disease (assessed at the start). Cohort Studies Risk is measured by the ‘proportion’ of those who get the disease relative to all participants in a sample cohort. This measure is called the Cumulative Incidence (CI). It refers to a particular interval of time: 1 year, 2 years whatever. Note that the cumulative incidence is a type of rate: change in disease status relative to change in time. (See Lecture notes L6 rates, ratios a
More Less

Related notes for BIOS 1500

Log In


OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit