CHLH 274Quiz 210/08 Lecture- Epidemiological studies are either observational or experimental - Experimental studies are considered true experiments and are the gold standard of epi research- Controls are assigned randomly in experimental studies - ways to categorize experimental studieso Individual vs community- treatment allocated to the individual or entire communityo Preventative vs therapeutic- prevent the disease or improve quality of life- Reference population- general group whom results of a trial should be applicable - Issues to be consideredo Sizeo Restrictions on whose eligible- Treatment assignment should be RANDOM- Goal of randomization is to achieve a baseline comparability between compared groups on factors related to outcomeo Compared groups should have same distribution of all characteristicso Equal distribution of known and unknown factors that are relevant response to treatmento Larger the group the better it works- Ascertaining the outcomeo Goals are high follow up rate/ don’t lose peopleo Uniform follow up for compared groups- 10/10 Lecture- Population: ecological studies- individualso case reportso case serieso cross-sectional studies- analytical studies- comparison groups- observational studieso case-control studieso cohort studieso experimental studies/intervention studies - person level data- looks are individuals and compare rates in the groups ( example smokers and non smokers and see lung cancer rates- aggregate- level (ecological studies)- classifies region as high level smoking region and other as low level smoking region and compare- Definition experimental study- involves the use of designed experiments to investigate the role of some agent in causation, prevention, or treatment of disease - Ambroise pare- randomized trial b/c used to put hot oil in wounds at war battles but ran out of oil, so some got the treatment and some didn’t and they saw those w/out oil were better- Randomized control= randomly split into 2 groups but decides who gets treatment or not - Cohort study definition- a study in which two or more groups of people are tracked forward in timer from exposure to outcome - Cohort stepso Start with a cohort group of individuals apparently free of the disease of interesto Cohort is divided based on possible exposure to the risk factor or lack of exposureo Followed through time to determine outcome of interest - Types of cohortso Prospectiveo Retrospectiveo Ambidirectional - Cohort studies are useful for exposures that have both short and long term outcomes- RR= (a/a+b) / (c/c+d)- Incidence rate/ density= # of new cases during a given time/ total person-time of observation - Framingham study looks at risk factors associated with hearth disease- Issues in the design of a cohort o All participants ( both exposed and unexposed) must be at risk of developing outcomeo Need a clear unambiguous definition of the exposure at beginning of studyo You need sources of exposure information and pre-existing records level of detail may not be accurate and not contain info on all variable of interest, people may not be detailed enough in an interview, o Who is the appropriate comparison and how can I make this an experimental study- Three possible sources of comparison groupso Internal comparison- unexposed members of the same cohort (people from same time and place)o Comparison cohort- a cohort who is not exposed form another similar populationo General population data- use pre-existing data form general populations as basis for comparison - For rare exposures you need to assemble special cohorts - Sources of outcome informationo Death certificates,o Physician, hospital, and health plan recordso Questionnaires (verified by recordso Medical exams - Strengths for cohortso Efficient for exposure rateso Good info on exposures ( prospective)o Can evaluate multiple effects on exposures o Less vulnerable to bias (prospective)o Can directly measure incidence or risko Clear temporal relationship b/w exposure and outcome ( prospective  directly measures incidence risk w/ cohorts - Weaknesses of cohorto Inefficient for rare outcomeso If retrospect, poor info on exposures and other key variables more vulnerable to bias o Expensive and time consuming particularly for a disease with long (blank) and latent periodo Validity of results can be influenced by loss to follow