Unformatted text preview: assess baseline behavior daily using standard tool behavior daily using standard tool Each Each child receives a soda daily; 11 times in 77 77 days a “challenge” soda with 7 food colors, otherwise, placebo. Double blind, placeboplacebocontrolled-Each child is own control Each child is control
Weiss Weiss et al, Science. 1980 Mar 28;207(4438):1487-9, McCann et al, Lancel 28;207(4438):1487370:1560, 2007
31 The The Experimental Study as as “Gold Standard”
Advantages: Disadvantages: Strongest Strongest evidence for Not Not “real life” causality causality High cost High cost If If randomized, selection Inappropriate/unethical Inappropriate/unethical bias and confounding bias and confounding for many questions for many questions removed removed Can Can be subject to poor If blinded, less observer If compliance loss to compliance, loss to bias followfollow-up Can pool results of Can several studies—“meta studies— 32 analysis” When When Can a Randomized Trial Design Design be Used?
Randomization Randomization is ethical when we do not know whether treatment is better than usual care. Need to resolve questions of benefit and harm and not subject subject people to side effects of unnecessary interventions or raise false hopes at tremendous expense. Need Need to obtain a truly informed consent. truly Need to plan for premature termination of a trial when benefits or risks become apparent early in the trial trial.
33 A Frequent Sequence of Frequent Studies in Human Populations
Case Studies/Ecological Associations/Case Series Sometimes referred to as a hierarchy of study designs. Cross Sectional Studies/Surveys Case-Control Studies Cohort Studies Randomized Trials
35 Skin Cancer Skin Cancer
Most common cancer in US Most common cancer in US > 1 million cases diagnosed in US each year. million 1 in five Americans will develop skin cancer in fi their lifetime. Basal Basal cell and squamous cell carcinomas are the two most common forms of skin cancer, but are easily treated if...
View Full Document
- Spring '08
- Epidemiology, Clinical trial, Randomized controlled trial, intervention intervention, Intervention Intervention Studies