University:
Johns Hopkins University
Course:
ME.250.699. | Medicine Elective
Academic year:

2023
Views:

81

Pages:

2
Author:

Yasin N.

Study DesignsisDescriptionAnalyticalDescube occuranceExamine etiology &g dls by time.NSD caused amountanPlane, personExperimentalVasi- experimentalNon. experimented(intermention studies)Innestigator lacksC observationalstuchesInvestigator intentionallyfull control overNo interventionthe interventionalter one I more factorbut conductsinvestigator obsento study the effectswithout inter mentsStudyUniont moluel tialsControlled trials-Cohort(Rot&PmTrial c out controlTenais circumCare-contaor companison group.groupCross-Seel"eg : Phare I or II clinal( ey Phane III (1.(mils)EcologicaltrialsUnit g ramdomizatncase -can-Individualor camonlycomm) Hield trails. unitHybridPrevalum surneysis - comm) christes- Nesled cas-contcase -series- care - whatSur vallance dath- Care - CrocianDescr $ analysis g- Senail crossroceturely collected dertaMsecunt(registries , mortality clataete)Randomized (RCTs)Quasi reconclomizedNon- raincommed- Allocation: using- Done cout- Intervention allocatedSchemies such asunderlyingrandomly. all participantsacc. to DOB,random procenname equal chance g benywo.g hospital records.allocated to each ofdate in which theyimmited to participateStudy groupin the study Relative ruilc (RR) / Prevalence natio / Risk entirMeasures how likely the exposed group will deneloa dis componed to imperposed groupRR = Inciclence in the exposed9/9+b=s(a) C+olIncidence in the unexposedIndesprotation :Participantle who smoked mere S times Whete hooto develop lung cancer compared to those who didnot smokeOdds ratioOdds g dls in enposed populhOE = 9/b.O dols g dls in non-enposed popul 15.Oo's c/d.odds g clls given emposureOR = OEOo.Bias in crocs actional study1. Selection bias - systemotic diff am thigh allocation & studyparticipants- Sampling biess : representativenessPrevented by- Prevaleme incident biar (Neyman him)Randomizeoceus when anglosme occus he annument disin mined- Rapons non response biers Couse clie earlySecreacy galhappento accur2. Measurement bies - systematin difference in morning outcomes- Misclamijid (prevension- Recall biasBlunding the- Lead time biasparticupantsin't measury- hength biased samplingculcomes