HLSC 4P99 Lecture Notes - Lecture 2: Prior Probability, Observer-Expectancy Effect
12 Feb 2015
School
Department
Course
Professor
Document Summary
Screening processes are not perfect- interference can be caused by the test itself or by the way the study was delivered. Don"t conclude on which paper is right based on how they concluded - all about the numbers. Based on the two articles but not based on your opinion. Need to talk about the disease before you evaluate the articles. Ramifications of not answering: you don"t know the utility, and it ends up not working well with the practice. If not a gold standard then it"s a reference standard. Two different tests, need to use both of them together. Ramification: reference standard- overestimation, no standard, we don"t know. Did this validation study include an appropriate spectrum of subject. Ideal spectrum- every possible person whom may have the disease is in the study. Ramifications: subjects have severe level of disease= overestimation, mild levels= underestimates. Bias- clinical trials are done in tertiary clinics, worst of the worst, sensitivity is overestimation.
