Thursday, September 19, 2019; 8:45 AM - 10:15 AM
Section outline
-
Lecture 1: Course overview, the diagnostic process, and measures of interobserver agreement
Introduction to the course and how we will teach it.
Evidence-Based Medicine: 1) refining disease probabilities; 2) quantifying treatment effects.
Diagnosis: assigning the right name to a patient's illness versus using test results to guide treatment decisions.
Inter-rater agreement: concordance, unweighted Kappa, linear and quadratic-weighted Kappa.
Faculty: Michael Kohn
Location: Mission Hall 1400-
Watch URL
-
Lecture 1 from 2017 URL
1:09:30 Weighted kappa.
1:27:00 What's a good kappa?
On a laptop or desktop you should be able to play back at 2X speed. This does not appear to work on a mobile device. If you use 2X, you can watch from 1:09 to the end of the lecture in 10 minutes.
-
Skip from "Reiliability of Continuous Measurements" on Page 12 up to "Using studies of reliability from the literature" on Page 21.
This section covers Bland-Altman plots (which you will cover later this quarter in Epi 203) and calibration (which we will cover in Session 5 on "Evaluating Predictions").
-
In section on 9/19, we will discuss these questions. Try to think about them and jot down some notes for yourself, but we won't be collecting or grading your answers.
-
Kohn (2013) Understanding Evidence-Based Diagnosis File